EMERGENCY PLANNING FOR
INDUSTRIAL HAZARDS

Proceedings of the European Conference on Emergency Planning
for Industrial Hazards, organised by the Commission of the
European Communities (Directorate-General Environment,
Consumer Protection and Nuclear Safety (DG XI) in collaboration
with the Joint Research Centre (JRC) Ispra Establishment) and
held at the Congress Centre, Villa Ponti, Varese, Italy, 4–6
November 1987.
TECHNICAL COMMITTEE
A.AMENDOLA
A.BAUN
G.CAPRIULO
G.DEL BINO

CEC-JRC, Ispra, Italy
National Police, Denmark
Department of Civil Protection, Italy
CEC-DG XI (Chairman), Bruxelles,
Belgium
M.GENESCO
Ministry of the Interior, France
J.HEFFERNAN
Department of Labour, Ireland
R.KAY
Health and Safety Executive, Great
Britain
K.B.KRISTOFFERSEN Civil Defence Corp., Denmark
P.LAGADEC
Ecole Polytechnique, France
L.ALCON
Ministry of the Interior, Spain
H.J.PETTELKAU
Ministry
of
Environment
Nature
Protection and Reactor Safety, Federal
Republic of Germany
E.L.QUARANTELLI
Disaster Research Center, USA
A.SAMAIN
Ministry
of
Public
Health
and
Environment, Belgium
H.SCHNADT
TÜV Rheinland, Federal Republic of
Germany
J.NICOLAU
National Service for Civil Protection,
Portugal
MRS P.TESTORI
CEC-DG XI, Bruxelles, Belgium
G.VOLTA
CEC-JRC Ispra, Italy
C.J.VAN KUIJEN
Ministry of Housing, Physical Planning
and Environment, The Netherlands
M.VASSILOPOULOS
Ministry of the Environment, Greece
B.WYNNE
University of Lancaster, Great Britain
LOCAL ORGANISING COMMITTEE
H.B.F.GOW
MRS A.MANARA

CEC-JRC Ispra, Italy
CEC-JRC Ispra, Public
Press, Italy

Relations

and

iii

MRS M.P.MORETTI CEC-JRC Ispra, Public Relations
Press, Italy
T.SMYRNIOTIS
CEC-DG XI, Bruxelles, Belgium

and

EMERGENCY PLANNING
FOR INDUSTRIAL
HAZARDS
Edited by

H.B.F.GOW
CEC Joint Research Centre, Ispra
Establishment,
Ispra (VA), Italy
and
R.W.KAY
Formerly Health and Safety Executive,
London, UK

ELSEVIER APPLIED SCIENCE
LONDON and NEW YORK

ELSEVIER SCIENCE PUBLISHERS LTD
Crown House, Linton Road, Barking, Essex IG11 8JU, England
This edition published in the Taylor & Francis e-Library, 2005.
“To purchase your own copy of this or any of Taylor & Francis or Routledge’s
collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”
Sole Distributor in the USA and Canada
ELSEVIER SCIENCE PUBLISHING CO., INC.
52 Vanderbilt Avenue, New York, NY 10017, USA
WITH 19 TABLES AND 39 ILLUSTRATIONS
© 1988 ECSC, EEC, EAEC, BRUSSELS AND LUXEMBOURG
British Library Cataloguing in Publication Data
European Conference on Emergency Planning for
Industrial Accidents.
Emergency planning for industrial hazards.
1. Chemical engineering plants. Accidents.
Emergency action
I. Title II. Commission of the European
Communities. Consumer Protection and Nuclear
Safety. Commission of the European Communities.
Joint Research Centre. Ispra Establishment.
Gow, H.B.F. Kay, R.W.
363.1′1966028
ISBN 0-203-21605-9 Master e-book ISBN

ISBN 0-203-27238-2 (Adobe eReader Format)
ISBN 1-85166-260-X (Print Edition)
Library of Congress CIP data applied for
Publication arrangements by Commission of the European Communities,
Directorate-General Telecommunications, Information Industries and Innovation,
Scientific and Technical Communications Service, Luxembourg.
EUR 11591 EN
LEGAL NOTICE
Neither the Commission of the European Communities nor any person acting on
behalf of the Commission is responsible for the use which might be made of the
following information.
No responsibility is assumed by the Publisher for any injury and/or damage to
persons or property as a matter of products liability, negligence or otherwise, or
from any use or operation of any methods, products, instructions or ideas
contained in the material herein.
Special regulations for readers in the USA
This publication has been registered with the Copyright Clearance Center Inc.
(CCC), Salem, Massachusetts. Information can be obtained from the CCC about
conditions under which photocopies of parts of this publication may be made in

vi

the USA. All other copyright questions, including photocopying outside the USA,
should be referred to the publisher.
All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording, or otherwise, without the prior written
permission of the publisher.

The conference was arranged in six main sessions which dealt with organisational aspects. lessons learned from past incidents and providing information for the public. the Directive does not specify the way in which these objectives are to be achieved and one aim of the conference was to discuss the approaches adopted by the different national authorities and other organisations.Gow R.Preface Recent events worldwide have again highlighted the need for effective emergency measures for hazards connected with the process industries and the large scale storage of dangerous substances. has therefore organised the European Conference on Emergency Planning for Industrial Hazards in order to promote an exchange of information on the current situation.F.B. exercises and auditing. Within the European Communities. H. the Directive on the Major Accident Hazards of Certain Industrial Activities (82/501/EEC) requires that on-site and off-site emergency arrangements should be made for certain potentially hazardous industrial activities.W. The Commission of the European Communities. appropriate techniques. The conference also provided an opportunity to explore the considerable research effort which is going on throughout the world on the improvement of systems for emergency planning. design of plans. However. in collaboration with the JRC Ispra Establishment.KAY . DG XI.

The Italian Situation Concerning the Monitoring of Industrial Activities with Significant Possibility of Risk and the Availability and Application of Associated Exterior Emergency Plans G. Guide for the Establishment of an Emergency Plan J.UTH 2 2. Emergency Plans in France R.INNES 30 4.J.BOISSIERAS 44 .Contents Preface Opening Addresses G.R.BINETTI 15 3.DEL BINO Session 1: vii xiii xv Organisations Implementing Emergency Planning 1.CAPRIULO and L. Protection of Areas in the Vicinity of Hazardous Industrial Plants in the Federal Republic of Germany H. Emergency Planning in the UK: A View from the Inside G.GROLLIER BARON 41 Session 2: On-Site and Off-Site Emergency Planning Design 5.BISHOP G.

On-site Emergency Plans G.GENESCO 95 Session 3: Exercises and Auditing of Emergency Planning 13.O.D. Emergency Response Planning Off-site of Chemical Plants B.POILLUCCI 110 15.ix Emergency Plan and Alert System at MONTEDIPE L.M. Emergency Plans According to the Law for Protection against Catastrophes and On-site Hazard Protection Plans According to the Major Hazard Regulations W.ANTONELLO 57 7. Industrial Emergency Planning in The Netherlands H.ESSERY 61 8.MÜLLER 82 11. Emergency and Intervention Plans: The French Experience M.L.COONEY 127 .MORICI and G. Exercise Study for an Emergency of Chemical Origin G.STEUER 70 9. Plan for Off-site Exercises A.VAN DER KOOI and H. Co-operation in Emergency Planning T.DICKIE 77 10.KIER and G.MACCHI.VUYK 90 12.K. A.CORIGLIANO and F.PARANHOS TEIXEIRA 102 14.C. Effective Organisation and Incident Control W.

Including Fixed Installations H. Expert System Technology to Support Emergency Response: Its Prospects and Limitations S.A.MAX-LINO.SIGALÉS and A. P.NEUHOFF 147 18. B.DE WITT.TRUJILLO 158 20. Effective Emergency Planning Design by Means of Risk Analysis Models 190 .SENYÉ. Exercises and Auditing: Experience Gained in the FRG S.G.HAGEN and P.BRENK and A. Assessing the Response Capability and Vulnerability of an Emergency Plan: Some Important Issues R. Emergency Management of a Gas Escape C.G.D.KNAUP 175 22. The Computer Program TIGRE and its Application to the Planning of Chemical Emergencies A.HESEL. Auditing and Exercising of Emergency Plans for the Danish Oil and Natural Gas Transmission System.HARRISON and C. H. Improved Emergency Response after Release of Toxic Substances: Application of the System SMART D.BELARDO and W.x 16.M.RAMSAY 140 17. H.PIETERSEN 181 23.WALLACE 161 21.JOHANSEN 154 Session 4: Techniques for Emergency Plans 19.

SILVESTRI 24.ANDERSEN and J.VASSILOPOULOS 285 31. The Accident at DSM: Learning from a Major Accident in The Netherlands M.NEUHOFF 235 27.RASMUSSEN 211 Session 5: Lessons Learnt from Emergency Management of Major Incidents 26.J.QUARANTELLI 242 28. L. Decision Support Systems for Emergency Management V.TOFT 289 .A.SICILIANO and E.LAMBARDI. V.xi A.LIND ARPE 267 29.L. Major Industrial Risks: Examples of the Technical and Predictive Basis for On.TURNER and B. Experience Gained from Recent Major Accidents in the Federal Republic of Germany S.and Offsite Emergency Planning in the Context of UK Legislation K. Organizational Learning from Disasters B. Community and Organizational Preparations for and Re-sponses to Acute Chemical Emergencies and Disasters in the United States: Research Findings and their Wider Applicability E.VAN DUIN 274 30.CASSIDY 196 25. Experience Gained from the Oil Pollution Control Operation at Læsø in 1985 F.DONATI. Lessons Learnt from Major Fire Accidents in Greece M.

Requirements for the Planning of Industrial Hazard Alarm Systems with a View to the Application of Modern Communi-cation Systems W.J. Hazard Protection Measures in the case of the Release of Toxic Gases: Principles and Description of the Concepts W.GUTMANN 334 Concluding Session: Panel Discussion and Conclusions 36.GROLLIER BARON 325 35.HALPAAP 318 34.ULRICI and G.xii Session 6: Information to the Public Prior to and During an Emergency 32. Communicating Industrial Risk in The Netherlands: Principles and Practice P. Summary of the Concluding Session 350 List of Participants 357 Index 369 .STALLEN 308 33. Industrial Risk and Information to the Public R.M.

BISHOP Director of the Ispra Establishment. Within the Commission the appropriate bodies are DirectorateGeneral XI for Environment. the intervention of the Commission of the European Communities is natural and desirable. a good neighbours’ policy. especially after the Chernobyl accident. It was decided purposely to exclude discussion on disasters in the nuclear field and in the transport of dangerous materials. In . the latter will be treated in another specialised conference to be organised later.Opening Address G. the former has been extensively ventilated.R. Consumer Protection and Nuclear Safety and Directorate-General XII-Joint Research Centre. As background to this event I remind you that the EEC directive on major hazards for certain industrial activities (Post Seveso Directive) imposes that Emergency Plans be prepared to mitigate the consequences of major industrial accidents. but scarcely a week passes without some minor or major alarum underlining the need for a fresh approach towards effective actions. Good neighbours do not throw their trash into each other’s backyards! As an initial step towards the definition of effective measures for dealing with such problems the Commission decided to organise this meeting at which experts can explain what their respective countries and organisations are doing to cope with hazards in their process industries. with the steadily lengthening list of accidents which serve to alert public opinion and anxiety. I do not need to regale you. The Commission has the task of formulating those policies aimed at promoting the harmonious development of economic activity within the community of Member States. Joint Research Centre It is an honour and a pleasure to welcome you as participants to this ‘European Conference on Emergency Planning for Industrial Hazards’. the experts in these matters. Since the consequences of industrial accidents are rarely confined to the location of their occurrence. and notoriously do not respect even national frontiers.

In fact the Commission has adopted expert reccomendations that the JRC look for other clients. service work. They chose 36 papers for oral presentation and 11 for poster presentation. but also the protection of the environment and improvement of safety. The Technical Committee was chosen from Member State nominees. by placing its specialised neutral and independent scientific potential at the disposal of organisations or industries in the Member States by means of research contracts.xiv broad terms the relationship between any DG and the JRC is that the DG is the policy-formulating body which can call upon the JRC for scientific and technical research support. While the Commission is today the main client for JRC work. Commission policies thus guide the JRC in the orientation of its research activities and in defining its priorities: these are not only the establishment of the Internal Market and its corollary the improvement of industrial competitiveness. invited experts and staff members of DG XI and the JRC. Over one hundred papers were submitted and a selection made by the Technical Committee (whose composition is listed in the conference brochure) in Brussels last June for oral or poster presentation in Varese. . there should be a healthy interest therefore in the Conference proceedings which will be published by Elsevier Applied Science Publishers Ltd. Attendance also surpasses expectation since more than two hundred registrations are made. In February of this year (1987) a call for papers was made after extensive publicity. One of these means is illustrated by the present conference. I wish you a successful and fruitful conference and trust that you will make durable new contacts or reinforce existing ones in an endeavour to render in concrete terms the intentions of the Commission policies. cooperative projects industrial clubs and other suitable means. Similar care was exercised in proposing the session Chairmen and the session Rapporteurs to provide as large a representation as possible of the Member States. this does not imply an exclusive recourse to the JRC in the assignment of Community tasks.

Brussels. I must say that this conference could not have been timed any better. It has been said many times.Opening Address G. that the Sandoz accident has further confirmed the need for international. action to prevent major accidents and to limit their consequences. moving at 7 km per hour. and I believe it is worth repeating here today. organised by the Directorate General for Environment. . implies the necessity.DEL BINO Head of the Division of Chemical Control. Last Thursday over 20 000 people had to be evacuated from their homes. to improve effective emergency measures both for hazards connected with process industries as well as large storage and transportation of dangerous substances. never mind the health hazard. and in particular Community. and in the process of fighting the fire. early in the morning of 1 November 1986 in Basel. somewhere between 10 and 30 tons of chemicals were washed into the Rhine—a major environmental accident if there ever was one. Furthermore. 250 metres above the ground. which then led to a yellow gas cloud. a thousand tons of chemical products caught fire on the premises of Sandoz. to have. Belgium INTRODUCTION I would like to welcome you today. of course. only a year ago last Sunday. to this European conference on Emergency Planning for Industrial Hazards. Fortunately the wind direction was such that the gas cloud was pushed towards the sea. Apparently the fire broke out in an 850 ton storage silo containing ammonium nitrate compound fertilizer. a major accident can pose. Consumer Protection and Nuclear Safety and the Joint Research Centre in Ispra (Italy). to develop. so the heavily populated city of Nantes was not at risk. and where appropriate an obligation. near to Nantes (France) because there was a major threat from a poisonous gas cloud as a result of a fire. This. 5 km wide and 15 km long. Industrial Risks and Biotechnology (DG XI–A/2).

and furthermore. but it does not indicate the procedure and ways in which the emergency planning should be prepared and carried out. including safety equipment. EMERGENCY PLANNING The planning and experience which went into the development of the Directive on major accident hazards from all parties involved. it was not appropriate to define all the necessary technical and scientific standards and details with regard to emergency planning. on-site competence for emergency planning is clearly the responsibility of the manufacturer. account being taken of the responsibility of the manufacturer. emergency services or regional or central government. the Directive on Major Accident Hazards does request. is responsible for ‘ensuring that an emergency plan is drawn up for action outside the establishment in respect of the notified industrial activity’. Thus. the Federal Republic of Germany and the Netherlands had legislation on on-site emergency planning. in the notification to be submitted by the manufacturer to the national competent authorities. On the other hand. that is to say ‘emergency plans. while off-site emergency measures are for the administrative authorities responsible. the Directive only establishes the general requirement that both on-site and off-site emergency planning measures should be taken.xvi THE DIRECTIVE ON MAJOR ACCIDENT HAZARDS As you are well aware. Thus. was not entirely comprehensive enough to foresee all the problems. Many Member States have already a substantial number of onsite emergency plans in operation and the Commission hopes that . information relating to possible major accident situations. the Directive also requires that the competent authority. All the other Member States introduced new legislation or are in the process of adopting it in order to comply with the Directive in respect of on-site emergency planning. ON-SITE EMERGENCY PLANNING When the Directive was adopted in 1982. whether they be local agencies. alarm system and resources available for use inside the establishments in dealing with a major accident’.

IMPLEMENTATION OF THE DIRECTIVE At the same time the Commission is looking for an effective implementation of the Seveso Directive in all 12 Member States.xvii in the near future. Ireland) or a committee (Luxembourg. common data banks. and in the case of a major accident. research and training programmes. technical workshops of national industrial inspectors. regional or central levels— while the competent authority may be an inspectorate (UK. CO-OPERATION AND COMMON UNDERSTANDING The Commission does see an essential need for an exchange of information. investigations to monitor the national and local situations. This will facilitate the development of common approaches throughout the Community. at Community level. There are. The level of responsibility for drawing up a plan varies from Member State to Member State—local. since the Directive not only requires site-specific emergency plans which protect man. on all matters concerning procedure and scientific and technical aspects of emergency planning. legal. However. We. struggle to ensure the implementation in practice and in aiming to do this we undertake all kinds of activities: regular meetings with national authorities. tackling the problems. exchange of information at all levels. Italy) or a governmental department (Denmark. Greece. all Member States will develop such plans for the industrial installations covered by Article 5 of the Directive. Most Member States have had to introduce new legislation in order to comply with the Directive’s provisions. of course. but also the environment. This is necessary in order to promote co-operation and common understanding and appreciation of the difficulties involved in planning emergency measures. administrative and scientific. in the Commission. the Commission believes that it would not be correct or fruitful to turn its efforts towards achieving a complete harmonization of the varied kinds of emergency measures and of the different levels of authorities involved in the various Member States of the Community. Belgium) or a local authority (Germany). problems which are to be expected where there are 12 different national realities and where administrative practices. scientific and economic measures and development are all rather heterogeneous. .

together with this Committee. looks at the practical and legal aspects. the idea for a conference of this kind arose out of discussions in the Committee of National Competent Authorities responsible for the Directive on major accident hazards. As I look at the programme in front of me and all of you here today. be interesting discussion. Thank you. . this process only started a few years ago and we are continually making progress. good luck. and fruitful ideas. Indeed. we look on the technical side to exchange information and explore new ideas and avenues in order to identify areas of possible future action. While the Commission. I am sure that over the next few days there will.xviii However. indeed.

SESSION I Organisations Implementing Emergency Planning Chairman: A.F.Gow (Joint Research Center Ispra. Italy) .BAUN (Danish National Police. Denmark) Rapporteur: H.B.

Neither the legal basis of the emergency plan system nor specifie problems connected with it will be discussed here since this will be covered in the course of the conference by other speakers from the Federal Republic of Germany. Berlin. . An analysis of events showed that when on-site safety precautions fail effective emergency plans for the people living in the vicinity of an installation become particularly important [1–3]. such as the ones in Bhopal. 2 HAZARD ANALYSIS In order to provide protection from risks it is first necessary to know what they are. the Seveso Directive) which will be briefly described here. Man and the environment in the vicinity of hazardous industrial installations can be endangered by fires.1 Protection of Areas in the Vicinity of Hazardous Industrial Plants in the Federal Republic of Germany HANS-JOACHIM UTH Umweltbundesamt (Federal Environment Agency). a regulation which implements on a national level the EEC Directive 82/501/EEC. Emergency plans can be a matter of life and death. explosions and/or the release of toxic substance. For quite some time now they have been part of a comprehensive system to protect areas in the vicinity of hazardous industrial installations from the harmful effects which can result from a major accident. Mexico City and Basle. The most important elements of this system were laid down in the ‘Regulation on major industrial accidents’ (Störfallverordnung [4]. have in recent times made the risks associated with the modern chemicals industry unmistakably evident. FRG 1 INTRODUCTION Spectacular industrial accidents.

are regarded as hazardous installations. — identification of the influences which can lead to uncontrolled release of substances (hazard sources). population density. At present 17 types of installation are listed.g. state of the ecosystem).1 Identification of hazardous industrial installations in the Federal Republic of Germany In accordance with the stipulations of the Seveso Directive the actual scope of the ‘Regulation on major industrial accidents’ [4] is established by drawing up a list of the installations and substances in question. 2. become less effective due to interaction with internal and external influences (e. — the conditions in the surrounding area (e. A comprehensive analysis of the threat posed to man and the environment therefore requires — identification of industrial installations using dangerous substances and processes. These precautions can. . explosions. In the event of accidental release of a dangerous substance the consequences depend upon — the quantity released.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 3 The potential hazard is directly linked to the specific properties of chemical substances. — analysis of the structure of the area which could be affected. This threat can become reality in the event of uncontrolled release of such a substance. earthquakes. Only installations subject to licensing procedures. which are included in the list and in which substances included on the substance list are handled or could be formed in the event of an accident. In order to prevent this a number of on-site safety precautions are taken. however. The substance list includes 142 individual substances as well as 3 groups of substances which comprise some 150 further individual substances.g. — the dispersion behaviour of the substance. fire). flooding. The ‘Regulation on major industrial accidents’ may only be applied to installations in which the given substances exceed a certain threshold quantity [5].

Types of hazardous installation in the FRG. Around .2 Installation structure In all. In only 1 % of installations are dangerous substances not formed unless a major accident occurs (as was the case in Seveso). Of the installations registered. 1. StöVO refers to Störfallverordnung (German hazardous incident ordinance) [4]. licensing in the Federal Republic of Germany under the Federal Immission Control Act (BImSchG). The majority (99%) come under the regulation due to the fact that dangerous substances are present under normal operating conditions. This number corresponds to approximately 1–5% of the total number of installations subject to FIG. Figure 1 gives an overview. approximately 850 installations were registered with the authorities (as of autumn 1985). the registration obligation was waived for around 10% because their stock of dangerous substances was below the threshold quantity laid down.4 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 2.

Table 1 Relative placings of 20 of the most common substances a Appendix II. The proportion of installations which belong to the chemicals industry heads the list with around 60%. Accident decree [4].3 Categories of substances Of the 145 substances/groups of substances (positions) which could be subject to registration on the basis of the list of substances. . The major dangers connected with these installations are fire and explosion. Approximately 30% of all installations registered fall into the groups of substances ‘flammable gases’ and ‘highly flammable liquids’. 43%). In particular. 50% of the installations are registered under 5 positions. as present in installations in the chemicals industry and the petroleum refining industry. only 61 have been registered (approx. 2.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 5 95% of all registered installations can be assigned to 4 categories of installation. Approximately one-third of all installations which could be subject to registration on the basis of the installation list are not registered. 85% under 20 positions (Table 1). can lead to moments of acute danger in these installations [6]. the combination of toxic substances and substances which pose a fire hazard.

The most frequently occurring substances are — chlorine (14%). — hydrogen sulphide (11%). all substances listed in Table 1. — bromine (8%). have been registered. with the exception of explosive substances. Figure 2 shows the ratio of the stocks of substances which pose a fire hazard and those which are toxic in those installation categories with the highest potential danger. 2.6 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. This is based upon the overall installation structure. — alkali cyanide (7%). In the chemical industry. Average distribution of fire hazard substances in the chemical and petroleum refining industry. . not on individual installations.

Average inventory in relation to the threshold B according to Ref.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 7 FIG. installations where there is a fire or explosion hazard and those where there is a hazard posed by toxic substances are often in close proximity to one another. 3. It is estimated that some 30000 tonnes of toxic substances are . 5. An indication of potential danger is also the quantities handled. This must be taken into account when planning safety precautions (domino effect). In view of the fact that the 850 installations registered under the Regulation are located on only 150 sites.

3 SAFETY CONCEPT IN THE FEDERAL REPUBLIC OF GERMANY The first point to be considered is the substitution of dangerous substances by innocuous or at least less harmful ones. Many individual installations outside the industrial centres are distribution points for liquid petroleum gas [8]. which might occur as the result of a major industrial accident. 3). situated mostly in the northern regions of the country (on the coast). By definition this implies a high potential danger [7]. which formed the basis for the ‘Regulation on major industrial accidents’ (see Fig. Those within the chemicals industry are concentrated in the areas traditional to the industry in North Rhine Westphalia. it will not be discussed in depth here but it should nevertheless be seen as part of a comprehensive safety concept. integrated three-stage safety concept was developed. They limit the effects of harmful .8 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS held in stock in the installations in question (not including potentially toxic substances in the groups ‘highly flammable liquids’ and ‘flammable gases’). 5) it can be seen that the threshold quantity is in practice exceeded to a great extent (Fig. This basic approach has its roots in Germany in the federal legislation on chemicals [9]. Stage 3 includes the measures taken off-site to protect the surroundings. 2. Hesse. a comprehensive. The large storage plants for highly flammable liquids are. Rhineland Palatinate and Bavaria. 4): Stage 1 includes all measures in the installation which assure the safe containment of dangerous substances and the prevention of inadmissible operating conditions. Stage 2 brings together all measures designed to limit the effects of fire. in those cases where they are not operated directly by the refineries. If one determines the mean volume of substances per installation in relation to the threshold quantity B (for a definition of B see Ref. From the elements of a hazard analysis mentioned above.4 Distribution The 850 installations are located at 150 sites. explosion or the release of chemicals. Here there are often several installations in one larger complex.

Specification Safe enclosure of inventory Restriction of emission Restriction of adsorption Step I II III substances. Scheme of the 3-step security system. In it all factors relevant to the safety of the installation must be analysed and proof provided that the safety obligations have been fulfilled. In view of the fact that detailed safety measures are dependent on the specifie requirements of the plant and its location. 4. heat radiation or the consequences of an explosion on the objects to be protected.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 9 FIG. 10. This takes place in the form of the safety analysis required by the ‘Regulation on major industrial accidents’. (For requirements regarding the form and content of the safety analysis see Ref.) . every installation has to be regarded individually.

etc. — internal regulations. licensing of dangerous production plants only if they are at a certain distance from residential buildings.g. the Dangerous Machinery Act. Ref. Economic considerations are an important factor here (cf.e. separation of industrial and residential areas. constructing and operating engineering plant. Most problems are posed by the regulations in Stage 2. — regulations made by the employer’s liability insurance associations (accident prevention stipulations). 2). The density of regulations tends to decrease from one stage to the next and general principles take the place of concrete stipulations. etc. e. Their basis is to be found in the ‘technical compendium of regulations’ which has to be taken into account when planning. the Federal Emission Control Act and its Regulations. to limit their effects. the limitation of the effects of a major industrial accident. Included in it are — state regulations. This begins with actions such as location of industry.2 Off-site safety precautions (Stage 3) In the event that the precautions taken under Stages 1 and 2 fail. — regulations made by trade associations such as DIN. Vdt.1 On-site safety precautions (Stages 1 and 2) On-site safety precautions are designed to prevent major industrial accidents from occurring or.10 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3. 3. VDI. The safety aspects begin with the planning of an installation and end with the organization of the operating procedure. Only in exceptional cases is any mention made of this problem in the technical compendium of regulations. These principles are laid down in the Federal Republic of . The amount of regulations depends upon the safety stage in question. The precautions can be of a technical or organizational nature. i. the Commercial Activity Act (Gewerbeordnung) and the Regulations issued under it. if they do occur.. the Chemicals Act. The crux of the problem is that a hypothetical accident has to be assumed in order to design the parts of an installation correspondingly. VDE. the dangers then posed to man and the environment can be limited if precautionary measures are taken.

in which checks are made to ensure that an installation complies with the latest technological developments.3 Flanking safety precautions Experience has shown that it is not sufficient to issue safety regulations and bans in order to achieve in practice an optimal safety standard. They begin with the licensing procedure. as well as measures to increase the motivation of the employees to act in a safety conscious manner (including training schemes). first aid provisions. and extend to individual assessments of the safety of special components. emergency plans have to be implemented. 3.g. This necessitates honest explanation of possible dangers and maybe involvement of the public in the plans [12–14. so as to monitor whether the regulations are being adhered to. The point at which the responsibility of the plant operator ends and that of the authorities in charge of emergency plans in the area surrounding the plant begins must be accorded special importance. The local population must be prepared in advance so that they would know what action to take in the event of an accident. e. use can be made of the data given in the safety analysis as prescribed by the ‘Regulation on major industrial accidents’ and of the description of possible areas of danger which can be expected in the event of an accident. There are various possible ways of monitoring adherence to the regulations. In incidents in which internal emergency plans no longer suffice to control the accident off-site.g. 11). This means in particular that suitable substances for extinguishing fires. fire/ explosion. must be readily available. release of toxic substances. For emergency plans to be effective the participation of people living in the vicinity who could be affected by an accident at the plant is essential. Further precautionary measures are taken within the scope of emergency plans which begin off-site. measuring equipment. etc. 16 ]. Control instruments are also necessary.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 11 Germany in the Town and Country Planning legislation (see e. Ref. . Any emergency plans must take account of the type of danger. The relevant stipulations within the technical compendium of regulations require alarm plans to be established for the employees which include descriptions of appropriate behaviour. When drawing up plans. etc. On-site and off-site emergency plans must be carefully co-ordinated.

This necessitates provisions requiring the notification of accidents in industrial installations. the obligation to compile safety analyses which also applied to ‘existing installations’ (those already in operation before the Regulation came into force) meant that all installations within the scope of the Regulation were looked at from a safety point of view. in smaller quantities). The necessity to make specific emergency plans has not yet become generally accepted. especially in rural areas. The local and regional authorities bear the responsibility for disaster precautions. any particularly dangerous plants are situated as far away from housing as possible. 16]. it must be said that emergency plans are often inadequate. . 4 EVALUATION OF CURRENT PRACTICE Since the ‘Regulation on major industrial accidents’ came into force 6 years ago the discussion of industrial safety has intensified and the safety concept behind the Regulation has started to be accepted as a general principle. Furthermore certain concrete principles of the Regulation had an effect on other areas to which they were not formally applicable (e. Work is being carried out at present to set up a system for the central evaluation of these incidents. Apart from a few exemplary plans for large chemical works which take special account of the specific dangers connected with particular chemicals. Special precautionary plans for protection from dangerous installations are the exception rather than the rule. This is particularly true regarding the willingness to provide the public with frank and comprehensive information and involve them in the establishment of the plans. As far as Town and Country Planning is concerned it must be said that in older. In larger industrial complexes.12 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The analysis of accidents is extremely important for the further development of safety technology [15.g. In the course of this process a number of defects were corrected. In some cases the stocks of dangerous substances were reduced and in other cases harmless substances were substituted. ‘traditionally evolved areas’ there is still often insufficient separation of industrial and residential areas [17]. There are several problems of a legal and technical nature which can be observed in practice regarding emergency plans. safety precautions for the storage of liquid petroleum gas. In particular. In many cases disaster control teams are ill-equipped and inadequately trained.

7–8 November 1985. C. Seite 772. pollution of the atmosphere by toxic pollutants). Nr. 1st Bhopal in der BRD möglich? Sicker ist Sicher. 298–306. BGB1. these accidents have made it clear that the Regulation must be revised to include protection of the ecosystems (water. Verordnung zur Durchführung des Bundesimmissionsschutzgesetzes. zuletzt geändert durch Verordnung zur Neufassung und . H.6. In doing this. TNO-Dossier 8727– 13325. 19 November 1984. 2. care must be taken to ensure that the general principle of regarding systems as a whole and not just a sum of parts must be adhered to in all stages of the safety legislation. that the emergency plans and disaster precautions are tailored to the specific (chemical) hazard. Whilst the Regulation’s main aim was to protect society from acute dangers (fire.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 13 5 OUTLOOK Recent experience with major or more minor accidents in the chemicals industry has shown that the discussion about safety and protection of areas in the vicinity of industrial installations was not definitively concluded when the ‘Regulation on major industrial accidents’ was issued. Instructions should be issued to the competent authorities to clear up the concrete difficulties involved with implementing the ‘Regulation on major industrial accidents’ [19. and that accidents which are notified are systematically recorded and centrally evaluated in order to be of use in the further development of safety technology. REFERENCES 1. (1985). In particular. London.1980.-J. it would seem necessary to extend the scope of the Regulation. (1986). 27. explosion. Mexico City. 3. Analysis of the LPG Incident in San Juan Ixhuatepec. which is at present being undertaken in the Federal Republic of Germany. soil) [18]. These incidents have shown up very clearly where the limitations of the Regulation lie. International Symposium. 32. UTH. 12. 37(6). and the compilation of instructions for the competent authorities will contribute to better emergency plans for the protection of areas in the vicinity of industrial installations. 4. 6 May.1. PIETERSEN. 20]. The up-dating of the ‘Regulation on major industrial accidents’. The Chemical Industry after Bhopal.1980. to improve the stipulations requiring notification of major accidents and to tackle the problem of compiling a compendium of safety regulations which take account of the specific nature of individual plants. et al.M.

Deutscher Soziologentag. Jahresbericht 1985.04. Sonderschutzpläne—warum und wofür? ZS-Magazin. Nr.1977 MB1. 229. gwf-gas/ erdgas. Bielefeld. 7. 11. Kronberg. (1976). Seite 205. NW Seite 1688/SMB1. (1980). (1981). UTH. Loss Prevention in the Process Industries. Gesetz zum Schutz vor gefährlichen Stoffen (Chemikaliengesetz). 18. 34.(Hg. A. BUNDESMINISTER FÜR UMWELT. (1981). S. MAGS. High Risk Safety Technology. GREEN. 24. p. 15. Probleme bei Sicherheitsanalysen. METREVELI. Abstände zwischen Industrie—bzw. Einführung in die Sicherheitswissenschaft. I. KUHLMANN. 18. INSTITUT FÜR LANDES. Umweltbrief Nr. 2/3. (1983). Ausgabe A. F. 25.UND STADTENTWICKLUNG NORDRHEIN WESTFALEN (1981).E. Nr. 19. 32. ALBRECHT.G. (1986). LEES. IG Chemie-Papier-Keramik Nr. Einführung in die Soziologie der Katastrophen. 2. Seite 178. Köln. 13. Kohlhammer. GMB1. CLAUSEN.1974/2. Abstandsregelung in der Bauleitplanung. K. 1982. 1985. London.-J. 12. 14.P. 14.10. H. Naturschutz und Reaktorsicherheit (Hg). 41. UTH. 20.1985. November. Verwaltungsvorschrift zur Störfallverordnung. Butterworths. Allgemeine Verwaltungsvorschrift zur Störfallverordnung. 10. Vieweg und Sohn. . L. Seite 1718. 1981. NW 280. Seite 1586. Düsseldorf.9–1. John Wiley.11. 16. 12 February 1987. Rhein-Bericht. Seite 10. 6. 16 September 1980. I. A. H. et al. 1. H. Bonn.1981. 8. Änderung von Verordnungen zur Durchfüuhrung des BImSchG.7.-J.) (1982). Gewerbegebieten und Wohngebieten im Rahmen der Bauleitplanung Runderlass v. 23. 1978.14 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 5. 28. BMB1. 7. 127(6). Kommentar zur Störfallverordnung. 27 April 1982. Sechs Jahre Störfallverordnung—sind Chemieanlagen sicherer geworden? Gewerkschaftliche Umschau. 1980. SCHÄFER. DEUTSCHER FLÜSSIGGASVERBAND eV (1986). BGB1. 17. (12). Katastrophenstrategie und Partizipation. 9. (1987). BGB1. 17.

This really high number gives an idea of the size of the problem presented by the production. without also . and many of these present greater or lesser danger to man and the environment. Of these. employment. All this required and continues to require the need for man to be able to live with these sources of activity and of progress. production and use at the industrial level. insufficient industrial prevention produces or has encouraged. the occurrence of chemical accidents which have sometimes caused irreversible damage. often without knowing either their characteristics and properties.CAPRIULO Ministry of Civil Protection. It is thus necessary for the risks that they present to be evaluated at different levels. The awakening of a critical knowledge has sometimes led to reactions which are extreme and not at all rational. about 10000 are of great importance from the point of view of quantity. to identify the conditions of acceptability without prejudice to the preservation by the health authorities of man and the environment. Until fairly recently the presence and use of chemical substances were accepted passively. or even their suitability.2 The Italian Situation Concerning the Monitoring of Industrial Activities with Significant Possibility of Risk and the Availability and Application of Associated Exterior Emergency Plans G. however. Rome. with a very strong emotional component. consideration has been given almost entirely to the benefit derived from the use of chemical substances. in some cases. On the other hand. In the past. Italy 1 INTRODUCTION There are about 100000 chemical substances on the market. Italy & L. use and elimination of chemical substances. Rome.BINETTI Ministry of Public Health.

by adopting the safety and control measures which are indicated in the emergency plan. however. For several years.16 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS taking account of the consequences of damage. as soon as possible during the initiative’s design phase. 3 PREDICTION In this respect the community standards mentioned above have fixed a series of objectives to be achieved. the identification of critical points. Chemical substances are therefore of prime importance in the programmes of the CEC and also Italy. an attempt has been made to redress the balance by evaluating the risk which the production and use of a chemical substance might cause. 2 DIRECTIVES 82/501 AND 87/216 OF THE CEC These represent the standards which the CEC has adopted to prevent and deal with important accidents. The practical application of these principles is. however. and have also indicated some general principles to be followed. left to the initiative of the various member States who will make their own choice depending on the interior structures available. which have sometimes been catastrophic. the prediction and prevention phase and the phase of emergency action in the case of an accident. the prediction of combinations of events which could lead to accidents. The choices made in Italy have been dealt with in a coordinated programme according to which. and the introduction of associated safety measures. To summarise. — to prevent any accident that occurs from having disastrous consequences. the probability of accidents which could occur. i. and during recent years important and concrete initiatives have been set up including. following successive stages which .e. in the first place. by conducting research into possible causes. the CEC Directive 82/501 which has recently been modified by Directive 87/216. The directives have two aims: — to detect. the Community standards treat the problem of industrial risk with a single approach and envisage two different phases.

The stages of this programme are as follows. at complete monitoring of the entire industrial sector involved in the problem of the risk of serious accidents. All the data collected (which are relative to about 10000 industries) have been examined. At the end of this last procedure the list may be conclusive and in line with the needs which it must control. one must arrive. The overall response obtained was considered satisfactory. in a reasonable time. List C collects together all the other industries which have provided data. evaluated and finally divided into three different lists: List A collects together all the industries which perform activities of deposition and/or production in the installations indicated in Appendix I of Directive 82/ 5101 and which use the substances listed in Appendices II and III of the directive but in quantities greater than the levels indicated. List B collects together all the industries which perform activities similar to those of list A but where the quantities of substances are lower than the same levels. which took place under the control of the Ministry of Civil Protection. 3. it was very close to the real national situation. one based on a system of self-declaration by the industries which took place under the control of the Ministry of Public Health.1 Identification of the national list of industries concerned The programme was started in February 1985 with two different procedures. in fact. At present the list is being revised to take account of the modifications made with Directive 87/216. the other based on the direct collection of data from industries identified with the banks available at different central and local public offices. by using other sources of information available. . In this respect a programme of inspections is underway which is being organised in coordination with the central and local public State organisations. In any case one should remember that for all undeclared situations the penal code will be applied.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 17 are linked together.

in September 1986 each industry involved was asked to prepare. 3. (b) The industrial activity is not totally acceptable but may be made to conform to acceptable levels with the introduction of new measures. in any case it is predicted that it will be complete by July 1989. in general.4 Evaluation of the safety reports and classification of the industries depending on their risk level The safety reports are being examined by the Italian public authorities. this programme has been given the biggest impulse in considering also the technical requirements of industries. the following cases: (a) The industrial activity is acceptable from the point of view of interior and exterior safety. safety reports for all the industrial activities to be notified.e. an appropriate technical guide was made available to all the industries. As a result. according to a programme which has fixed priorities for the most dangerous situations either because of the complexity of the installations or because of their proximity to an inhabited centre. At the conclusion of this examination.2 Risk maps Using data from the list. a final judgement will be given which will apply. the activities covered by Article 5 of Directive 82/501. To facilitate this procedure and make the various reports homogeneous.3 Safety reports As they form the fundamental knowledge base for the definition of the external emergency plans. (c) The industrial activity is not compatible with the place. 3. . i. The verification programme must finish as quickly as possible. as well as having available a complex series of data and technical information. it has been possible to define the first risk maps which take into account all of the most dangerous industrial activities. within a year. and consequently must be closed and/or moved.18 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3.

after having evaluated all the installations concerned. In this programme. all the useful information. one can easily see the importance of being able to bring them under uniform control. 4 PREVENTION When each safety report has been examined and evaluated. The possibilities . which could differ from case to case. A formal communication must be sent to the public authorities. must establish if it is (A) included in the field of application of Articles 3 and 4 of Directive 82/501. As the ratio of these activities to notified activities is roughly 10:1. for those covered by Articles 3 and 4 the situation is less well defined. if this is the case it must be established if the greatest existing risk — is limited to inside the establishment. according to the case. measures for closing it and at the same time for providing economic support must be adopted. it would be a good idea to know about all possible risk sources which could have some effect. Considering that emergency plans must be set up for these activities. If the activity is considered incompatible at the internal and external safety level. — could also affect the outside. first of all. to adopt all the measures. In fact these activities are included in the areas where there are almost always activities to be notified too. a simplified procedure has been laid down according to which each industry. it will then be necessary.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 19 3. The Italian authorities are therefore in the process of deciding on a programme for the monitoring of the safety of all industrial activities covered by Articles 3 and 4 of the directive. (B) not included in the field of application of Articles 3 and 4.5 Industrial activities with no obligation of notification (Articles 3 and 4 of Directive 82/ 501) While for activities to be notified the CEC directive has laid down a series of actions which allow the public authorities to have available. which should come into force on 31 July 1989.

dry distillation of coal and lignite. 5. then a series of appropriate technical measures must be adopted which the industry must satisfy within an appropriate period. production of metals and metalloids. If the activity is compatible with the territory but the safety level is not considered sufficient. a national sectorial plan called CHEMIC has been prepared to come to the aid of the public and the environment in the case of a disaster due to industrial risk. CEC 501/82. petrochemical.20 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS of moving it to a more suitable place must also be evaluated on the technical and policy level. 5 THE EMERGENCY PLANS The obligation to furnish external emergency plans for the industrial activities. and gas and inflammable liquid depots). Only after this complex programme has been completed will it be possible to define the external emergency plans with the precision necessary. As has already been said. on the other hand. While it is being carried out. the manufacture and storage of explosives. and their storage inside and outside the establishment. Such plans must be based on (a) elements contained in the safety reports on the evaluation of security. on their evaluations and on the measures consequently adopted. The CHEMIC plan also considers power stations. both phases of the survey programme will be completed in July 1989. as is that of giving the necessary information to the public. is the responsibility of the Prefects. In the range of industries mentioned above. (b) the elements deduced from simplified communications laid down for industries which are not at high risk according to Articles 3 and 4 of CEC 501/82. powders and munitions. release of solid or liquid substances by combustion or chemical decomposition. installations for the disposal of toxic or dangerous wastes. treatment of energy gases. The CEC directive refers only to certain categories of industry (chemical. which are covered by Art. extraction activities and other mining activities. it provides a framework for provincial and local plans. the CEC directive only considers the activities of transforming and handling toxic substances. . The transport of these goods is.

— flexibility. The transport of dangerous substances through oil and gas pipelines. lakes and pools.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 21 governed by the RID (International Regulation for Railways). The general criteria to be followed when drawing up and organising plans are well represented by — clarity and conciseness. — human resources and materials available. or explosions. — hierarchical organisation. — concrete definition of tools for emergency management. may lead to pollution of watercourses. not ratified by Italy) for sea transport. FIATA has created FIAT-SDT (shipper’s declaration for the transport of dangerous goods). and by IATA (International Air Transport Association) standards for air transport. (b) intervention phases: — pre-alarm. — alarm. — revision and updating of emergency management. following losses or breaks. by the International Code of IMCO (Intergovernmental Maritime Consultative Organisation. — involvement of all public organisations. it laid down (a) procedures: — definition and location of forces available. Law 979/82 was passed to defend the sea from pollution. The following emergency actions are laid down for accidents which occur in various situations: . who is the most appropriate person to warn the carrier against the dangers which the goods described may pose to people or things. — criteria of use. — mode of intervention. From the operational point of view. by the standards of the European agreement (ARD) stipulated at Geneva in 1966 for road transport. common to any type of transport which is the direct responsibility of the compiler. The accidental or deliberate pouring of hydrocarbons into the sea or the leakage or destruction of containers of toxic substances into the sea because of shipwreck may cause pollution of beaches and territorial or international waters.

etc. —put health regulations into practice by advising the public and giving instructions for behaviour. absorb.). — protection and re-establishment of water supplies and public services (bridges and embankments. (b) In means of transport. etc. ENEL.). . municipal enterprises. — use chemical or other solvents. if necessary. (c) At sea or in ports: The harbourmaster has primary responsibility for actions. recover. He will act with his means and with those of the fire services and other administrations who have the necessary means. etc. or dispose of the pollution where possible. — containment (fire services. — demand for consultation with experts. —in the case of pollution which is lighter than water and which does not mix with it. — take samples for analysis on the surface and in the pool. analysis of the help of intervention by firemen. — protect agriculture and livestock.22 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (a) At establishments or depots: — evacuation. — clean the beaches. — re-establishment of road and rail networks (ANAS and FFSS). —set up temporary dams. floating barriers or other forms of barrage. specialised actions. — neutralise it chemically where possible. suck the surfaces with pumps and drainage pumps. — isolation of the zone (police. — chemically identify the dispersed substance. —contain. and will thus proceed to — isolate the zone. SNAM. — the technical action of plugging and non-pollution will mainly be the responsibility of the manager. — stop the drinking water supply if necessary.). oil and gas pipelines: — as in the preceding case.

Those already in operation and which are of particular interest for industrial risk include (a) The data bank at the local level which collects together. and the software which includes the departmental data bank and the mathematical models. the CASI (Centre for the Application of Software Studies) which has a data processing centre. high risk industries. suitable and efficient action in the case of an emergency. 6 DATABANKS To guarantee rational. Many systems are already in operation and others are being developed. by introducing. This particularly concerns potential areas where the risk factors are aggravated by the movement of dangerous goods by sea and by land and by the existence of a pipeline. initiatives are underway to set up a global approach for certain areas where there are important industrial installations which involve dangerous activities not covered by the CEC directive. either statistical or related to resources. on the basis of a stored distribution of resources. which is used to evaluate the needs for assistance by locating the nearest available resources and suggesting the best route for the arrival of assistance. (e) An assistance model on the basis of a diagram reproducing the road network schematically and. Under development are . the disposal of toxic wastes. It manages the hardware which comprises two computers. the Civil Protection Department uses a specific service. (c) An automatic map-drawing programme which can create thematic maps using certain programmes. graphic equipment and terminals. etc. in real time. on a statistical basis.. debris. (b) The risk source data bank which collects together information from surveys performed by the CPD (or for the CPD) on dams or barrages. (d) A meteorological model which is used to provide. humidity etc. some hundreds of items of information. a spatial distribution of temperature. for the 8050 Italian local authorities.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 23 In addition to these general standards. the data detected by the MA (Military Aviation) meteorological stations.

and in particular with the Health Ministry (availability of health structures) and with the Superior Health Institute (dangerous substances data bank). (c) An accident data bank when catastrophic events occur. assistance models.24 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (a) The perfection of area data bases already available. the Health Commission. seismic models. which records and processes all the news on accidents and events in Italy and abroad which are of interest to the CPD (Civil Protection Department). thanks to portable processing systems. (b) The organisations which assist the Minister in his activity: EMERCOM (Operational Committee for Emergencies). (b) An accident and event data bank. . (d) The development and improvement of existing models. and management of health and logistics assistance for them. Emergencies. Now a Minister for the Coordination of Civil Protection has been appointed. it has no special personnel or budget and is the only example of a State organisation which acts by function and not by subject. the Committee for voluntary activities. This department includes (a) The Cabinet Office. Budget and AAAA. which allows deduction from files on site. (e) Liaison with data banks of all the administrations of the State concerned in the emergency. the Coordination Committee for activities concerning safety in the industrial sector. 7 CIVIL PROTECTION STRUCTURES The organisation of Civil Protection at the central level is made up of a Department of the Presidency of the Council of Ministers which is responsible for coordination. Public Works. (c) Four services: Coordination. and meteorological models are systematically updated and adapted in the light of experience acquired when real events occur. of the number and type of people injured following a catastrophic event. is being planned or set up. and the report of all the data no longer by administrative unit (Local Authority) but by a specific geographical point referred to by its coordinates.

State water services.) and telex links (RAI. and maintain contacts with regional and provincial operation rooms.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 25 EMERCOM plays a particular role. for fighting fires.) — of Transport* (i. responsible for the Academies) — of Industry and Commerce (i. It coordinates the administration of all the organisations and the emergency service. the road network) — of Education (i.e. press agencies). normal telephone links. Minister of the Interior). railways. Fire services. public works.e. (d) The Sanitary Emergency Service which deals with the sanitary aspects of each event. links by radio (fire service. etc. responsible for bridges and embankments. particularly by aeroplane. etc. which has already been described. (c) The COEM (Sea Emergency Operations Centre) which coordinates and plans action at sea in the case of serious pollution or an aeroplane accident. Organisations subordinate to the Department help in the coordination of actions: (a) Ministry of Defence* (b) General Command of Police and Financial Agents (c) Police (d) Fire Service (e) Ministers — of the Interior* (i. responsible for Provincial Work Offices and the Regional Work Inspectorate) . responsible for the Prefectures. respond in the first instance to safety requirements.e. ham radio operators. it has a telecommunication centre which has telephone links pointby-point (armed forces. Carabinieri. (b) The COAU (Unified Air Operations Centre) which directs assistance by air.e. Police. inform Department heads of events which are occurring. (e) The CASI. The emergency service has the task of ensuring rapid action and assistance to people affected by catastrophes. it puts the Minister’s orders into practice and looks after liaison with the scientific organisations which monitor major risks using the following operational services: (a) The CESI (Situation Centre) formed of a group of operators which receive and evaluate news 24 hours a day.

From the operational point of view. the peripheral structures of organisations described above go into action each with prescribed tasks. in the case of an emergency. * Ministers who have their own representatives in EMERCOM. responsible for the Directorates of Post and Telegraph. The peripheral Civil Protection organisation is formed of (a) At the Regional level: the person regionally responsible for civil protection and the Regional Operations Centre. — SNAM (national company for water and gas supply=ENI).e. national companies for telephone services) — of Agriculture and Forests* — of Health* (i. (c) At the Local level: the Mayor and the Local Operations Centre. responsible for national health structures) (f) Associations of Volunteers: — International Red Cross (IRC) — Ham Radio Operators’ Centre (CER-ARI) — Italian Alpine Club (CAI) — Caritas — Sovereign Military Order of Malta (SMOM) (g) the large service companies such as — ENEL (electricity). responsible for the harbour masters) — of Post and Telecommunications* (i. (b) At the Provincial level: the Prefect and the Provincial Operations Centre.26 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — of the Environment — of the Merchant Marine (i. .e.e. — etc. — SIP (telephones).

through the Prefectures. those which belong to the second must satisfy special conditions. obligations and responsibilities concerning the use of materials. can be caused by chemical. of emergency and helping services. productive programming. monitoring procedures. precautions to be adopted to avoid internal pollution which. procedures for monitoring the safety of the environment and the state of health of workers. checks and needs of the working environment. as well as external. The Minister of the Interior takes charge. training courses. In this case it covers all the aspects concerning ndustrial risk. with health and safety in places of work. such as the unitary character of the safety objectives in the place of work and life. The first is represented by the Minister of the Interior (Law 996/70. 24 of Law 833/78 delegated the Government to issue a Unique Text on the subject and laid down the directives to which the exercise of the procural must conform. The second group is at the Ministry of Health and the Ministry of Labour (TU LLSS 1934. selling and using dangerous products. In any case the installations must be constructed respecting numerous standards which concern the environment.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 27 8 CIVIL PROTECTION TODAY AND THE SNPC ACCORDING TO THE DRAFT BILL BEING EXAMINED BY PARLIAMENT The standard is applicable to four main groups: 1. in Articles 216 and 217. The following have been prepared: updating of standards. special procedures for specific .833/78 and DM23 December 1985). physical and biological poisoning factors. whether they are the responsibility of Regions or other institutional organisations or whether they are the responsibility of the Minister for the Coordination of Civil Protection when he decides to intervene with extraordinary powers. DPR 303/56. 2. criteria and ways of acting when there is a serious and imminent risk. The authorities may refuse authorisation for the setting up of installations or impose special conditions. DPR 66/81) and concerns all the situations which involve serious damage or danger for people and goods. in particular. activities and fire. Art. L. They deal. In particular the TU of 1934. ways of producing. classifies industrial installations which produce vapours which are unpleasant and dangerous to general health in two groups: those which belong to the first must be sited in areas away from inhabited centres. assistance for people affected by natural catastrophes or accidents.

on 18 Dec. with the task of: analysing the risk situation and of drawing up safety reports. regional and local authorities. with greater efficiency than national. training of personnel and safety of places of work. which has the same aims. However. on the site of the emergency. Finally there is the DM of 23 December 1985. In this case the Minister may himself take charge of action against the emergency. which because of its seriousness and extent cannot be tackled by the administrations in the field of their ordinary competence.28 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS risks. Those concerning wastes (445/75 and 319/78) have been received (DPR 915/82 and Order of the Interministerial Committee in Art. 1985 of the Health Ministry. setting up organisations for the prevention. must work as part of this committee. 4. Civil Protection is a pyramid-shaped structure with territorial base. 1985. in relation to the nature and gravity of this event. therefore. institutes and voluntary organisations for preparation in case of emergency. divided into successive channels. to satisfy each need of people affected by an event. checking safety reports and emergency plans. having exceptional powers. The third group concerns the decisions taken by the Minister for the Coordination of Civil Protection who. the President of the Regional Court and the Mayor may issue contingent and urgent orders on hygiene. control and monitoring of industrial installations. Overall. risk cards and internal and external emergency plans. 3. starting from the local step. ways of determining and updating limiting values for noxious factors. reorganisation of public offices and services responsible for safety at work. the temporary standards which already exist for industries. organisations. which come into action by steps. the procural has twice reached expiry and must be renewed. public health and veterinary policy. encouraging Regions and Prefectures to adopt uniform criteria in the development of administrative action. The Commission created by the DM of 23 Dec. The fourth groups concerns the CEC directives. coordination between State and territorial organisations and the definition of problems of standardisation. coordinates all the public administrations. 5 of the same DPR) and those which concern industrial risk (501/82 and 216/87) have not yet been received. . Article 32 of the same law lays down that the Health Ministry. The Minister also presides over the committee which coordinates the security activities in the industrial sector created with CPDM.

1982 by Deputy G. 1980. as well as through the promotion of civil voluntary actions. As has already been said. The Unified Text of the draft law presented on 5 Feb. a State of Emergency is declared. and approved by the II Permanent Commission. 23 Nov. is intended to perfect the organisation of Civil Protection by eliminating the gaps in the present legislation which emerged during the earthquake in Irpinia. In the draft there is a substantial revision of the organisation criteria presented in 996/70 and a new orientation towards the creation of a modern conscience on Civil Protection through the psychological and operational preparation of the public. Provinces and territoral and institutional public organisations will participate with the State in setting up the SNPC.Spadolini. The Regions. the power of the orders is put into action for him and he can use the Fund for Civil Protection.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 29 If such an event becomes a disaster or a catastrophe. and if the resources prove insufficient to satisfy demand. . In this text assistance is not the only component of Civil Protection: prevention is re-evaluated. Even if it remains the responsible organisation. the Region and Province have a permanent function of coordination on their respective territories. and the mayor and local community are of great importance. the structures for the management of smaller emergencies are the basis for interventions in catastrophes. Zamberletti and Senator G. the Minister for the Coordination of Civil Protection assumes responsibility for action. voluntary organisations are given the chance to participate in the Committees laid down by law 996/ 70 in relation to prevention activities.

generally. UK HISTORY First. Yet. over the years. In the UK no single organisation is responsible for making contingency plans to deal with major emergencies whether for peace or war. Civil defence expenditure by local authorities is . by virtue of political cohesion and the judiciary the bureaucracy has.3 Emergency Planning in the UK: A View from the Inside GEORGE INNES London Fire and Civil Defence Authority. Emergency planning for peace is referred to simply as ‘civil emergency planning’. Local authorities therefore prepare emergency plans for peace (civil emergencies) and war (civil defence) and are empowered by Parliament to spend public money to these ends. somehow or other. it will be helpful to explain the terminology used. London. In this connection it is noted with interest that a draft document published by the European Commission in April 1987 covering civil emergencies contained references to ‘Civil Defence’ policies. in the final document issued in June under reference 87/C176/01. Traditionally. house the homeless and. This may seem a bit untidy but could be said to be typically British. In 1986 the Government decided that the term ‘Civil Protection’ would be adopted to cover emergency planning for both peace and war. However. the generic ‘Civil Protection’ has been substituted for ‘Civil Defence’. in the UK the term Civil Defence means emergency planning for war. Unlike most other Member States of the European Community. generally produced effective programmes to implement the decisions of democratic government. It is the responsibility of local governments to feed the hungry. It is common knowledge that the UK has no written constitution. restore the aftermath of a major emergency to normality as soon as practicable. the primary responsibility for caring for the local populace—whatever the contingency—has been devolved from central to local government.

so far as people and property in the vicinity are concerned. Sir John Anderson’s historical assertion remains valid. in formulating civil defence policy. THE DEVELOPMENT OF CONTEMPORARY POLICIES The intellectual development of British contemporary civil defence cannot be easily broken into discrete phases. Sir John Anderson—after whom a domestic air raid shelter was named—is on record as saying: ‘I cannot too strongly emphasise that whilst it is the Government’s business to fight the war. However. so the Air Raid Precautions (ARP) system came into being during the war. came to much the same conclusion as its predecessors had in planning to deal with Napoleon’s threatened invasion of 1803/4. The then Lord Privy Seal.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 31 almost totally refunded by central government under a grant-inaid scheme. Thus. Perhaps understandably for an island people. in 1914 a government committee. the consequences could be remarkably similar. Civil defence in the United Kingdom has quite a long history. For instance. Shortly before the Second World War a number of publicspirited people in politics. Emergency planning. Even then the planners had looked back to the arrangements made to deal with the Spanish Armada of 1588. as for what people are to do it is the business of the community to prepare itself’ (author’s emphasis). industry and commerce became determined that the British people should not have to face the prospect of aerial attack as unprotected against its effects as the people of Spain had been during their civil war. for peace and war. the causes may be different but. One could easily go on to draw the parallel here between a bomb dropping on a populated district and a 10 tonne road tanker full of flammable material exploding and burning. the armed services. Thus the first links were forged between local government and its now well established responsibility for co-ordinating responses to major emergencies whatever their character. there has been a natural tendency to look to past examples which then serve as precedents for the future. established to consider what should be done in the event of an invasion. arrangements for the defence of our shores by the populace as a whole have figured prominently in our history since the Middle Ages. hence. there are . This organisation was locally funded with links into the respective local authorities and enjoyed government support. is local government business.

the Home Department issued a policy circular to local authorities (since embodied in that department’s Emergency Planning Guidance to Local Authorities. namely. Another consequence of the Flixborough accident was that the Government’s Health and Safety Commission appointed the Advisory Committee on Major Hazards who. or so-called ‘catch all’. the Local Government Act 1972 already empowered local authorities to incur expenditure to ‘avert’. However. The post-war years have seen tremendous strides in modern technology leading to the development of new products and materials. Local authorities were thus inter alia recommended to prepare general local emergency plans to deal with any emergency. At least the main ingredients of the plan are available for implementation. and partly in response to the major accident which destroyed the caprolactam plant at Flixborough. Unfortunately a by-product of these developments is the growing dependence of industry and commerce on the use of hazardous materials. For example. the 1968 dismantling. ‘alleviate’ and ‘eradicate’ the effects of any emergency or disaster which could be foreseen as representing a danger to life or property within their jurisdictions. and the 1980 ‘enhanced commitment’ subsequently reinforced by statutory Regulations made in 1983 and monitored since then by the Home Department in what is referred to as the Planned Programme for (their) Implementation (PPI). the 1972 care and maintenance policy. Lincolnshire. 1985) suggesting that officers employed on civil defence planmaking should be employed on making plans to deal with civil emergencies also. general plans do suffer from some selfevident limitations in that they are not specifically addressed to a specific scenario in a specific locality. because it is to match them that a body of professional emergency planners has emerged (specifically in the top tier local authorities) upon whom falls the responsibility for making emergency plans. Partly as a consequence of this. in 1976. if a school is identified as a reception centre for those who suffer as a result of a local accident. . in 1974. it matters little whether the activation of this part of the plan has been occasioned by an explosion in a nearby site or a chemical spillage from a road tanker passing through the locality. These developments are especially relevant. In addition to the promise by the Home Department of grant-aiding the staff costs for a modest amount of this civil emergency activity.32 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS some identifiable turning points. These general. not only for the war emergency but also for other major emergencies. plans are still in being in many authorities and have proved their worth. the Civil Defence Act 1948.

London had its own government. During 1985 in Greater London. became law in 1983. with the enactment of the Civil Protection in Peacetime Act. enforces the regulations. 7 site-specific plans were made by the Emergency Planning Division of the now abolished Greater London Council. came the coining of that term. In pursuance of these regulations some 220 such sites have been identified by the Health and Safety Executive which. among its other functions. for example. namely fire-fighting and civil protection embracing preparedness for peace and war. liquid petroleum gas and natural gas sites. The Committee also recommended complementary changes to planning legislation to deal with development at. but with the reorganisation of local government in 1986 the GLC and the six Metropolitan County Councils in England were abolished and their duties were inherited in the main by a lower tier of local government. such installations. Two exceptions were fire and civil defence (including CIMAH) which passed to successor Fire and Civil Defence Authorities. 1902. Thus the London Fire and Civil Defence Authority. the now well known Seveso Directive was issued in 1982 by the European Council. in which the author is serving. in turn. . the Greater London Council (GLC). The costs of the emergency planning staff are met by 100% grant-inaid from central government. THE SEVESO DIRECTIVE AND THE UK RESPONSE Following the disasters in Italy at Seveso and Manfredonia. Also in 1986. requiring site-specific plans to be made for certain specified industrial hazards. It is of interest that. This culminated in 1978 with the issue of draft regulations for notification and survey of hazardous installations which.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 33 recommended in their First Report the introduction of a notification and hazard survey scheme for installations where specified quantities of certain dangerous substances were involved. when the CIMAH planning cycle started. These plans included chlorine storage and processing areas. commonly referred to as the CIMAH Regulations. The Control of Industrial Major Accident Hazards Regulations 1984. and in the vicinity of. More sites are still being identified and it is expected that by the end of 1988 some 10–12 sites in London will be embraced by the CIMAH regime. This directive was translated into UK Statutory Instrument No. now has two functional responsibilities.

— told of the significance of the audio warning. Some 2500 people died. Moreover. but nobody knew what the signal meant or what action to take upon hearing it. It is not proposed to address such matters now. neither is it appropriate to discuss the medical treatment of the affected population. of which the last is most important. On 3 December 1984 a lethal cloud containing some 15 tonnes of methyl isocyanide (MIC) covered some 30 square miles of the Indian town of Bhopal. but one relevant question remains: Why was there no off-site emergency plan to provide for such an eventuality? Had there been one. . In Bhopal a public warning siren was actually sounded. These three measures. the public would have been given prior advice to do just that.34 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS EMERGENCY PLANS Why do we need emergency plans? What does a professional emergency planner aim to achieve by making a plan in the first instance? What would happen if an emergency of one kind or another did arise and there simply was not a plan to achieve appropriate and coordinated responses to it? These questions can perhaps best be answered by reference to the recent and well publicised Bhopal disaster.. according to one reliable source. it should have addressed the various measures outlined below. the siren was sounded only briefly lest the public became alarmed! The measures that a professional emergency planner would have included in his plan would have been to ensure than the people within the vicinity of the site were — made aware of the existence and nature of the hazard. albeit late. Much evidence has been gathered about this accident—even if some aspects of it are still shrouded in mystery—and of course the human interest element has been widely covered by the world media. notably the World Conference on Chemical Accidents held in Rome in July 1987. Because it is generally accepted that the best protection in a toxic environment is to be indoors behind shut windows etc. but doubt remains on some aspects. The technical details of the disaster have been extensively ventilated in the scientific press and at seminars. would have constituted the basic preliminary elements of the Bhopal plan. — told what to do immediately on hearing the audio warning.

the hazard analysis should state that the worst credible event would . Flixborough. if a relevant hazard analysis of the Union Carbide operation at Bhopal had been undertaken. thereby reducing the risks to such a level as to make the execution of an off-site emergency plan only a remote possibility. the first step is to obtain from the user of the hazardous substance(s) a hazard analysis that has been endorsed by an independent and competent agency such as the Health and Safety Executive. hopefully.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 35 It is common knowledge that some parts of Bhopal comprised temporary and frail domestic shelters and that there were even many pavement sleepers. If Bhopal and. For example. It should not be just a statistical risk appraisal of probabilities. and record a suitable synopsis in layman’s language early in the planning document. Seveso. before that. but there were also some buildings which could in the event have been used for protection. effective contingency plan-making would have significantly minimised the damaging consequences for life and property. in the case of a toxic substance. perhaps including the contingency provision of temporary shelter. Plant procedures and safety measures would have been examined in depth and. but it should clearly and consistently reflect the scenario or scenarios as to what could hypothetically constitute the worst credible event(s). or the determination of mean time between component failures. Indeed. and others sadly becoming too numerous to mention. the emergency scenario would almost certainly have been foreseen and appropriate steps taken. had been embraced by the Seveso Directive regime. EFFECTIVE PLAN MAKING How then do the professional emergency planners involved in CIMAH plan-making in the United Kingdom go about making effective and meaningful plans with a view to achieving coordinated and cohesive responses by all the agencies which have contributions to make in given emergency situations? Step 1: Hazard analysis As already indicated. action taken. The existence of a hazard analysis would have had a fundamental effect. even if they had. and not all the people slept in the open. or the assessment of how many times per million years this or that might happen. the disasters now taking these names into the history books might not have occurred.

y min. the emergency planner moves to the second step. The plan maker will simply provide for the interactive coordination of all the agencies involved.36 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (a) cause a release at x kg/min for a duration of. Every one of the planned activities must contribute to the attainment of the aim. z tonnes. The emergency planner should not. and (b) involve a potential total quantity of toxic gas released of. of course. The aim must therefore be realistic and. Step 2: Aim With the hazard analysis in his possession. population density etc. and. The aid must be to minimise the worst potential effects of the emergency. the planner should first outline the resources needed to attain the aim. When this information is available it is possible to determine the area which could potentially become affected and to adjust the data in relation to weather. namely the ‘execution’ or implemention. attainable. it should answer the questions Who?. In the case of a massive release of a lethal substance. Specifically. however. concern himself with how it is to be done—this is strictly a matter for the professional disciplines concerned. In this section. including the important matter of the arrangements for command and control. Here he determines the ‘aim’ of the plan. and follow with a description of the resources actually available from contributions by the various agencies. The implicit assumption in the plan is that the ‘worst credible accident’ will happen. above all. to devise a method of applying these variables to a real world situation. topography. What? and When? This should be followed by an outline of the operational organisation needed to apply the available resources effectively to the hazard. . Step 3: Execution This third step is the very heart of the plan. say. The aim of the plan should be clearly and concisely stated before the ‘execution’ section is considered. it would be a mistake to convey the notion that all the lives in the affected area must be saved—if such an idealised objective cannot be fulfilled. say.

Emergency plans are not born fully formed like Venus rising from the sea but are subject to a process of continuous evolution. It is absolutely essential in crisis management— because it is what a major emergency requires—that at least the principal officers concerned should be able freely to communicate with each other and be supported by adequate equipment and material. Such exercises undoubtedly generate lessons for all to learn—not least the plan maker himself.ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 37 What the plan should also usefully provide are convenient checklists for the participants covering those generally relevant matters outside the specialist disciplines. This is usually done by running an exercise which has the further advantage of serving to train the contributors in their respective roles. where to go. All plans need to be maintained and updated on a regular basis. what equipment will be needed. Step 5: Plan validation/training/exercising/ maintenance Finally. once the plan has been made it must be validated or tested. progressing from a broad rudimentary start to greater and greater detail and coverage but in which perfection is never likely to be achieved. or what sort of protective clothing will be necessary in the face of the particular hazard. For example. Step 4: Communications and logistics Even the best plans will fail to achieve their aim(s) unless the contributors have adequate communications and appropriate logistics back-up. in particular the police. The emergency planner It follows from the obvious importance of the emergency plan that the emergency planner must be a professional trained to analyse the problem and to identify the resources and actions required as well as those available in practice to solve or counter it. He must then devise an organisation to apply the resources and actions to the solution of the problem in the most effective and efficient (in . Such information would also be of value to other contributors. the checklist of the firefighter responding to a given emergency would advise him which route to follow.

The total number of officers directly employed on emergency planning in a full-time capacity is not readily available. join either the County Emergency Planning Officers’ Society or the Association of Civil Defence and Emergency Planning Officers. the required talents. for it is a sad reflection that it is presently only at such establishments that the skills of planning are thoroughly taught to the necessary extent. been successful in obtaining people with many. While membership of the Society is restricted to . In terms of quality of work experience. Figure 2 shows the differing backgrounds of formal higher education and. if not all. the Authority’s retiring age is 65). and with team leaders who are either former scientists or people with considerable backgrounds of relevant experience. The LFCDA Emergency Planning Division is organised into teams of 4 with as great a mixture of talents and experience as possible. Age structure. however. for those exuniformed service.38 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. on the whole. Figure 1 shows the breakdown by age and reflects the general high level of experience (incidentally. London has. 1. Most officers. Figure 3 shows the differing work experience and the considerable number of officers with experience of more than one of the relevant categories. just over 50% of the officers are members of the Institute of their previous professions while all but some 13% of the ex-uniformed members of staff held senior rank. The illustrations show the mixture of talents and qualifications brought together in the London Fire and Civil Defence Authority (LFCDA). formal staff college training. that order) manner. a very high proportion of the staff is engaged upon second careers.

the heads of the emergency planning teams formed in local authorities. 3. of course. Formal academic qualifications. vice versa. Twenty-four per cent of the staffhave had the benefit of post-graduate Staff College Training. many members of the Society are members of the Association but not. membership of the Association is open to all officers employed full-time on emergency planning. FIG. Indeed. Thirty-three per cent of the staff have had previous experience in the field of Emergency Planning. Nature of previous work experience. Current membership of the Society is 58 and of the Association it is 392. .ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING 39 FIG. 2.

among other things. the award of a diploma. It is to be hoped that the Commission will not be slow to take the initiative in that regard. in other words. it would reduce the considerable dependence on second careerists.’ Clearly. may one inquire. The principal thrusts of the function lie in the areas of civil defence and industrial major accident hazards. the creation of the organisation referred to is the function of the emergency planning profession. There is a growing need for a centre of learning to address the subject of emergency planning. have a structure and chain of command like a military division on the battlefield. though membership of the Society of Industrial Emergency Services is believed to be on the increase. A formal course for aspiring emergency planning officers is needed leading to. Where. there are smaller but equally important areas of activity such as oil pollution and nuclear radiation.40 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS CONCLUSIONS Emergency planning in the United Kingdom is an old activity which has only relatively recently developed into a full profession. In the same article there was a reference to the new European Disaster Medicine Centre at San Marino ‘…so that at European level there should be a single corpus of theory on the organisation of medical help…’. One sentence in the article reads as follows: ‘If a rescue operation is to be efficient it must be properly organised. is the single corpus of theory on the organisation of multi-discipline responses to major emergency and disaster situations being formulated and taught at the European level. exchanges of persons responsible for civil protection as part of training programmes undertaken by the Member States…’ is a most encouraging contribution to the development of the profession. Probably the largest number of professional emergency planning officers is to be found in local government employment. A properly developed career structure would also be beneficial to both the producers and receivers of emergency plans. in cooperation with the Commission. However. the article ‘Major catastrophes: our vulnerability’ in the Council’s May 1987 edition of Forum was of particular interest. at least. In that connection. On the wider issue of improving the quality of our emergency planner training. or even at the national level? . this initiative by the medical profession is to be applauded. the Resolution of the European Council in June to ‘encourage.

called the ORSEC plan. The EEC Seveso Directive makes emergency plans mandatory for member countries. This plan has been elucidated by the following appendices: ORSECRAD for nuclear activity ORSECTOX for chemicals ORSECHYDROCARBURES for petroleum products In 1985 a new interministerial instruction entitled ‘Technological Risks’ stipulates plans of action for chemical and petroleum installations. The 1976 Law of Classified Installations replaces the former 1919 Law of Classified Installations. the mobilization of means. while stipulating a certain number of principles. It has two parts: . and starting with certain amounts of products. aims at ensuring the command.GROLLIER BARON Institut Français du Pétrole. and stipulates the responsibilities of the head of the establishment and those of the Public Authorities. The 1952 interministerial instruction. requires a defence plan to be drawn up under the authority of the Prefect to define the course of action to be taken in case of accident. The 1967 petroleum regulation. Vernaison.4 Emergency Plans in France R. the inventory of these means and their distribution. It applies to activities covered by the Seveso Directive. resulting from the accident at Feyzin. France French regulations have long provided for measures to be taken in the case of an accident in industrial establishments. For some dangerous activities. but the Prefect can extend its application to other establishments. manufacturers must obtain an operating authorization including the description of what is to be done in case of an accident. and there are labour laws concerning personnel training and control facilities that such establish-ments must have. which was improved in 1971.

in particular: — Alarm — Firefighting — Safety perimeter — Access routes — Instructions to the public — Assistance and relief for the wounded — Pollution control — Relations with the media — etc. POIs and PPIs reinforce one another. worked out under the authority of the Prefect. and it defines the public alarm to be set off. Likewise the implementation of public means also requires good coordination of means in keeping with the definition of missions. thus requiring solid coordination at the level of how they are worked out. The general manager of assistance and relief is the Prefect or his delegate. In addition. We are interested in their consequences and not in preventing them.42 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — The in-house operating plan (POI). . for an establishment or industrial complex when the accident has or-may have consequences going beyond the boundaries of the establishment. concerning what is to be done inside the establishment when an accident occurs. A PPI has two parts: (1) the organization. worked out under the responsibility of the head of the establishment. to searching for supplementary means. It is based on scenarios leading to the assessment of the needs. it deals with the POI/PPI interface during the time the accident is increasing in gravity. The scenarios are constructed from reference accidents. (2) the means. property and the environment. A leader and means correspond to each mission. Exercises must serve to check proper procedures. to the identification of the means available. — The special action plan (PPI). Command is ensured by an operational command post and a logistical command post. Its aim is to enable the Public Authorities to ensure the protection of persons. Among the missions included in the PPI are. Mr Boissieras will discuss POIs in Chapter 5.

Portugal Rapporteur.S.SESSION II On-Site and Off-Site Emergency Planning Design Chairman: J.NICOLAU National Service for Civil Protection. R.A. France .GROLLIER BARON Institute Français du Pétrole.

and of actions to be taken when there is an accident situation. previously inventoried. . and it contains all the information to be given to Public Authorities to carry out the Special Intervention Plan (SIP) prepared under the Authority of the Commissaire of the Republic: — Starting from a study of the potential dangers presented by the installation. For the Establishments alluded to in the Order ORSEC Plan: Technological Risks. — The SIP aims to ensure the safety of the public and to protect the environment when the accident entails or may entail danger outside the Establishment’s boundaries. and the necessary means which the operator must put into practice in the case of an accident. the IPO defines the organisational measures. to protect personnel. the Emergency Plan corresponds to the requirement for the establishment of an Internal Plan of Operation (IPO). the public and the environment. Safety Directorate.1 Definition The Emergency Plan is the guide on the setting up of internal and external equipment at the Establishment.5 Guide for the Establishment of an Emergency Plan J. Lyon.BOISSIERAS Rhône Poulenc. the methods of intervention. France 1 GENERAL COMMENTS 1. for which the Head of the Establishment is responsible.

Although it is intended to allow mastery of serious situations (fires.3 Field of application The Emergency Plan concerns any situation involving the putting into practice of safety actions or of protection of people. the decree of 2 Sept. 10) and an internal operation plan. 3–5). To avoid improvisations. — The ORSEC Plan: Technological Risks. Art. in France. For installations covered by the Petroleum Regulations and their extensions. which has been applicable to the Member States since 8 Jan. including material for the rescue and evacuation of personnel (R23338/39/40/41). 10): — Texts of application proceeding from the CEC directive 82/501 of 24 June 1982 (Seveso) concerning the major risks of some industrial activities. fulfilling the needs of the Seveso directive. pollution). The Emergency Plan also has a legal foundation based on: — the Working Code through the obligation to give safety training (R231–34/35/36/37) and to put into practice fire-fighting equipment. 1967. it appeared necessary not to exclude from its field of application accidents or incidents of . an interministry order of 12 July 1985 on intervention plans in the case of accidents which. it is a good idea to predict and to plan. the ORSEC Hydrocarbons and ORSECTOX plans. Decrees of classification may lay down specific protection measures. In the case of an accident it is too late to develop a strategy of actions taking account of all the possible consequences. 1977 applying law 76. — the Regulations of the Classified Installations. 1972 on hydrocarbon deposits.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 45 1. replaced.2 Why an emergency plan? The Emergency Plan forms part of a policy of prevention and protection of people and goods. Art. of business and its environment. toxic emission. a defence plan must be organised (decree of 4 Sept. plus an overall defence plan (decree of 9 Nov. 1984.663 of 19 July 1976 in particular envisages the organisation of safety means (Art. goods and the environment. in agreement with the general policy of Public Powers. 1.

and generally under the coordination of the Safety Head. one can see the essential role of permanent monitoring (caretakers. etc.46 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS medium seriousness as far as they need non-routine actions of help or information. particularly the classes of compulsion. — the organisation. structures. Account should be taken of all the situations linked to the presence or absence of personnel in the Establishment (working hours. — the environment. importance and psychological impact of the accident or incident. Outside the Establishment the manual will be sent to the people responsible for Civil Safety.) and of its accessibility. .). from this. compulsion. social conflicts. — means of intervention. The corresponding procedures will be distinguished as a function of the nature. and be able to deal with the specific risks of each situation.4 Manual The Emergency Plan is formalised by the drawing up of documents. To facilitate checking and updating. In the Company there will be a control copy as evidence of updatings. which contains all the information necessary to manage an accident situation. which are collected together in an ‘Emergency Plan’ manual. This manual is drawn up by the Establishment Head. These can be made when intervention exercises are being performed. periods of leave. The Establishment staff who are asked to organise or take decisions in the case of an accident must know and have mastered the contents of these manuals. all the copies will be numbered and named with the function. in liaison with the Public Powers. remote observation. It must be operational whatever the situation. 1. who is responsible for updating it as a function of the evolution of — risks or knowledge. etc.

(2) For each possible and probable scenario. human and material. (4) Organise the whole in an ‘Emergency Plan’ manual. 3 EMERGENCY PLAN MANUAL This may act as a specimen scheme for the internal operation plan.1 Alert — Description in the form of an organigram of the progress of the alert from the first sign through to the services concerned — List of the telephone numbers and addresses as an appendix . to be put into operation to limit the consequences. with — nature and location of the risks.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 47 2 STAGES FOR THE REALISATION OF AN EMERGENCY PLAN (1) Examine the accident situations which merit actions to be taken into account in the Establishment Emergency Plan. (3) Draft the procedures of — intervention. — gravity of the consequences. as well as the advice for carrying them out. This evaluation results from the ‘danger study’. — information. define the internal and/or external means of intervention. 3. Take the advice of those concerned. which is obligatory for all classified installations subject to authorisation. — returning to normal after the accident. which must be in the form of single sheets which can be used by anyone involved.

zone or workshop. — the zones to be protected in an emergency.5 Organigram of the services Organigram of the assistance services. where it comes from. 3. fighting against the accident .2 Geographical situation — Plan of the location positioning the factory in relation to its environment on the scale of the major risk and showing the access and assistance routes — General plan of the Establishment with the reception sites — Meteorological data such as the wind rose of the site 3. with the names of the people in charge and the staff needed to ensure the following five tasks: — Operation: stopping units and making them safe. — List of the external public and private defence equipment. with its potential. and the time needed to make them operational (after the call) 3.3 Risk For each unit. the plans of the inside of the establishment show in particular — the possible ways of access. where they come from. — the zones which might be affected by a toxic cloud or by a shock wave caused by an explosive cloud.4 Means of intervention — List of the Establishment’s fire-fighting equipment with its potential.48 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3. and the time needed to make it operational (after the call) — Private or public water services — List of the various materials or products with their potential.

media. the media and the public — Operations to be carried out. fuel.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 49 — Communication: making and keeping available means of communication — Logistics: supplies of material.6 Operations For each unit. and each intervention sheet will include the following elements: — Risk definition: — Fire/Explosion — Escape of toxic gas or liquid — Escape of inflammable gas or liquid — Aims of the fight against the accident — Location of the PC and the organisation of the assistance — Evacuation and counting of personnel — Staff necessary and their role — Means necessary and where to find them — Information on the administration. men. etc. functional services. . Appendices — Inventory of dangerous products and files produced — Plans of the installations Remark This schema may be changed depending on the characteristics of the Establishment itself.… — External relations: administration. area or workshop. public. — Observation: preparation of the log-book. the strategy will be studied in relation to the potential danger classes. files. … 3.

Staff should be trained to give a brief and precise warning message indicating the place. Permanent readiness is ensured by the existence of a ‘fixed point’ where the tasks and means defined in the organisation lead at least to the transmission of the alert.50 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 4 RECOMMENDATIONS 4. guard post. The means of intervention corresponding to the type of accident are triggered by alerting — the Establishment’s permanent or auxiliary fire service. or outside firemen (possibly put on pre-alert). or who can come in quickly when called). It should therefore be possible in any circumstances to summon — Personnel who can take decisions (personnel present. Generally the message is received by another person at the Fixed Point (assistance centre. . — medical service and/or first-aiders.1 Emergency plan always operational The modes of work of Establishments vary. Safety functions must always be envisaged in the organisation. The list of people who must fulfil these tasks in the case of an accident must be available and the people must be informed. in principle using alarms which are inside or outside the Establishment. The personnel outside must be easily contactable and be supplied with passes in cases where traffic restrictions are set up.2 Alert It is the duty of any witness of the beginning of an accident or of an anomaly which might lead to an accident to give the alert and to act with the means at his disposal and within the limits of his ability (1 st intervention step). — Material and human means of intervention. but an accident can occur at any time and often in the least favourable situations. or remote sensing centre). telephone switchboard. The alert is the information given to ask for assistance. according to regulations. 4. type and seriousness of the accident.

— with what means (location and performance of the materials). The Working Code obliges the Establishment Head to ensure — a minimum of protection against fire (extinguishers.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 51 — people with a specified task (e. — how to coordinate the performance of the intervention. reinforcing guards). including staff training. — the training of first-aiders in all the continuously established teams. The most pessimistic scenarios which are least likely to occur may be taken into account. the assistance systems which must be ensured are both internal (those of the Establishment) and external (those of the Public Powers or Mutual Assistance).g. In general the quarter-Heads will be first-aiders. . The coordination of actions implies delays which must be taken into account in evaluating the efficiency of the assistance. In the Establishments subject to the Classified Installation rules. depending on whether it concerns — a zone of the Establishment. Section 4. etc. Practically. instructions. how he can be informed. — outside the Establishment (cf. the preparation of procedures and information for the mastery of each risk must allow definition of — who acts (name.3 Means of intervention The means will be adapted to the nature and importance of the risks. 4. — the entire Establishment. In all cases.…).…). address.10). … There will be differentiated alert levels (internal and/or external) depending on the type of accident and on its potential consequences. The alert level is decided by the most senior person present at the time. the means of intervention are defined in relation to risks examined by the study of dangers and by the scenarios considered possible for accidents.

directed by the person responsible for the intervention — Then. The person responsible for Exterior Assistance receives a task from the Director of Assistance. stated in the SOP. near to the accident. in a room where liaisons with the advanced CP on the one hand and the outside world on the other will be organised In an Establishment which has a Security organisation and means of intervention. this task is the responsibility of the most responsible person on the site — Put into action the means of intervention (strategical choices=controlling the accident.4 Triggering of the emergency plan As soon as the alert is received. the Assistance Directorate comes under (as part of the IPO) the Establishment Chief or his representative.…). this task is the responsibility of the intervention head (2nd rank and others) These two tasks can only be accomplished with perfect coordination between the operation hierarchy and the different intervention levels. . and without waiting for the starting of the SOP. zones to be protected. He deals with its performance and puts the necessary means into effect. in the case of an accident extending outside an Establishment. two tasks must be ensured: — Define the actions for treating the accident (tactical choice=stop the unit. the Establishment Head must act outside his Establishment under the responsibility of the public authority and in the framework of previous and clear agreements with this authority.…).52 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 4. If the accident cannot be mastered in the framework of the workshop. if necessary. the following must be set up: — An advanced operational CP. put into a safe state. a Central CP housing the Assistance Directorate. evacuation. The ORSEC Plan: Technological Risks envisages that. help for people.

— visitors. If necessary these manoeuvres will be performed with a small number of personnel and/or means of protection such as individual masks. Emotional reactions might lead some people to flee. the personnel must go to a prescribed assembly point and be counted. administration. messenger. . Lines should be reserved for outgoing messages (be careful of the risk of blockage of lines which go through a guard-post).6 Means of communication Inside the Establishment the people in charge will generally have radios for communication. To do this. loudspeaker. the workshop must set up a state of emergency.… Communication with the outside will be mainly by telephone. in particular of products. Taking account of the risk of breakdowns. without forgetting — part-time staff (maintenance.…). — outside organisations. as well as the telephone. The orders to be given to the staff should be given by means of telephone. As far as possible the staff will be counted through the hierarchy. It is advised that radio posts should be made available for outside assistance (firemen). including home address and telephone number.5 State of emergency: internal evacuation. When the order for evacuation of a zone is given. while admitting the possibility of some leaks. counting If the nature of the accident implies the evacuation of a zone. Telephone lines must thus be available to organise assistance. radio. Part of in-service training is for personnel to acquire the right reflexes.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 53 4. megaphone. special instructions will be established to make the workshops safe with a small number of manoeuvres and without any risk of making the accident worse. An up-to-date staff list. 4. there should be two independent systems for communication with the Main Assistance Centre. must always be available.

various categories of people might come to the plant. Information for families should be organised quickly. it is a good idea to set up quickly a technical analysis which allows the drawing up of an accident report. The information must be given by the Establishment Director or a person designated by him and trained for this job. The aim of the information is to inform people of the nature and the consequences of the accident.7 Reception Following the accident. if not. Curious onlookers must be kept outside and they will be asked to keep away from dangerous zones while waiting for barriers which might be set up by the public authorities. They should then be confirmed in writing. Do not formulate any assumption on causes or responsibilities. The information will be the subject of a precise plan incorporated in the emergency plan and structured depending on the seriousness of the accident. especially if there are injured people or people who are being kept on the site. and limited to the facts. 4. 4. towards the site of the accident (operational CP). . A scheme is proposed in the Appendix.9 Analysis Independent of any legal enquiry.54 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 4. — the media towards a specially set up room. These guards will direct — the assistance towards a pre-established assembly point or. — the authorities towards the Directorate’s representative. It must be quick. or the usual person who receives visitors (special room). objective. Spontaneous eye-witness accounts should be collected quickly. so a reinforced guard is needed at the Establishment entry. if possible on tape. accompany journalists who are authorised to visit accident sites.8 Information procedures These should not be confused with alert procedures or with the orders given as part of the intervention procedures.

10 General information of the public When the consequences of an accident may overflow the limits of the Establishment and affect neighbouring populations (e. signs perceptible from outside. The public must be informed of this alert procedure in advance by the distribution of a card giving details of the security measures and the procedures to follow. toxic emission). immediately inform — the Company hierarchy following special instructions. if they have not been informed as part of the alert. There should be at least an instantly developing camera in the first-aid vehicle. . risks that the accident might extend outside.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 55 To do this.…). an alert procedure must be set up in the Special Intervention Plan. On 22 July 1986 the UIC published a technical circular recommending the conduct for this subject. then. If there is damage which temporarily puts the installation out of action. even those which a priori appear harmless. APPENDIX: EXAMPLE OF AN INFORMATION PLAN For any incident or accident. and emission or danger of emission of pollutants. — the police. Then one will inform — the factory inspectorate. one may use audiovisual apparatus (camera. — the advising engineer of the CRAM. — the local authorities. in the most serious cases. as quickly as possible. video camera. as quickly as possible. and.g. — the inspector of classified installations and. and in any case as soon as the alert is given. the factory Director or his representative will inform. film camera. 4. as a courtesy. the Commissaire of the Republic.

to inform the public.56 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — neighbouring factories. If personnel are kept on the site or if there are injuries. one will warn — the family of the personnel concerned. a communique will be drawn up in agreement with the Public Authorities. Finally. for serious cases. .

CORIGLIANO & F. we report here the case of the larger MONTEDIPE factory (i. Milan. Of course.6 Emergency Plan and Alert System at MONTEDIPE L. referred to a limited event whose possible consequences are confined to the plant where the accident takes place. The zonal emergency. The several situations of MONTEDIPE factories (MONTEDIPE is the petrochemical and polymers company of the MONTEDISON group) have made necessary different approaches to achieve this goal. 2. the development of process plants demands a continuous updating of the emergency rules and procedures. the number and type of plants. This concern is clearly expressed in the legislation and particularly in the EEC Directive on the major accident hazards of certain industrial activities which also requires linking of on-site preparedness emergency planning with off-site Territorial Civil Protection Plans. when the area affected by expected consequences can include several plants. considering the latest available studies and technologies. considering the extent of factories. when the area that can be affected involves all or most of the factory. the organisation and constitution of personnel. Italy Provision of appropriate information about the actions to carry out and the correct behaviour in case of accident has always been a primary requirement in the chemical industry. The general emergency. checking their feasibility and adaptability to different possible events. The three aforesaid levels are covered by the specific Department Plan and the General Plan. As an example. which is linked to the Territorial Civil Protection Plan for chemical Hazard. The local emergency.e. Obviously the three steps are inter-connected in such a way as to allow the gradual transition .ANTONELLO MONTEDIPE. P. etc. for which three levels of emergency have been defined: 1.to Marghera). 3.

Assessments of occurrence likelihood are made.58 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS from one level to the higher one in the case of anomalous evolution of the situation. other plants involved in the situation. . The method of notifying employees of the actual situation is the first concern in the case of accident. The Department plan is developed on the basis of a range of reasonable hypotheses of specific events. public authority. and on the basis of a predetermined check. — a map of the plant on which is plotted the route to follow in order to perform the task. In this case. As a general rule. — plan of the factory. Panic caused by uncontrolled rumours and feelings or lack of knowledge can also create an emergency. — a list of Departments and Managers to give notice about the event and its possible evolution. in relation to the development of the situation. — the most important technical files for all the different plants and utilities. and the emergency planning coordinator. an Emergency Committee will meet in a centre equipped with — personal protection equipment. from the oper-ations centre to the control rooms of the plants and other specific workshops. the Department plan (first level) includes — the indication of charged employees who must accomplish the emergency operations and the relevant tasks. This plan provides in particular for quickly warning the in-house fire brigade. . one may extend the emergency to the second level. — the ultimate meeting point from which to abandon the plant in order to evacuate. Afterwards. For this reason the factory has a departmental selective communication warning system. — direct phone lines to Company headquarters. Such a procedure allows shutdown and evacuation of the plants or the areas affected by the emergency without spreading alarm and without involving needless general emotional upset of other departments. and possible consequences are evaluated providing adequate operations and measures both to reduce the likelihood and to mitigate the consequence. particularly in the second and third levels of the emergency plan. emergency services.

and to improve expertise. The trainers may choose from among several typical hypothetical events and test. self-contained breathing apparatus. Moreover. gas . in order to train the operators. It is therefore possible to check the reliability and functionality of the warning system. From the keyboard it is possible to check the availability of the system at any time. the preparedness of the operators in the receiving stations. Periodic simulations of different types of emergency are performed. The operator may find — typical emergency cases. — the appropriate procedure to adopt. simulating the alarm. be heard in all the departments in the case of a widespread emergency. consisting of software for a microcomputer.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 59 The warning system. can represent on the display the map of the factory or the part of it concerned with the event. In fact. or of other companies in the event of accidents during transportation or in other sites. A set of emergency protection equipment is available to deal with different situations. which have specialist teams working on shift. This affords the possibility to supply information about the accident and to issue the appropriate guidance how to operate. for some years now the emergency teams of some works have proved themselves able to act on the request of the authorities. The warning system is connected to meteorological stations in order to collect data on wind direction and velocity. illustrating the likely affected area according to the chosen type of accident. allowing the spread of preparedness recorded messages that are appropriate to the actual situation. These areas have been estimated by computerised mathematical models simulating the consequences of chemical accidents in the typical range of reasonable hypothesis. — the area that can be typically affected. e. the warning system is connected also with the network of alarm points. to test the correctness of the procedures. The fire-fighting and the first-aid within the works are entrusted to the fire brigade and the medical department.g. A continuous alarm signal (sirens) will however. it has proved to be of use also for emergency measures taken outside the works itself. in addition to the availability of appropriate equipment. This type of organisation has encouraged the development of a suitable professional attitude to the problem.

face shields. — kits for repairing leaks. . A public authority. working together with the appropriate authorities to prevent harmful consequences and to mitigate the effects of accidents occurring even hundreds of kilometres away from the works itself. etc.60 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS masks with universal and specific filters. MONTEDIPE runs. chemical safety goggles. on behalf of Federchimica. at any time. such as the fire brigade. e. there will be an expert who will answer questions after a quick scan of the database. a database of more than a thousand substances to comply with any emergency situation. — neutralising products which inhibit the evaporation of toxic or harmful substances. — hydraulic pumps ensuring safe operation even where flammable mixtures are present.g. — a flare that can be assembled on the site of the accident to burn any flammable gas. named SIET. Moreover. acid-proof and gas-proof clothing. can call this database. For some years now it has therefore been possible to respond efficiently to all the requests for help received. Moreover. the teams have special equipment at their disposal.

use of non-flammable rather than flammable solvents) 2. strict adherence to design codes. requires inter alia that a site on which is stored more than specified quantities of certain hazardous materials shall have an on-site emergency plan. its consequences are minimised. Billingham. UK 1 INTRODUCTION The Seveso Directive. pressure vessel design. provision of bunds. constructed. 2 and 3. Overall the protection of people both on-site and off-site who might be affected is best achieved by using procedures designed to ensure that the risk of a serious incident is low and that.7 On-Site Emergency Plans G. etc. The use of multi-stage hazard studies. Hazard reduction (e. which was enacted in the UK as the Control of Industrial Major Accident Hazard Regulations 1984 [1].L. These procedures will normally be directed towards: 1. use of smaller inventories in process and storage) 3. if it does occur. Incident mitigation if all else fails To achieve items 1.g. In most cases these plans had been in place for many years. Hazard elimination (e. because manufacturers had recognised their value in mitigating the effects of serious incidents. plants processing hazardous materials need to be designed. Equally important is the selection and training of operators. Hazard containment (e. proper maintenance and regular tests of key plant items and safety devices all contribute to ensuring that plant hardware is reliable.ESSERY Imperial Chemical Industries.g.g. the provision of operating . operated and maintained to high standards.) 4.

In the case of a minor incident which affects a single plant area only. Where the effects of a major accident could extend beyond a site boundary. Guidance on off-site planning is available from the UK Chemical Industries Association [4]. their maintenance with audit and review and good supervision. It is not possible to provide a detailed description of an on-site plan in this short paper. this aim can generally be met fairly easily. and more general advice is available on emergency planning for Local Authorities [5] and for industry [6]. or more likely an unforeseen series of events. However. What follows is therefore an outline of some aspects that need to be considered in preparing an on-site plan. the procedures and training must prompt and condition people to act correctly. it is essential that the on-site plan should integrate effectively with those of the external Emergency Services. a sound emergency plan is essential. this may not be true of a major incident which affects a much wider area. which could be used to give a rapid consequence assessment of a toxic gas release. As many of the people who will be . Unlike a normal plan. which starts from a single clearly defined point. In this case the people trying to contain the incident at source are often unaware that the effects of the incident are causing problems further away and the people trying to ensure that the correct actions are taken elsewhere often have great difficulty in obtaining enough information to enable them to give sound advice and instruction. accurate information is essential.62 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS instructions and safe working procedures. even with high quality plant and adherence to carefully considered procedures. could lead to a serious incident. Such integration can be achieved only by close liaison with the planners and the operating teams from these services. Guidance on emergency planning has been published by the UK Health & Safety Executive [2]. Should this occur. an emergency plan has to be able to start from any one of many abnormal situations. As rapid. it is still possible that an improbable event. However. and another paper [3] includes a simple quantified approach. 2 EMERGENCY PLANNING: AIM AND GENERAL PRINCIPLES The prime aim of an emergency plan should be to restore normality as far and as quickly as possible with minimal adverse effect on people and the surroundings.

It needs to be simple so that those called upon to implement it can do so readily. in the initial stages of an incident. flexible approach will ensure an easy dove-tailing of the on-site plan with the off-site plan prepared by the Local Authority in conjunction with the Police. it is advisable to establish both an Incident Control Point and an Emergency Control Centre. It will also be necessary to warn people who might be affected by the incident. The next stage will be to consider the appropriate countermeasures at the scene and elsewhere. it is essential that the emergency plan should be simple and flexible. Simplicity will also reduce training requirements and increase the likelihood that all those covered by the plan will react in the desired manner in the event of an emergency. it may sometimes be necessary to consider the normally absent products of a runaway reaction such as dioxin at Seveso or the toxic products of combustion following a major fire. He will be responsible for ensuring that the actions outlined above are taken and. The Site Main Controller will operate from the . and it needs to be flexible to allow for easy escalation or de-escalation to match changing situations. Where containment has been lost. Possible contamination of the environment may also need to be considered. with whom there should be consultation during the planning stage and following regular exercises thereafter. Furthermore the simple. Apart from possible fire. The Incident Controller (who needs to be readily recognised by the Emergency Services) is likely to be the shift supervisor until he is relieved by a more senior manager with knowledge of the area in which the incident is occurring. There will have to be a rapid means of calling in both the Emergency Services and the Company personnel required to respond to the incident.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 63 called upon to perform key roles in the major incident response team will not normally be involved in dealing with emergencies. but it should be recognised that this ‘warning’ might come after an explosion or a gas cloud has affected the people concerned. he may also have to act as the Site Main Controller. it is necessary to postulate the various types of major incident which could arise and the likelihood and severity of these hypothetical events. despite the wide range of start-points. it should be reestablished as quickly as possible. explosion and toxic release scenarios. In dealing with a major incident. Personnel must be accounted for or searched for and where necessary casualty treatment must be initiated. Before starting to draw up the plan. Fire and Ambulance Services.

data preservation. the team should consider how the superimposed effect of a partial or total loss of essential services would influence their procedures for re-establishing containment. resource availability. The main objective of his support team will be the seeking out of this data to aid joint decision-making with the Emergency Services. traffic movements. To perform these tasks he will need information about the current situation.64 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Emergency Control Centre once this is established. advising the Emergency Services about the way the incident is likely to develop. appropriate counter-measures need to be considered. and arranging backup support for the Incident Controller. Similarly. In addition. 33 IN-PLANT INCIDENTS: SOME ASPECTS TO BE CONSIDERED As has been long appreciated by fire-fighters. Good communi-cations with the decision-makers in the Emergency Services is therefore essential. it is necessary for the local management team to assess the various types of incident which could occur in its area. likely weather changes. For example: Would remotely operated valves still be operable? Would the shutdown purge systems still be operable? Is the emergency lighting adequate? Do the communication systems have standby power supplies? Having identified the various undesired events and their potential consequences. a plant operator is often able to limit the scale of an incident provided that he has been made aware of the potential non-routine situations and given appropriate training and practice—real or hypothetical. a quick response by a well trained person can prevent a minor fire escalating into a major conflagration. For example: — Are the plant fire-fighting resources adequate for the initial attack prior to the arrival of the professionals? Is the equipment adequately maintained? Is the plant team adequately trained? Are there some situations where a fire should be allowed to burn or be contained rather than extinguished? . etc. He will be primarily concerned with co-ordinating the required actions across the site other than at the incident. shift changeovers. etc. In order for this training to be soundly based. The Site Main Controller should also ensure that adequate consider-ation is given to the provision of data for the Public Relations team.

and in addition there will now be a need to ensure than an Emergency Control Centre is established as quickly as possible. Some people like to believe that they can cope with an incident on their own. It will act as a data- . This reluctance to ask for help early enough can lead to an incident not being tackled in the best way and to a shortage of the information needed for advising the Emergency Services.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 65 — Can the drains cope with the fire-fighting water? (Inadequate drains can lead to the spread of hydrocarbon fires. This centre will act as the main communication link with the Emergency Services—particularly the Police. SOME ASPECTS TO BE CONSIDERED All the comments relating to in-plant incidents are still applicable. it is worth commenting on a very natural human reaction which sometimes leads to a less than optimum response. 4 MAJOR INCIDENTS. drivers and visitors? — Is there more than one method of communication? — How is additional help summoned? — How is the decision taken to declare a major site emergency which will require the call-in of the Emergency Contol Centre team? Before moving on to consider some aspects of dealing with a major incident. It is better to have far too much help at hand than to be just a little short.) Should the fire-fighting water be contained to avoid contamination of water courses? — Are instrument and electrical cable ways adequately protected against fire? — Has similar thought been given to structural steel and stock tanks? — Are relief valves and vents sized for fire conditions? — Could sub-zero liquids enter systems not designed for low tempera-tures. especially if escalation could occur unexpectedly. evacuation assembly and roll-call procedures adequate for the events postulated? What about contractors. thus causing failures which could exacerbate the incident? — Can all the necessary isolations be made easily? — Should more isolation valves be motorised so that they can be operated quickly from a safe location? Should some isolation valves be actuated automatically by gas sensors? — Are the plant alarm.

where appropriate. (On occasion it may be advisable to allocate a radio channel solely for these three users. preferably dedicated to and well equipped for the purpose.) Difficulties arising from interference or ill-disciplined radio procedure can also be overcome by using telephones in preference . A more senior Incident Controller may be needed. it may be necessary to have two centres allocated. 4. possibly using audible (and zoned?) alarms backed up with a looped tape to emergency telephones. They must be given the location and nature of the incident and. All these calls need to be made quickly when the Works resources are fully stretched dealing with the emergency. Good information flow is essential. The site needs to be told of the incident.66 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS gathering point for information gathered from many sources. many actions are required in quick succession: 1.) 7. Remember also that police road-blocks might restrict the access of those called into work. A mobile analytical team for monitoring the environment may be needed together with a readily usable means of estimating the extent of a toxic cloud. (Media interest even in minor incidents is much more rapid and demanding than it used to be. providing that it can be manned quickly and that it has excellent communication facilities. Can the calls be automated or cascaded? Remember that the travellers may need to be told of the direction from which the wind is blowing. 2. 3. to identify a safe approach route. Additional telephonists may be required to re-route calls to the Police Casualty Bureau or to the Public Relations team or to the Emergency Control Centre. Bearing in mind the difficulty of establishing what is happening in the first hour or so of a major incident. 5. it is essential that good communication links are established between the Emergency Control Centre. the mobile analysts. Having decided that a major incident is occurring. It is therefore advisable for non-uniformed personnel to carry some form of identification. It may be prudent to locate this centre some distance from the site. they should be told the direction from which the wind is blowing. particularly if problems arise from interference with other users on different channels. 6. The Emergency Services need to be called. the incident and. A Public Relations team will be needed. A Site Main Controller and his team may be needed. To take account of different wind directions.

some of which may be of poor quality. However. In the event of there being many casualties. and as shown in HSE papers [7.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 67 to TELRAD or radio. it may be beneficial to have a company personnel presence in the police casualty bureau. Another useful training aid is a video prepared for . and the police representative is able to obtain a better picture of the Works situation. Furthermore each individual should know what to do in the one or more posts which he may be called to occupy in an emergency. difficulties could arise with incidents in interface zones or in pipe trenches if areas of responsibility for dealing with major incidents are not adequately defined. In most cases the advice for the protection of the public from fire. (The absence of a telephone near the incident is easily overcome with a modern plug-in telephone and a long extension lead carried on a Works emergency vehicle. This should be resisted as it could lead to the giving of the wrong advice.) A police presence in the Emergency Control Centre is also strongly recommended. These cases must be recognised and appropriate arrangements made. 8]. the police may have information which will assist the control room staff. it is then important to go outdoors once the gas cloud has passed. It is also important to provide the police with information in a form which will help them. In assessing a situation based on a limited amount of information. While in-plant incidents are clearly owned by plant personnel. They need to know the effect of a particular gas concentration rather than its actual value. there is sometimes a temptation to make assumptions. everyone with an active role in the control of a major incident should be aware of the emergency procedures overall. explosion and toxic release will be to go indoors and shut all windows and doors. Practice in radio procedures may also be necessary. in the event of a toxic release. This representative should be able to work directly with his personnel colleagues in dealing with the next of kin without needing to involve anyone in the Emergency Control Centre. A training package [9] which gives guidance on emergency planning and includes seven in-depth case studies is now available. The use of checklists can be of great assistance when under pressure in an abnormal situation. 5 TRAINING Apart from the in-plant training needs identified earlier.

In many cases over the years there has been close co-operation on a voluntary basis between industry and the Emergency Services in the drawing up and testing of the various emergency procedures. These procedures should be simple. it has also been recognised that the potential for a major incident still exists and that appropriate pre-planning for such an incident should be done in order to minimise its consequences. the exercises help in building understanding and trust between the participating teams. and they act in accordance with the information received at the place required of them by the emergency procedures. The most essential training will be in conjunction with the Emergency Services. preferably held out of normal office hours. Table-top exercises in which most of the participants are present in one room are useful in that they cause minimal interference with plant operation and they can be conducted at a speed which allows the testing of the team’s reaction to more than a single train of events. It shows a major ‘table top’ exercise with the involvement of several external organisations. The process of identifying possible release mechanisms and rates should result in a full appreciation of the possible major incident situations. and in the interaction of these procedures. all necessary communication systems can be tested properly.68 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS CONOCO called ‘April Storm’. Equally importantly. The Seveso Directive and the associated UK CIMAH . planning to avoid a major incident has been the normal practice of industry for many years. In the essential joint debriefing sessions after such exercises. which in turn will allow the drawing up of procedures to handle such situations. It should take the form of a carefully designed and scripted exercise to test the emergency procedures which have previously been prepared following detailed discussions with the Emergency Services. With this arrangement. 6 FINAL COMMENTS For sound business reasons. there is also scope for a helpful interchange of ideas to overcome some of the problems which have been identified. In this case the participants are called out as they would be in a real incident. However. A somewhat more realistic testing of a site’s major emergency procedures is provided by a ‘Control post’ exercise. These joint exercises are of particular value in that they expose any weaknesses in the procedures to which the several teams of participants are working. flexible and widely understood.

Health & Safety Series Booklet. 3. Health & Safety Series Booklet. PURDY. 42. The Home Office. Refinement of Estimates of the Consequences of Heavy Toxic Vapour Release. 1986. 1985. 5. DAVIES. Guide to the Control of Industrial Major Accident Hazard Regulations. 1982. LYNSKEY. . European Federation of Chemical Engineering Publication Series No. P.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 69 Regulations now require that both on-site and off-site emergency plans are completed. & PURDY. Emergency Planning Guidance to Local Authorities. with benefits to all concerned.C. The Development of an Effective Emergency Procedure for a Toxic Hazard Site. 9. 2. European Federation of Chemical Engineering Publication Series No. Guidelines for Chemical Sites on Offsite Aspects of Emergency Procedures. G. 7. 8 January 1986. Control of Industrial Major Accident Hazard Regulations: Further Guidance on Emergency Planning. 1987.. 1986. Society of Industrial Emergency Services Officers. 4. IChemE Symp. Preventing Emergencies in the Process Industries.. 8. 6.C. HS (G)25. P. As these plans must operate together. Industrial Emergency Planning Manual. Toxic Gas Incidents—Some Important Considerations for Emergency Planning (HSE). G. Toxic gas risk assessments: The effects of being indoors. REFERENCES 1. 47. 1985. 1984. a video training module available from the Institution of Chemical Engineers. this requirement has brought industry’s emergency planners into even closer contact with the local Emergency Services.. & DAVIES. P. HS(R)21. Chemical Industries Association.

may be the source of a catastrophe risk.’ The local disaster prevention authorities and district disaster prevention authorities shall draw up and update a description of the hazards of all establishments which.’ . the operator of a plant shall. above all. FRG 1 GENERAL Instruction systems are a vitally essential prerequisite for the management and control of unusual malfunctions and catastrophes. These plans shall. units and establishments.8 Emergency Plans According to the Law for Protection against Catastrophes and On-Site Hazard Protection Plans According to the Major Hazard Regulations W. as well as other organizations. to be called upon for assistance in the event of a catastrophe. by way of their special nature. dated 27 June 1980. In the North Rhine-Westphalian Law for Protection against Catastrophes. specify the alarm procedure. the preparatory measures and all authorities.STEUER Bayer AG. Article 5 of the 12th Directive on the Implementation of the Federal Air Pollution Control Act (Major Hazard Regulations). Para. defines the requirements for limiting the effects of incidents: ‘In order to fulfil his obligation arising from Art. 3. in particular. the legislators specified the following in Article 18: ‘The disaster prevention authorities shall prepare and update plans for protection against catastrophes and Emergency Plans for particularly high-risk objects. Leverkusen. prepare and update on-site alarm and hazard protection plans which are harmonized with local disaster prevention and hazard protection planning. passed as long ago as 20 December 1977. 3 (safety obligations). Without such systems the task forces cannot be deployed with maximum efficiency.

The Emergency Plan. The necessary details are described more precisely in the appendices . 2 PLANNING HIERARCHY FROM THE VIEWPOINT OF A LARGE WORKS In large works and with large plant units. form the Hazard Protection Plan of that plant. The following paper (Chapter 9) will give a more detailed description of Emergency Plans in accordance with the Law for Protection against Catastrophes and on-site Hazard Protection Plans according to the Major Hazard Regulations. Town Clerk. 3 EMERGENCY PLAN ACCORDING TO THE LAW FOR PROTECTION AGAINST CATASTROPHES On the basis of the legal requirement. Finally. This plan will be presented here. it is possible that various buildings of one and the same plant have their own alarm codes which. Emergency Plans are only intended for use in conjunction with the general disaster protection plan of the authorities and the alarm plan of the operator. which has been introduced by the President of the Cologne regional administration as a guideline for drawing up Emergency Plans for this administrative district. the Office for Fire Protection. not only have the terms ‘Emergency Plan’ and ‘on-site Hazard Protection Plan’ been defined. If there are several Hazard Protection Plans for plants within a works. adopts the contents of the Hazard Protection Plan of the works which are of importance for the activities of the authorities. they are combined to form the Hazard Protection Plan of the works.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 71 As a result of these statutory regulations. Rescue Services and Civil Defence of the City of Cologne drew up a specimen plan. District Clerk) and contain the special information required for implementing measures to protect the public in the event of a catastrophe. The Emergency Plans are always kept by the competent district disaster prevention authority (e. The Emergency Plan is compiled in the form of a checklist. but also their general contents. The Emergency Plans are updated annually by the district disaster prevention authority. taken as a whole. the Disaster Prevention Plan of a community also includes all the existing Emergency Plans. a cooperative effort of the authorities and the works.g.

General rules 2.2 Establishing the endangered area 3.4 Warning and informing the public These items cover the essential elements of what is known as the Leverkusen Model. Appendices A little more information will be given on the following subsections of Part 3 ‘Immediate actions in the event of an emergency’: 3. The Plan consists of five sections: 1. Description of the object 2. as well as internal and external emergency services. This includes the notification and/or alarming of management staff. Special risks situations 4. 4 HAZARD PROTECTION PLAN OF THE WORKS The Hazard Protection Plan of the works is the plan ranking above the Hazard Protection Plans of the individual plants. Immediate actions in event of an emergency 4. Addresses . Tasks of the management areas 3. The Hazard Protection Plan of the works regulates tasks and competences relating to deployment of the fire brigade and other unusual malfunctions where decisions coordinated between several managerial areas have to be taken. It represents an overview and summary of the essential measures in relation to the entire works and establishes the link to the Emergency Plan. This checklist gives the Head of disaster prevention a concise management tool allowing him to delegate orders to the members of his disaster prevention management team. Reporting and alarm-raising paths 3. The Emergency Plan consists of five sections: 1. Follow-up actions 5.3 Measuring the gas concentration 3.72 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS to the Emergency Plan. while the latter use the appendices to initiate and monitor the individual measures.1 Establishing the type of hazard 3.

a distinction is made between a main line and a secondary line. Section 5 lists all the plans. names on- . the Hazard Protéction Plan covers all potential on-site hazards (e.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 73 5. Section 3 describes the defensive measures in the event of particular risk situations. Section 4 contains the telephone numbers and addresses of the management staff of the works nominated by the areas to be responsible for the organization and implementation of all the tasks specified in this Hazard Protection Plan. tanker collisions). accident) and off-site hazards (e.g. Regardless of whether or not a plant is subject to the Major Hazard Regulations. It contains information required for planning and updating defensive actions in the event of danger. The external agencies and authorities are also listed. The tasks of the Technical Task Force (TEL) and the Works Task Force (WEL)—the central managerial bodies—are shown. it ensures the greatest possible degree of protection of life and property in the event of danger. including those of the public disaster prevention authorities. malfunctions in neighbouring plants. for which separate regulations or rules exist. Section 1 contains the legal principles on which the specifications made are based. Section 2 contains the tasks and competences of the individual managerial areas. explosion. 5 ON-SITE HAZARD PROTECTION PLANS ACCORDING TO THE MAJOR HAZARD REGULATIONS It should be pointed out that there are currently some Hazard Protection Plans which have to be drawn up because a plant is subject to the Major Hazard Regulations. fire. As regards the responsibility for informing. the contents of the Hazard Protection Plans are identical in both cases. The Hazard Protection Plan is part of the on-site safety organization of a plant. together with the general scheme for informing and reporting to agencies outside the works. Thus. Appendix Among other things.g. particularly in relation to the required notification of other or high-ranking areas. In a system of coordinated organizational measures. while others are set up by the plant in an effort to achieve the greatest possible degree of passive protection.

The specimen Hazard Protection Plan consists of two parts. It serves. as a basis for the fire brigade deployment plan. also showing the alarm scheme for deployment of all the necessary on-site technical task forces. an organizational part and an informational part. The following subdivision is recommended: — Processes (e. The organizational part contains all the measures in the form of on-site instructions governing the special behaviour of plant staff in the event of an emergency.74 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS site safety facilities. In the event of an emergency. and specifies special behavioural measures for the plant staff in the event of an emergency. following changes in the materials used. The informational section contains all the information on the plant which might be of importance in the event of an emergency. The specimen Hazard Protection Plan presented here can be regarded as a recipe for drawing up such a document. The informational section is subdivided into three parts: Plant overview. special requirements of license) — Block plan — Buildings plans . which must be coordinated with the plant management. clearly arranged and easily manageable. It is important that. whenever the plan is updated. These plans should be simple. e. Thus this plan also serves as a document for plant staff training and drills. All information on the plant which might be of importance in the event of an emergency should be listed in plans.g. the updated contents are also made known to the competent task forces. Lists of actions describing the behaviour to be adopted in the event of special hazards are enclosed. who should put their colleagues in a position to react correctly in the event of an emergency. among other things. Drawing up the Hazard Protection Plan Ideally. the Hazard Protection Plan should be drawn up by the person responsible for the plant in cooperation with the Fire Protection and Occupational Safety departments. The Hazard Protection Plan is directed at the senior staff of a plant. it tells the senior staff what they need to know and what actions should be taken. type and location of the plant. A brief description of the plant is provided here to give the task forces a general idea of the size.g.

whatever form of subdivision is used. During normal plant operation. Consequently an overview must be compiled. condensed or expanded. The task forces need this overview in order to coordinate their actions in the plant. safety officers. plant engineers. and to facilitate rapid orientation in the event of an emergency. A knowledge of all on-site hazard sources—and of those in the immediate vicinity—is an essential prerequisite for successful operations in the event of an emergency. The following subdivision is recommended: — Systems — Equipment An overview of these systems is also necessary to allow the responsible person in the plant to check the presence and condition of such systems and equipment more easily. Be prepared to sacrifice this time. hazard sources are prevented from becoming dangerous by means of prophylactic safety measures. as does any planning.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 75 Hazard sour ce overview. in the event of unusual malfunctions. Thus the list of recipients of these Hazard Protection Plans can be based on target groups. fire prevention and occupational safety officers receive the complete Hazard Protection Plan. You may then already have won half the battle for safety. However. This document should list all equipment and systems which are designed to ensure the safety of the plant staff and the use of which is governed by the safety instructions. The structure in Table 1 is just one possibility. The following subdivision is recommended: — Hazardous substances — Hazard areas — Technical facilities which may be potential hazard sources. Safety equipment and systems overviews. However. Depending on the requirements of individual plants. Drawing up plans takes time and effort. they can create a hazardous situation and lead to property damage and personal injury. Heads of department. the subdivision of the sections can be adopted as described. plant managers. it must always be ensured that the groups and/or departments requiring to be instructed and/or informed are included as target groups. .

76 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Table 1 Recipients of Hazard Protection Plans .

9 Co-operation in Emergency Planning T. have formed a voluntary Major Incident Control Committee (MICC) with the original. hosts a number of companies in the oil. Since 1968 these companies. petrochemical. together with the Emergency Services and statutory bodies. examining the organisation in each works for dealing with it. Grangemouth. exploring the integration of the various systems and setting out methods for controlling the emergency should a Major Incident occur’. on the River Forth. UK 1 INTRODUCTION The town of Grangemouth. purpose of ‘examining the material in each works and its hazard potential.DICKIE BP Chemicals Ltd. and still valid. The following list of the current members may give an indication of the Committee’s scope: Borg Warner limited BP Chemicals Limited BP Oil Limited BP Oil Grangemouth Refinery Limited Central Regional Fire Brigade Calor Gas Limited Central Scotland Police Central Regional Council Enichem Elastomers Limited Falkirk District Council Forth Ports Authority . chemical and associated businesses.

police and ambulance services etc. Each industrial member’s site is linked to the Police Station by a dedicated Omnibus telephone system which allows communication between the Police Station and either individual sites or groups of sites. but it works and can be tested readily and .’ 3 MAIN COMPONENTS OF THE EMERGENCY PLAN The Committee has long recognised the importance of having the individual emergency plans of the Emergency Services and the industrial organisations dovetail into the overall MICC plan.78 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Health & Safety Executive Macgas Limited Imperial Chemical Industries Limited Rohm & Haas (UK) Limited Ross Chemicals and Storage Company Limited Scottish Ambulance Service 2 DEFINITION OF A MAJOR INCIDENT The definition of a Major Incident has developed over the years as earlier definitions were tested by incidents which appeared to approach the borderline. beyond the normal call-out attendance. and perhaps rather old-fashioned. and/or (b) requires external aid in fire-fighting. the current wording is ‘A Major Incident is defined as an industrial incident which (a) is likely to affect. This is very simple. Main Control is housed in a suite of 3 rooms designed. There is a central room used by the Police with one adjacent for the use of Technical Advisers and another for officers of the Fire and Ambulance Services. equipped and maintained specifically for the purpose. The plan provides for a ‘main control’ centre in Grangemouth Police Station and forward control nearer the scene of the incident. or is affecting. and for this and other reasons the overall tone is that of simplicity. the safety of people outwith the undertaking. after consultation by these Emergency Services and site managements.

The plan is co-ordinated with the Central Regional Council offsite contingency plan prepared under the CIMAH Regulations. each week. supplemented by the radio communications of the Emergency Services and the industrial members. To eliminate any doubt or hesitancy at the outset of an incident. It also has a section listing the equipment which can be made available from member companies on a mutual aid basis if required in the event of an incident. The Committee was also used as the vehicle for a joint approach by the top-tier sites through Falkirk District Council to implement the requirements of the CIMAH Regulations regarding notifications to the public. the first call to the emergency services goes through via the normal telephone system in the standard way. The Committee is used to obtain the agreement of the companies to the off-site contingency plan and any amendments to it. one of the industrial members initiates a small exercise by calling the Police Station on the Omnibus telephone. for gas releases and for radioactive substances. but it is quite common to encounter difficulties with radio trans-mission/reception in and around plant structures. There are three levels of ‘exercise’ in use at present. The reference to Technical Advisers above comes from the plan which provides that. replacement or stand-in staff to become familiar with its use. each industrial member sends one senior member of staff to Main Control to assist with technical advice and information on the availability of mutual aid. notifying the Police of an incident and asking for assistance which may be in one of several forms such as setting up road-blocks or . but provides a regular opportunity for new. of course. The plan provides for an industrial gas detection service by which three of the industrial members provide radio-equipped vans manned by suitably trained staff to monitor the level of gas in the atmosphere in any given area. 4 EXERCISES The main purposes of exercises are to train staff in the procedures and to test and improve the procedures themselves.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 79 regularly. in the event of a Major Incident being announced. Also. This very simple test not only ensures that the equipment is working properly. It is. The plan has specifie sections for the Grangemouth Docks area. Each week the Police Service tests the Omnibus telephone circuit by calls to each member.

and the plan has been modified to reflect this.80 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS having another member provide some specified emergency equipment. and the Emergency Planning Officer arranged a suitable training course making use of speakers from the MICC. This aspect belongs to the Regional Council’s off-site contingency plan. Less frequently. e. The Emergency Planning Officer of the Regional Council used the incident as a springboard for further training of Local Authority organisations such as the Housing or Education departments and voluntary organisations such as the WRVS in the roles which they might be called upon to play. in the last year the arrangements for information flow within Main . this is simple but ensures familiarity with the system. on communications or on updating the manual. allowing staff of the emergency services and the technical advisers to concentrate on dealing with the implications of the incident. 5 EXPERIENCE OF AN ACTUAL INCIDENT In 1986 the plan was used in earnest when a small leak of bromine gas from one of the factories affected people in neighbouring premises and threatened to affect people in houses nearby. more ambitious exercises are held. Again. 6 UPDATING THE PLAN The Committee meets quarterly with ad hoc Sub-Committees. currently about twice per year. due to restricted availability of manpower. For example. The plan itself is not static. The arrangements worked well and it was clear that the time spent in developing the plan and training people meant that the administrative side fell into place. These involve setting up and manning forward control and main control and working through a ‘scenario’ devised by an exercise sub-committee. These exercises have involved the deployment of police to road blocks but. One lesson learned which had not become apparent from the exercises was that the Department of Environmental Health of Falkirk District Council became a main contact point with the public for the flow of information. meeting separately as required. have not involved the deployment of fire and ambulance personnel beyond the attendance of officers at forward and main control.g.

the Committee members recognise that. while they strive to ensure that the plan is effective and that personnel are trained in its implementation. (Experience has shown that wind direction and speed can vary considerably within the area. and the locations of anemometers held by members have been listed in the manual. .) In conclusion. it represents one part of a wider safety process which has the prime aim of preventing emergencies as well as containing any which may occur.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 81 Control have been improved.

which to some extent occurred periodically in specific areas. The majority of these events were usually natural disasters. The organisation. Generally. as well as the selection and training of suitable personnel for these tasks. large fires. additional steps were taken to prevent these disasters and plans developed to limit their effects. and a preventive protection plan on the other. health and property of those affected. Urgent decisions were taken by those responsible in each case without any forward planning of support services. these plans encompassed the provision of aid for combating disasters. such as floods. scope.10 Emergency Response Planning OffSite of Chemical Plants BENNO KIER & GÜNTHER MÜLLER Rheinisch-Westfälischer Technischer ÜberwachungsVerein eV. on the whole. the rescue and the care of the population. The type. emergency response and contingency planning was. There was no evidence of any major link between the actual causes of accidents and sources of hazard on the one hand. FRG 1 INTRODUCTION In the past. the recovery of and caring for victims. This procedure was also adopted in principle for those additional dangers which arrive with the increasing mechanisation of traffic installations and industrial plants. based on lessons learnt from previous events. as well as the evacuation and accommodation of parts of the population in special cases. Specific disaster management planning was first undertaken with the introduction of nuclear power because of all the Radiation hazards associated with it. chronological sequence and range of possible accidents were incorporated into . avalanches or earthquakes. Because of the great danger this caused to the life. Essen. training and equipment needed for disaster management were geared to the aftermath generally to be expected from serious accidents in these installations— mainly to the fighting of fires.

This concept allows detailed prior planning of all necessary defence measures against each particular incident. In addition. as well as to compile the major requirements for plantspecific emergency response plans. It can be assumed that. operators and authorities. Even though emergency response plans have already been drawn up available for a range of installations in the Federal Republic of Germany.g. and also a reality-based testing for such measures. as were the special technological aspects of the individual plant and its respective location. both in Germany and abroad. experiences accumulated to date are being evaluated and supplemented. without any secrecy restrictions. in which hazardous substances are produced. to the concentration of an increasing number of different installations and works on single sites. orography). the safety of such installations regarding incidents and the associated effects has been constantly adapted to keep up with the developments in process technology. . to harmonise them with the plant-based safety precaution plans and to ensure an exchange of information between the public bodies responsible for contingency planning. stored and handled. or replaced by new processes. It is especially necessary to adapt the emergency measures to the situation at the specific plant (e. Even so. population density. size and hazard potential) and at the specific location (e. the Federal Office for the Environment in Berlin instructed us to conduct a study in 1981. For this reason. increasingly. The aim was to throw more light on the use of the existing emergency response plan. traffic routes. especially in the chemical industry. and preventive measures have been organised in certain areas between the operators and the authorities responsible for the prevention of catastrophe under special agreements. at the same time. show that accidents with adverse effects on the surrounding area cannot be completely eliminated by acceptable technical means. existing installations near residential areas are. developed. thanks to the efforts of manufacturers. being renovated.g. Conventional industrial plants. In order to draw up suggestions for a common guideline. both in Germany and abroad. there is still a shortage of uniform regulations. treated. type. basic scientific considerations and actual far-reaching incidents in chemical installations.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 83 this plan. also have a high hazard potential Over the past few years this technical development has led to ever larger production units being erected and.

On the basis of general disaster management planning. Specifically. descriptions and prepared specific measures. Dynamic emergency response relies on high flexibility in the leadership and services without the detailed prior planning of corresponding individual steps. the orographie conditions prevailing and the current meteorological conditions. detailed planning provides for highly differentiated instructions. This is because of the high physical and psychological burden on all those involved in a hazard situation. It became clear that the legal stipulations represent a substantial prerequisite for effective emergency response planning and are particularly important for the allocation and demarcation of the tasks and areas of competence of those involved. It concentrates on the decision-making capacity of the leadership of the services. It provides decision-making aids with regard to immediate measures and provides all the necessary information for further procedure. authorities and competent associations were questioned and the relevant literature was evaluated in order to determine the state of emergency response planning in Germany and abroad. the rapid mobilisation of personnel and the constant availability of technical equipment and aid. Planning must also take account of the residential and traffic structure of the site. In contrast to this. depends to a crucial extent on the circumstances of each individual case. The investigation revealed that the spectrum of planning variants ranges between two basic views. there were found to be considerable differences with regard to the measures to be taken to meet these requirements. Such measures and preparations relate in each particular case to the specifie features of the particular objects whose storage facilities. Accordingly. Converting them into concrete plans. production facilities or special properties mean that the possibility of a hazard arising cannot be discounted. within the legal framework. hazard and location) . the official emergency response plans include specific and specialised measures and preparations with regard to particular hazardous objects. however.84 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 2 RESULTS OF THE INVESTIGATION Operators of plants. This applies to both the organisational requirements and the personnel and technical equipment needed. the following aspects were evaluated: — Structure and content of disaster response plans — Characterisation of hazard potential (description of object.

on the other. animals. on the one hand.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 85 — Cooperation between operators and authorities — Organisations involved. by disaster management legislation and. This obligation is covered by a series of regulations on a Federal and State level. substantial disadvantages and substantial burdens which may be caused in some other way. The existing regulations contain the legal possibilities for emergency response for such planning as well as clear statements regarding their requirements. traffic control. The government bodies are also to ensure that the effects of the residual risk are countered without delay when there is an alarm. inasfar as plants are concerned which are subject to licensing. by anti-pollution legislation. equipment and aids available — Service strategies (determination of situation. This requirement is put into concrete form in the Incidents Ordinance (12th BImSchV). also from hazards. This . alerting of the general public. and to prevent damaging environmental influences from arising’. Already this statement of the Act’s purpose directly imposes an obligation to ensure preventative protection against hazards. initiation of measures and control of the defence operation — Interaction of various types of service teams — Documentation of the incident chronology and the defence measures taken 3 CONVERSION OF THE TEST RESULTS FOR A DRAFT GUIDELINE The legal framework for the preparation of hazard prevention plans in the vicinity of conventional plants with high hazard potential is laid down. but its possibility cannot be excluded entirely. entailing not only state-of-the art safety precautions but also emergency response measures. in conjunction with the ordinances and administrative regulations passed on the basis of the Act. ‘to protect humans. The Federal Anti-Pollution Act is mandatory for the plant operators. The Disaster Response Act (Katastrophenschutz-Gesetz) obliges govern-ment bodies to go beyond the safety of plants and transport facilities by also ensuring that the effects of any residual risk remain limited. evacuation. vegetation and other objects from damaging environmental influences and. Its purpose is. assistance) — Practical instructions and decision-making aids for the alerting of disaster services. Such a risk may be rated as a very minor factor in the technical rules.

Thus a hazard analysis should be considered as a first step.86 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS ordinance lists. For the individual substances. From this we obtain starting points for objectrelated or plant-related emergency response. The Second General Administrative Regulation to the Ordinance Decree (2nd Störfall VwV) expressly mentions related details of organisational protection facilities and operational emergency response plans. the substances to be anticipated are allocated to the individual plant parts. such as — Apparatus and vessels for hazardous substances and hazardous processes — Possible but undesirable reactions of substances being handled — Pipelines and pipe bridges for hazardous substances — Loading sites for hazardous substances — Sites with high fire hazards Where there is a hazardous situation. and on the special conditions and prerequisites in the area around the site. it is crucial to have information on the type and properties of substances with which the emergency services or general public will be confronted. An object description is drawn up first of all. Such plans have to be drawn up for special hazardous objects within the framework of general disaster management. among other things. on the soil. strategy and logistics of the plant-specific emergency response have to be based on the type and scope of the potential hazards for a particular plant. operational alarm and emergency response plans to limit the effects of incidents. These starting points are reflected in the requirement for special protection plans in the disaster response acts of a number of federal states. instructive details should be provided of — The properties of the substances — Their possible reactions — The effects of the substances on the human organism. further information is useful to ensure emergency response. Further federal and state regulations can be consulted to back up and fill out specific prevention planning in the surrounding area of hazardous plants. on animals and vegetation. on agricultural areas and on food supply operations . In a works analysis. especially chemical plants and refineries. on water and fish. plans which are to be drawn up and updated to fit in with the local disaster management and emergency response planning. including a list of all identifiable hazard sources and hazards. In addition. The planning.

0·03 bar.g. and in addition the objects requiring special protection in the potential impact area have to be determined. in the case of a tank explosion) — Explosion. take measures — Details of the probable chronology of hazard effects (e. details of the chronology of an event are also important. which is to be drawn up by the producers of hazardous substances. propagation characteristics — Shock wave with distance parameters for the effects of the pressure (e. with specific reference to individual substances. discharge rate.g. In addition to the effects of substances. Objects requiring special protection include. the maximum range of the effects has to be estimated in detail. 0·01 bar) — Devastated areas (e. the following details are also relevant at this stage.g. deflagration — Contamination of soil and water. warn the public. can serve here in particular as a source of information. the conceivable forms the hazards could take should be listed: — Fire over a very wide area — Gas discharge: possible quantities. filters) — Possibilities for counter-measures — First aid measures for injured persons — Therapy in the case of hospitalisation Finally. duration of discharge. in order to cope with the consequences resulting from a discharge: — Protective facilities (e. Such details include — Estimation of the time spans between when a hazard situation arises or is detected and conceivable effects in the surrounding area — Establishment of the probable remaining time to alert the emergency services. in the case of airborne emissions in various meteorological conditions) To delimit the surrounding area which the plans are to cover. with indication of the effect boundaries With a view to the fact that there may be an alarm or evacuation of the population affected by a hazard.g.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 87 The DIN safety data sheet. 0·35 bar. for example .

social welfare facilities.e. they make possible a reliable and complete check of the measures to be taken. i. Specially suitable for this purpose is the documentation of accidents and incidents where the events have been recorded in standardised form. waterways and stretches of water. alert calendars should be drawn up and made available at the individual points. A valuable aid in hazard analysis.88 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — Installations and facilities which themselves have a hazard potential — Densely populated residential areas — Hospitals — Schools. 3 of the Incidents Ordinance The details are listed in a practical form and presented in plans. When an alert situation has been established. lodging facilities — Drinking water source areas. meeting places. To enable the competent bodies to take necessary steps in a correct and complete fashion without any delay. accidents and incidents. . as has been done in the systematic recording of facts relating to damage events. as needed in the specifie situations. suitable measures have to be initiated immediately. Paragraph 2. The alert procedure should be divided into stages which involve the relevant services according to the increasingly hazardous nature of the situation. in particular with regard to determining the causes. bodies such as — Management of the disaster services — Police — Fire brigade — Medical services — Social welfare services Since they can be used as check-lists. Alert schedules are a list of the conceivable activities and arrangements for the various bodies. where such facts are important for damage prevention and emergency response. is the evaluation of past events. No. Events which necessitated major counter-measures are especially useful in indicating how to draw up new plans and updating existing emergency response plans. the chronological sequence and the spatial extent of the effects of accidents. traffic routes — Plants and stores connected with foodstuffs — Other objects of high value in the meaning of Article 2.

through high concentrations of hazardous substances in the ambient air. subsequent explosion. A preventive evacuation of areas in which a hazardous impact may arise. The preparation and implementation of any evacuation measures and the decision to set them in motion have to take account of this time dependence. spread of a fire. A decision on the relevant procedure to be adopted can be taken with the help of an estimation as to the impact alternatives to be expected.g. spreads). toxic deposits.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 89 Evacuation of large areas in the vicinity is an especially difficult task in emergency response. short-lived incidents which have an impact lasting only up to a few minutes. On the other hand. assumes the most detailed knowledge possible of the hazard situation and its further development. it must be noted that the exposure of these persons and the disaster services to the effects of an incident during evacuation may also constitute a considerable hazard. Evacuation after the onset of an incident is only meaningful if persons in enclosed rooms are at direct risk because of a persistent exposure of the area concerned (e. It may be necessary to evacuate neighbouring plants or residential area if — after the onset of an event it is found that no adequate protection is provided by staying in enclosed rooms there. severe thermal radiation). or — it has to be feared that. or can be expected as an incident develops (e.g. change in the direction of propagation of toxic substances). this protection can no longer be guaranteed The possibilities for evacuation are thus determined to a crucial extent by the time-related nature of the incident’s level of impact on the surrounding area. as the disaster develops (e. With sudden. the smaller the effort involved in the evacuation and the risk for the persons and services affected. evacuation can be practically discounted because of the practice already described and the preparatory time needed.g. The earlier such preventive measures can be initiated and concluded. .

this act gives the obligations to the authorities to draw up emergency plans with respect to disasters which can happen at undefined places.). As a second point. the obligations are laid down in the Act on Calamities.11 Industrial Emergency Planning in The Netherlands H. the authorities have to set . In case the effects of such a disaster extend beyond the municipal borders this act lays the competence at a higher administrative level (e. the Provincial Governor or even the Minister of Home Affairs). The Netherlands 1 INTRODUCTION Although there are nowadays many chemical factories.O. Since January 1985 this act has been in force. In this act it is stated that in case of calamities.VAN DER KOOI & H. Thirdly. There are two main parties involved in the action to be taken in emergency situations: — Public authorities — Companies With respect to the activities of the public authorities. producing thousands of dangerous chemicals. only in certain big companies are there extended emergency plans and schedules. particularly the Mayor. small and medium-size chemical industries have not such well developed emergency plans. the aftermath of large-scale incidents at a chemical plant. Voorburg.g. However. Generally. An explanation will be given about the practical situation concerning the Industrial Emergency Plan in Holland. it is the duty of the authorities in the European countries on behalf of the EC Seveso Directive to ensure that there should be emergency plans in companies with major hazard risks. the supreme responsibility in righting the disaster is in the hands of the public authorities. etc. railway accidents.VUYK Ministry of Social Affairs.K. of whatever nature (airplane crashes.

This requirement can be ordered to major hazard companies. explosion hazard (c) The possible extent of the effects of pressure waves. in accordance with the so-called notification in the EC Seveso Directive.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 91 up so-called calamity fighting plans in case the risk of such a calamity is confined to a distinct area and connected with industrial activity. 2 INDUSTRIAL EMERGENCY PLAN In our opinion the following starting points are important in making an Industrial Emergency Plan: — Evaluation of the risks based on safety analysis (a) Insight in the amount of fire-fighting and rescue equipment (b) Nature of the dangers as toxicity. heat radiation. With respect to the activities inside industry. The requirements of an industrial emergency plan are given in a publication ‘Guideline for the setting up of an industrial emergency plan’. dispersion of toxic clouds (d) The probability that such a calamity can take place — Preceding research of the company’s own means and services — Relation with the organizational set-up of emergency response systems by the public authorities — Participation in mutual assistance systems Among the basic elements of an Industrial Emergency Plan are (A) Management and coordination. The designation of the major hazard companies takes place on the basis of having certain minimum amounts of dangerous chemicals. (B) Industrial coordinator at place of accident. He is located in a head coordination centre which is equipped with all the necessary facilities. there is a requirement in accordance with the Working Environment Act of November 1980 to have an industrial emergency plan. In most cases this function belongs to the managing director of the company. It is preferable to coordinate the actual activities of assistance and fighting by a . Industrial Head Coordination has the overall management and coordination in the emergency activities inside the establishment. These companies are already obliged to draw up an occupational safety report. which will be officially published by the Directorate General of Labour in the near future.

giving recommendations concerning safety . (a) Industrial fire brigade. registration of visitors and announcing the special situations to them. 3 FIRE-FIGHTING AND ASSISTANCE IN EMERGENCY CASES Various industrial services have a duty in fighting. He declares the state of emergency on the industrial area and eventually orders shut-down of installations. Each service acts according to a prepared action-plan. the prevention of escalation. (c) Industrial safety service. Such an action-plan varies in accordance with the severity of the emergency situation. (b) Security service in emergency cases. The Industrial Coordinator at Place of Accident contacts the assistance services and the head coordination centre by means of telephone or orderly. emergency power provisions and. Assistance in the evaluation of the appearing hazards. The production manager is most suitable for doing this job. The plan of attack should describe the detailed situation in the plant: storage and amounts of chemicals. the protection of apparatus and the rescue of victims. being informed about the admission of emergency staff and people and also the authorities etc. properties of the chemicals. fitting into the total industrial emergency plan. The tasks of fire brigades are of course the firefighting. knowledge of the electrical power system. and controlling an accident. (C) Overall management and coordination in the affected area. removal of unauthorized persons. These tasks and instructions should be recorded in an action-plan fitted into the Industrial Emergency Plan. assisting at. All people who take part in fighting the calamity are under the command of the Mayor. Control of the traffic through the gate. blocking the disaster area. keeping the roads free for fire-brigade and other assistance. extinguishing foam and special appliances. In the case of a great catastrophe this is in the hands of the Mayor of the municipality. In any case there must be rules of consignment to be sure that at any time such a functionary can be committed immediately after the event. the relations with the Industrial Coordinator at Place of Accident and the Head Coordination Centre. The Fire-master is in charge of the operational management.92 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS person closely involved in the industrial activity. provisions of firefighting water. not least.

Also procedures should be described for the call-out of key personnel and for mobilizing emergency services. There should be a technical emergency action-plan which describes the emergency power provisions. MOBILIZING PROCEDURES AND COMMUNICATIONS Procedures must be available for triggering and sounding the alarm if the latter is initiated from a central communication post. to the Industrial Coordinator at Place of Accident and the Industrial Head Coordinator. Plant control and emergency shut-down are his responsibility and should be described in emergency procedures. as well as the systems and devices for communicating information to staff and plant personnel. the disposition of electricians for securing of apparatus for safe extinguishing. the coordination during the various stages of the accident. In case another PR functionary has been designated. For effective medical assistance a medical emergency plan is desired. It can be important to have certain specialists at hand in case specifie hazards are present. (f) Engineering services. major emergency) each particular alarm will have to be described . the transport priority to hospitals and the consultation of the National Toxicity Information Centre should be described. and the relations with the other services should be stated.g.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 93 measurements. the procedures for alerting the health service. (g) Specialists. the medical help. During a major emergency the Industrial Head Coordinator will report to the media. the assistance of other organizations. their identification. Red Cross places for the interception of wounded people. 4 ALARMS. e. Specific instructions must be available for this functionary. and gas surveyors for determination of explosive and toxic gases. (d) Industrial health service. (e) Production departments. specialists on radioactive materials. Besides that. fire. e. Reports all necessary information concerning the accident. the regional health service and the ambulance service.g. If there is a hierarchy of alarms (minor. In such a plan the organization of this service. (h) Public relations. gas specialists. provisions must be made for his accommodation. gas. etc. The production manager has the coordination about all activities on plant-operational level in case of an emergency. determining the dispersion of toxic emissions. causes and course of the fighting.

This is to be developed from simple to fully simulated exercises for various scenarios in which external services participate. In case of a calamity/emergency situation it should be obvious for everyone in the company when. In most cases the fire brigade is alerted at first. an emergency situation exists. The decision of evacuation is made by the Industrial Head Coordinator in consultation with the Industrial Coordinator at Place of Accident. An evacuation action-plan must describe how workers can be evacuated safely.94 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS as well as alternative means or back-up systems in case the normal system fails. in accordance with the particular alarm type. should be indicated.2 Training and exercises Having established the emergency plan. Procedures should also be included with respect to giving the ‘all clear’ signal. it is essential that all people involved are trained and exercised with adequate frequency in emergency situations to create expertise and confidence in the emergency plan. Also the surrounding companies should be informed about the situation. and in which area. At the same time the public authorities should be informed about the emergency situation. . 4. Also the required action of key personnel and emergency services. 4. In major emergency situations the decision will be taken by the public authorities.1 Declaration of the major emergency situation and evacuation This is a primary task of the Industrial Coordinator at Place of Accident.

whatever the degree of . As regards hazards connected with the storage. generally throughout an entire department when local emergency arrangements prove inadequate. the experience of recent years has shown that accidents can occur which exceed the risk management capability of a single department. This revision of operational planning to deal with technological hazards was essentially the outcome of the following factors: — The development of modern industry and technology had engendered new hazards which. To deal with situations resulting from such major disasters it has been decided to apply the principle of the ORSEC plans to larger areas (zones de défense). The operational validity of the concepts on which these plans. However. transport or processing of chemical and toxic substances.12 Emergency and Intervention Plans: The French Experience M. France Emergency planning for natural or technological disasters is at present based on the ORSEC plans for each of the French geographical departments and the technical annexes to those plans. The emergency plans make it possible to mobilise and commit resources on a large scale. which have moreover been adopted by several foreign countries.GENESCO Direction de la Sécurité Civile. are based has been successfully put to the test on numerous occasions. will not replace planning for specifie accident situations. to cover accidents within industrial installations capable of causing damage to the population or the environment. which establish in general terms how official intervention is to be organised. these new operational arrangements. new emergency planning was necessary similar to that already undertaken for radiological hazards. Paris. However. This is the objective of the new departmental plan known as ‘ORSEC risques technologiques’ which will replace the plans known as ‘ORSEC hydrocarbures’ dating from 1967 and ‘ORSECTOX’ dating from 1973.

in close collaboration with the government departments concerned. i. the ORSECTOX plan. local authorities and the media.e.96 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS prevention or of industrial safety and reliability. A considerable part of these plans deals with emergency warning arrangements and information for the population. In this context. together . in particular. prescribed various measures relating to both prevention and protection of the population of which account had to be taken. concerning major accident hazards posed by certain types of industry. despite the strict preventive measures imposed by French legislation. both to provide information and to carry out relief operations. the special protection plan — The ever increasing complexity of the chemicals industry has meant an increasing concentration of risks and required better emergency planning to cope with all eventualities. For this purpose.g. It was sent out in July 1985 to all local authorities concerned. including stocks of hydrocarbons with a capacity of over 600 m3. led to the creation of petrochemical complexes within which there were a variety of plans with the same objective. some 300 in France. hydrocarbons) on the lines of those prepared since 1978 in regard to radiological hazards. The departmental plan known as ‘ORSEC risques technologiques’ is therefore the outcome of inter-ministerial work over a two-year period involving the Ministries of the Environment. drawn up operational emergency plans for technological hazards (chemicals. Henceforth the indispensable collaboration between industry and government in the event of accident. notably by the conclusion of information agreements between the commissioner of the Republic and the plant operator. will take the form of a narrow interface between an ‘on-site emergency plan’ (POI) for which the operator will be responsible and a ‘special (off-site) intervention plan’ (PPI) drawn up and implemented by the commissioner of the Republic. Defence. Health and the Interior. It will be accompanied by an evaluation or revision of the risk assessments for each establishment and correlatively by an exhaustive and detailed analysis of possible accident scenarios. This new form of emergency planning will apply to industrial establishments or plants covered by the Seveso Directive. the heads of the emergency services have. could never be entirely eliminated — Industrial change had. the so-called Seveso Community Directive of 24 June 1982. Industry. e. the Hydrocarbure plan.

. The PPI (special intervention plan) deals primarily with how outside relief is to be organised and with information to be provided to people living near the site. or in terms of the volume of hazardous substances transported. and will represent a significant additional capability in this area. in 15 departments in 1986 and a similar number in 1987. the plan has three levels of application depending on the scale of the accident. specialising in nuclear and chemical hazards. Level 1: Non-toxic accident Level 2: Toxic accident confined to the installation Level 3: Toxic accident with off-site consequences The PPI also covers situations arising as a result of unlawful acts perpetrated against installations. The POI naturally reflects the pre-established risk assessment and lists appropriate resources and facilities at the disposal of the operator to deal with the situation.e. — Rapid emergency warning and information arrangements have been made for populations which could be affected by the consequences of an accident. under the responsibility of the commissioner of the Republic or his representative. numerous arrangements have been or are about to be made: — Mobile chemicals emergency (CMIC) have been set up in the departments which are most at risk. — An intervention unit of the emergency services (UISC) has been created. how the various command posts are to be activated. To ensure that the new plan is operationally effective. Such resources are initially to be implemented under the authority of the plant manager and subsequently. it states how emergency warnings are to be given. owing to the number of installations requiring a PPI. should the accident spread or threaten to spread outside the plant. i. It also stipulates exactly how internal and external emergency resources are to be used. The POI (on-site emergency plan) lays down what the plant operator must do to safeguard his installation and workers. — Heads of agreement have been established between the authority responsible for emergency relief operations and local radio stations. In addition. Like all plans of this type based on the ORSEC approach.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 97 with a time-table for the preparation of a POI and a PPI for each installation affected. and what is to be done by local and regional government services.

Cl2. Application threshold of the Seveso Directive 2. in particular: a) protection of individuals. Lessons to be learnt from this event 1. APPENDIX: THE NANTES ACCIDENT A. . Immediate evacuation within the security perimeter 2. A nearby silo containing 850 tonnes of ammonium nitrate did not explode. Preliminary confinement of the population to their houses 3. The combustion of these products gave rise to emissions containing. Evaluation of the consequences: — difficulty of discovering the origin and composition of the products (industrial secret). 45 000 inhabitants including 8000 school-children and several old peoples’ homes) 4. NO and NO2. — need to be able to mobilise specialised means quickly. Interruption of traffic on the Loire and the Nantes-St Nazaire railway C. — difficulty of quick disposal because of the nature and concentration of the toxic products generated by the combustion. and NH4.98 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — Lists have been prepared of computer data-banks providing information about hazardous and toxic substances so that relief workers on the spot can be better informed. HNO3. The following appendix deals with the fire which occurred at Nantes just a short while ago (29 October 1987). Evacuation of the population of 7 parishes (i. B. Protection measures taken 1. in particular. and b) analysis and identification.e. Products involved NPK fertiliser in the form of chloride and nitrate.

— effectiveness of the ORSEC plan.36 Beginning of the fire alert 10.08 10. Sources of supply lacking End of toxic emissions Return of evacuated people The checks continue to their houses throughout the night The ORSEC plan terminated . — exemplary reactions of the population.20 11. The wind direction changes.ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN 99 3. — solidarity observed between the evacuating parishes and the welcoming parishes.00 07.00 Fertiliser silo (Frensit storage) Installation of the mobile First sampling of toxic PC vapours Setting up of the 500m Evacuation of 3 poisoned safety perimeter people Alert of the national authorities. Supplementary means of The evacuation continues protection for people and accommodation proposed Fire not extinguished.20 11. Chronology for 29 October 1987 09.10 17.00 22.15 16. Measures for protecting the population: — benefit in having radio communications available.00 18. Setting up of the crisis PC Sending of airborne Preliminary confinement of monitoring systems and the population specialised means of analysis Triggering of the ORSEC 43 000 people involved plan 20 000 evacuated Evacuation of the 25 taken to hospital population Results of the measurements Fire contained The toxic emissions continue.00 16.40 14.30 19.

100 .

HEFFERNAN Department of Labour. Ireland . FRG Rapporteur: J.SIGEMUND State Ministry of the Interior.SESSION III Exercises and Auditing of Emergency Planning Chairman: H.

consequently. Within these installations. When there is a major accident. responsibility is assumed: — at — at — at — at the the the the municipal level (mayor) by the SMPC.PARANHOS TEIXEIRA National Service for Civil Protection.13 Plan for Off-Site Exercises A. national level (prime minister) by the SNPC. or a catastrophe which has considerable effects on large areas and their populations. . The Portuguese National Civil Protection Service is a very decentralised service and. responsibility is assumed by the National Service (SNPC) which draws up plans and carries out civil protection operations. and for which one can see. that the lowest level will not be able to deal with it. however. Lisbon. regional level (regional government) by the SRPC.M. Portugal 1 GENERAL COMMENTS 1. district level (prefect) by the SDPC. or the massive escape of several tonnes of toxic gas in the Estarreja region. as soon as it begins. the company where the accident takes place is responsible. We will now deal with this last example.1 Responsibility for the civil protection plans The drawing up of civil protection plans and the carrying out of civil protection operations in Portugal is limited to the effects of major accidents outside manufacturing installations. As an example we may cite a serious earthquake at Lisbon.

The dominant winds generally blow. is situated north of Aveiro in the coastal plain where the height above sea level is less than 50m. The average speed is about 4–5 km/h. Some 3 km to the NW of the town there is a group of four chemical factories which produce several very toxic. — from E to W for about 1 month per year but irregularly. and all the safety standards imposed by these authorities are respected. Cl2 (toxic) — Isopor factory: phosgene. CH2CHC1 (explosive and toxic) (Chlorine and phosgene were used as combat gasses in the First World War. .2 Characteristics of the Estarreja case Estarreja. and sometimes explosive. with some regularity: — from NW to SE for about 8 months per year. the industrial region is surrounded to the east and west by large tree-covered areas (Pinus pinaster and Eucalyptus globulus). NH3 (toxic and explosive vapours) — Uniteca factory: chlorine.) Account must also be taken of how the gases react together. except for the SW winds. The nearest heights above 50 m are to the east of the town about 2 km away. of which the average is about 20–30 km/h with gusts which may reach 50–60 km/h. — from SW to NE for about 3 months per year.EXERCISES AND AUDITING OF EMERGENCY PLANNING 103 1. The population of the town and its surroundings is about 15 000 inhabitants. who are mainly employed in the agricultural sector. Their activities are in accordance with the legislation in force. These factories have been constructed according to all the standards imposed by the law and are authorised by the Portuguese administrative authorities. as the region is one of the most fertile in the country. The town and these heights are separated by the Antua river which is thus an obstacle to rapid evacuation of the town’s population to the higher surrounding ground. The region is crossed by roads and paths in several directions and by the Lisbon-Porto railway in the N-S direction. gases which are heavier than air: — Quimigal factory: ammonia. This railway crosses an industrial area situated NNW of the town. COC12 (toxic) — Cires factory: vinyl monochloride. with local government offices.

an emergency plan. could also be affected.) . though probably less serious. etc. Avanca. Thus the plan. As well as the risk already mentioned. The situations which follow must be envisaged from the need to have available. anticipates and governs the use of means (human and material) and resources of the region. These risks. still worry the Portuguese administrative authorities. one may also say that all the industrial equipment is up to date and conforms to the safety standards established by the law. boy scouts. One must never exclude the unexpected occurrence of violent events which could suddenly lead to a major accident. in advance. These means are: (1) On the local scale—municipality coordinated by the SMPC): (immediate action — The factories (companies) — The fire service — The Red Cross — The security forces (police) — Health structures (public and private) — Private service organisations (business. Anjeja and others of considerable economic importance. such as Murtosa. One may thus assume that it is very unlikely in normal conditions that a major accident which goes beyond the limits of the industrial area would occur—but it cannot be ruled out. transport) — Other voluntary organisations (amateur radio operators. this plan envisages several other risks which. 2 THE EMERGENCY PLAN The emergency plan for the town of Estarreja and its surroundings was prepared by the SNPC in collaboration with the SDPC and the SMPC. to deal with the events. which are considered minor. Neighbouring villages.104 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Furthermore. cover a vast collection of accidents and catastrophes which range from a road accident of a tanker carrying dangerous materials to forest fires and problems of pollution of the environment. or of the entire country. based on assumptions which have some probability.

that a dangerous event has occurred. burial of dead animals. factories and other places where there is a large concentration of people During: by the automatic alert. Observers from Member States may attend. as quickly as possible. re-establishment of normal living conditions (of health. as part of the European Year of the Environment. during and after the emergency: Before: based on the evaluation of risks and information and the informing of the population by means of conferences.1 Assumptions This exercise. and the performance of exercises and progressive training. purification—cleanliness. rehousing of evacuated people. takes one of the situations envisaged in the . it has been decided to carry out an exercise in the region. considering the importance of the industrial area of Estarreja and the fact that its emergency plan has not yet been approved. diffusion of individual measures to be taken.) 3 THE EXTERNAL EXERCISE Although similar exercises have been carried out in Portugal. etc. 3. The establishment of such a plan must lay down and regulate the actions to be performed before. in the case of a gas leak.EXERCISES AND AUDITING OF EMERGENCY PLANNING 105 (2) On the complementary district (prefect) and national (government) level (coordinated by the SDPC or SNPC): — The Armed Forces — Hospitals and health organisations — Public service organisations — Social assistance organisations — The fire service and the Red Cross — etc. burning of debris. which will be preceded by a vast campaign informing the public. and by the immediate triggering of the measures laid down in the plan After: by decontamination and neutralisation of chemical agents deposited on the soil. especially in schools. in February 1988.

by the ‘communicating vessel’ principle. — This release produces a toxic cloud (plume) which has roughly the shape of an irregular and elongated ellipsoid. fire service. where it can remain for up to 18 h). and informs the external authorities (SMPC. 3. hospitals). are as follows: — length (in the wind direction) 3500 m — width (transverse) 600 m — height 50 m — This cloud (plume) is formed from a core of pure gas and moves at wind speed (4–5 km/h). The gas may infiltrate into the houses. heavier than air. on the radio and television . The factory where the gas leak occurs immediately activates all its warning systems (sirens. evacuation of gassed people by helicopter. particularly simulations of the evacuation of children in schools. its approximate dimensions. as well as in the houses which are lower than the cloud. telephones) and its own means of assistance. 1st phase (preliminary preparations) Educating and making the population of the region aware by: — Conference to be organised in the communities. — It is impossible to repair or remedy the damage before the total outflow of the gas. in these places and at this precise moment. temporary interruptions of road and rail traffic. until it is completely dispersed (apart from caves and the lowest points. windows and drain pipes. It dilutes towards the surface and lasts for about 6 h. This exercise is prepared in three phases. being able to reach the villages in its path. Red Cross. etc. at soil level. from a sphere where it was enclosed at a pressure of 10 bar. but there may also be several practical actions on the ground. actions of decontamination. According to the prevailing meteorological conditions.106 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS emergency plan— perhaps one of the more severe—based on the following assumptions: — A violent event leads to the uncontrolled release of 50 tonnes of a toxic gas.2 Preparation of the exercise This is a ‘planning’ exercise. — its effects will be felt along a ‘corridor’ which the cloud follows. at the beginning. which it is estimated will take 1 h. by the doors.

etc. school children and. indirectly. on the envelope one will write only: (a) the name of the issuing organisation (b) the time of the opening and the consequent transmission of the message — Preparation of a dossier which must contain: . the fire service. health services. the Red Cross. (c) allow more efficient joint action This phase must begin at least 3 months before. 2nd phase (preparation) — Preparation preparation events — Preparation information of a general framework for the exercise.EXERCISES AND AUDITING OF EMERGENCY PLANNING 107 — Leaflets to be distributed to people. which will include on: (a) the creation of several accidents: their nature and sequence in time (b) description of each event and its placing in the categories of accidents envisaged (c) indication of the organisations which report the events and those which must receive the information. there will also be information which will not have a prescribed and available receiver — Preparation of messages on the events and of respective envelopes. the civil protection service. the aim is to make people aware of the risks and the measures without dramatising them — Installation of supplementary alerting systems — Perfecting of a system for obtaining meteorological data — Perfection of telecommunication between factories. their parents. — Performance of several sectional sub-exercises to: (a) give various sectors proper training outside the general framework of the exercise (b) avoid dramatisation of the situation by frequent tests. by the and characterisation of categories of dangerous of a sufficiently detailed ‘guide’.

These receivers will have total liberty to take their decision. evacuation of people to nearby higher points. It represents the top level and can in turn either satisfy the requests presented or not. their reactions to messages. They are free to involve their subordinates and to ask the upper echelon or others for help. evacuation of gassed people by ambulance or helicopter. according to the graduations established by the ‘guide’. it will have representatives at the various levels of performance and decision centres. given by the factory chosen. The exercise is directed by the National Civil Protection Service. prohibition of all movement.g. when they feel it opportune. it will not be involved in decisions but will only observe and remark on the behaviour of the performers. and note the correct and incorrect procedures as well as their ‘reaction time’ — Preparation and setting up of places for the performance of the exercise and the different means of communication 3rd phase (performance) — The exercise will begin at hour H of day D — It will begin with the warning. these messages must lead to the immediate reactions of the various participants to whom they are sent. They are thus under the permanent observation of members of the arbitration team. e. by means of a siren and of telephone calls to the support centres reporting the events — The centres will alert the public — They will proceed simultaneously to activate all the decision organs and the helping and assistance services Adoption of measures envisaged in the emergency plan The evolution of the exercise is ensured by the transmission of messages containing the accidents. and also create new accidents such as .108 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (a) the ‘category’ of each accident and the ‘situation’ which it may cause and which will be envisaged by the civil protection (b) the theoretical solution which must be adopted according to the principles laid down for each of these situations — Formation of an ‘arbitration team’ and training of its members. etc.

It is also envisaged that means for decontaminating and neutralising toxic products can be used.EXERCISES AND AUDITING OF EMERGENCY PLANNING 109 the temporary interruption of one or several means of communication. asking the fire services for help or using neutralising products dropped from aeroplanes. Final phase After the exercise there will be a ‘critical review’ organised by the exercise directorate and by the members of the arbitration team who were placed at the various echelons and command posts. .

The selection of the site has been made in order to point out the most important emergency features. A. in case real-time information support is not available. The site where the exercise will take place is a real one in Italian territory.POILLUCCI Directorate of Nuclear Safety and Health Protection.14 Exercise Study for an Emergency of Chemical Origin G. This particular plan has to be considered as an addendum to the Provincial Contingency Plan. The performance of the exercise will be based on the following references: (a) The Piano Provinciale delia Protezione Civile (Provincial Contingency Plan of Civil Protection) established by the competent Prefettura (the Prefecture is the provincial body of Central Government) (b) The Piano di Intervento (Particular Contingency Plan) established by the Prefecture for toxic hazards originating in the specific industrial plant involved in the simulated accident. Rome. . This tool has to enable the Emergency Coordinator to define quickly and accurately the actual areas at risk. ENEA. Italy 1 INTRODUCTION The scope of the study is the arrangement of a preventive technical tool for a specific emergency exercise following an accidental event with release of a toxic substance.MACCHI.MORICI & G. The definition of the areas is aimed to optimize the necessary emergency provisions.

Schematic topographical map of the site. . 1.EXERCISES AND AUDITING OF EMERGENCY PLANNING 111 FIG.

2. Scale layout (partial).FIG. 112 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

is widely used at international level. In particular. . Some of these cars contain hazardous materials. flammables. more details are contained in the plot plan in Fig. based on the Cox-Carpenter model. including an industrial area with chemical and manufacturing plants. A recent revision of the Probit coefficients [4] has been taken into account in the present study. the accidental sequence described in the following Section 4). These probabilities refer to population staying outdoors and in absence of escape reactions. Furthermore. considering also the short duration of the exposure. the railway traffic is based on a FFSS station adjacent to the tank farm of a chemical plant. 1. developed by Technica Int.3]. The code provides also the evaluation of the toxic effects in terms of probability of lethality at given distances by the Probit equation approach. such as chlorine. parts of the WHAZAN package. the maximum distance for a concentration of 25 ppm has been reported. Beyond this limit the impact has been regarded as negligible. expressed as the maximum downwind distances to have probabilities of lethality (LTL) of 50%.EXERCISES AND AUDITING OF EMERGENCY PLANNING 113 2 SITE DESCRIPTION The selected site is located in essentially flat territory near some small towns. The results are reported in Tables 1–4. The study supplies a parametric evaluation of the areas at risk following the dispersion of chlorine clouds originating in an instantaneous release from pressurized vessels (cf. 2. ammonia. Ltd in collaboration with the World Bank [1]. 5% and 1%. The Dense Cloud Dispersion code. For this purpose ENEA/DISP has used the codes ‘Adiabatic Expansion’ and ‘Dense Cloud Dispersion’. The movement of raw materials and finished products takes place both by railway and by road. Normally several rail cars are in transit or parked in the station. 3 PREVENTIVE EVALUATION OF THE AREAS AT RISK ON A PARAMETRIC BASIS ENEA/DISP has performed a study to be used by the Emergency Coordinator as a source term in exercise preparation and as a decision making tool during the exercise itself. The calculations have been performed for several values of the released mass and for the most likely meteorological conditions at site (weather category and wind speed). even if many others are available [2. A schematic topographical map of the site is shown in Fig.

representing the maximum impact distance due to the initial expansion of the cloud (Table 5) — The spreading angle (function of the meteorological conditions only).114 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Table 1 Distances to a given hazard level (LTL 50) for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions In order to enable a more accurate evaluation of the areas at risk. according to the following procedure: . two more parameters have been defined: — The back-distance (function of the released mass only). this parameter does not take into account the possible oscillation in the wind direction On these bases the Emergency Coordinator is able to identify the contours of the actual areas at risk on the topographical map. representing the envelope of the transverse impact distances due to the spreading of the cloud during the translation (Table 6).

3 and a possible oscillation of 40° in the wind direction. 3 shows the case relating to a release of 45 tons of chlorine in meteorological conditions D.EXERCISES AND AUDITING OF EMERGENCY PLANNING 115 Table 2 Distances to a given hazard level (LTL 5) for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions (a) Drawing of the circumference. The availability in real time of the actual meteorological conditions is a necessary premise to a correct evaluation of the . Fig. with centre in the release point and radius equal to the back-distance (Table 5) (b) Drawing of the angle at risk with vertex on the said circumference in the upwind direction and opening given by the spreading angle (Table 6) plus the possible oscillation of the wind direction (c) Drawing of the circular sectors with centre in the release point and radius equal to the impact distances (Tables 1–4) As an example.

116 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Table 3 Distances to a given hazard level (LTL 1) for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions areas at risk. it is anyway necessary to perform an estimation on-site with due caution. In particular with regard to the weather category. developed by UKAEA SRD [6]. are reported in Table 9. limited at the maximum distance to 25 ppm. the Emergency Coordinator may use some rough estimation developed for this purpose. one of which is shown in Table 7 [5]. and the code developed by the Ontario Ministry of the Environment [7]. whose availability as prompt information from meteorological services is normally poor. The results of the DENZ code are reported in Table 8. In case insufficient information is available from some meteorological services. A comparison has been carried out between some results of the WHAZAN code (Tables 1–4) and the results of two other codes for the dispersion of dense clouds. namely the DENZ code. . while those of the Canadian code.

. 25 ppm) for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions Table 5 Back-distance in relation to released mass .EXERCISES AND AUDITING OF EMERGENCY PLANNING 117 Table 4 Distances to a given hazard level (LTL negl.

118 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Table 6 Spreading angle in relation to atmospheric conditions It is clear that some discrepancies that exist in the results of the different codes cannot be overcome at the present state of the art [2. On the other hand. casually or depending on precise advice. 3]. a sure meaningful overestimation is introduced by any one of these codes. In fact a significant part of the population. not taking into account the mitigation effect due to the indoors sheltering. .

release of 45 tons of chlorine in meteorological category D. 3. Risk contours.3.EXERCISES AND AUDITING OF EMERGENCY PLANNING 119 FIG. Table 7 Flow chart for selection of weather category .

Table 8 Distances to a given hazard level for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions (as per DENZ) 120 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

illustrated in Fig. . This comparison shows not only good general agreement. 4. This perspective has to be very well borne in mind by the Emergency Coordinator in using the present tool. with the well known recommendations expressed by the US DOT (Department of Transportation) about the isolation and evacuation areas in the case of major chlorine release [13]. The quantification of this mitigation effect has been made the object of recent studies [8–10]. as already done in some particular cases [12]. and particularly to enable an optimized use of limited resources. but also the greater completeness and applicability of the present tool. for the purpose of the present study. the intermediate results of the WHAZAN code have been regarded as a ‘best estimate’. In this sense an overestimation of the distances may be negative because it will induce the available resources to spread over an area bigger than necessary.EXERCISES AND AUDITING OF EMERGENCY PLANNING 121 Table 9 Distances to 25ppm for instantaneous releases of chlorine from pressurized vessels in relation to released mass and atmospheric conditions (as per Canadian code) will stay indoors during the cloud passage. an underestimation may also be negative because it will induce neglect of some indispensable provision. On the other hand. Therefore. Further support for this selection is given by the direct comparison. As the aim of the study is to support the decision-making process in an emergency. These aspects should be taken into account by the most used codes. without significant loss of simplicity and speed of use. as well as the influence of particular characteristics of vulnerability of the population [11]. the selection of the useful results has been performed accordingly.

the possibility of getting the fire under control is remote. Among others.122 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. including the water reserves. 2). within about 4 hours. releasing to the atmosphere 45 . As thermal collapse of the chlorine rail car must be expected under these conditions. For this reason. a rail car containing chlorine is parked in the station (Fig. Within about 20 min (that means 4½ hours after the sequence initiation) the rail car collapses. 4. At this moment. before that moment all the available firefighting resources. a comparison with the US DOT. are directed exclusively to the protection of primary targets inside the plant. its protection by water sprays is attempted. the heat radiation makes it impossible to approach the rail cars and to move them to a safe place. Risk contours for a release of 45 tons of chlorine. on the other hand. 4 DESCRIPTION OF THE ACCIDENT SCENARIO The initial event of the accidental sequence is a major fire in the tank farm of the chemical plant adjacent to the railway station. The probable evolution of the situation makes a major escalation of the accident inside the plant possible.

while an Emergency Control Centre is set up at the Prefecture. with the aid of the documentation prepared by ENEA/DISP (see Section 3) and on the basis of the actual meteorological and accidental data. advises the Fire Brigade about the specifie hazardous situation existing inside the station and undertakes the necessary provisions regarding the railway traffic. and the demographic and logistic situation — Identification of the area whose population has to be advised to stay indoors — Identification of particularly vulnerable targets (schools. At this moment. The initiation of the exercise will start with the activation of the on-site emergency plan. the special procedures contained both in the Provincial Contingency Plan and in the on-site emergency plan are activated. The Fire Brigade converges on the site. due to the potential release of a toxic substance. the exercise is held according to the procedures contained in the Provincial Contingency Plan of Civil Protection of the competent Prefecture and in the on-site emergency plan of the chemical establishment involved.) inside the maximum impact area For illustrative purposes only. due to the cooling by water sprays for as long as possible. The last named. a possible line of action may be as follows: . being in the charge of the competent Local Authority. followed by the warning given to the personnel of the railway station. etc. evaluates the areas at risk. after a preliminary evaluation. hospitals. the available resources. On the whole. 5 (valid for an instantaneous release of 45 tons of chlorine in category D.3). This one. 5 REALIZATION OF THE EMERGENCY EXERCISE The study of the exercise operating details is not a part of the actual work.EXERCISES AND AUDITING OF EMERGENCY PLANNING 123 tons of chlorine. and referring to Fig. It has to be noted that no release of chlorine from the safety valves will take place up to this moment. Knowledge of these areas enables proper consideration to be given to the following aspects: — Evaluation of the possibility of a safe evacuation and its extent based on the expected accident evolution. where the Emergency Coordinator is operating.

the possibility of getting assistance in real time from the ARIES Emergency Centre of ENEA/DISP will be checked. — Immediate evacuation of the back-distance circle and the area A50’ with the exception of personnel provided with adequate protective equipment — Indoors sheltering of the population in areas A5 and Al. . The ARIES Centre is already operating at national level for nuclear emergencies [14]. Example of risk contours on a schematic topographical map. in addition to the meteorological services normally available. with priority warning action in area A5 — Warning to the hospitals and evacuation of the schools in area A0 In the course of the exercise. 5.124 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG.

provides for re-entry of the population. the necessity for specific training of key personnel will be verified. Progr. WARLEY. waiting for an adequate real-time information system to be organized and implemented on the whole national territory. Mitigation of accidental toxic gases. Manual of Industrial Hazard Assessment Techniques. for fire and explosion hazards only [15]. 2. On the other hand. Eng. as already done by the Ministry of the Interior with the SIGEM system. Oct. Evaluating emergency response models for the chemical industry. REFERENCES 1. (1987). Technical papers. A second objective of the exercise will be to check the adequacy of the technical aids made available by ENEA/DISP to the competent Prefecture: 1. Feb. . Assistance in real time by the ARIES Emergency Centre Prospectively the exercise should give the necessary indications in order to prepare a proper format for the realization of other preventive parametric evaluations. Chem. (1987)..M. Jan. Washington. The World Bank. University of Sydney. referring in particular to the clear and correct interpretation by the Emergency Coordinator 2. McNAUGHTON. 6 CONCLUSIONS The emergency exercise will give indications covering many different aspects.. 1985.J. 4. Aug. These measures should be used ad interim in dealing with real or simulated emergencies. the Emergency Coordinator reestablishes normal conditions and.G. G. P. 1986. The scope of the exercise will be first of all to test the general adequacy of the procedures included in the Provincial Contingency Plan.EXERCISES AND AUDITING OF EMERGENCY PLANNING 125 To conclude the exercise. as well as their capability to deal with the technical/scientific aspects of the problem. 3. Int. and particularly their applicability to this type of emergency. & BODNER. D. on Preventing Major Chemical Accidents. Preventive parametric evaluation of the areas at risk. Warren Centre. where evacuation has been undertaken. Major Industrial Hazards. HARRIS. C. Symp.

Il sistema informativo SIGEM: la gestione delle emergenze. 13. 7. & NUSSEY. ARIES: a computer based system for the real time monitoring of atmospheric dispersion in nuclear emergency.C. E. UKAEA-SRD. Series.126 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 5. 14.G. Antincendio. Toxic gas incidents: some important considerations for emergency planning. (1987). 11. P. FRYER.C. L.. P. 93. (1984). WITHERS. J. C. J. (1979).P. ROMANO. Isolation and evacuation distance table. 6.P. (1985). G. (1986). 12. The development of an effective emergency procedure for a toxic hazard site. J. HAASTRUP. PAPE. Risk Analysis. & PICCININI. A basic approach for the analysis of risks from major toxic hazards. . US DOT Guidebook. No. The portable computing system for use in gas emergencies. Conf. (1986). R. 5. R. IChemE. Occupational Accidents. IChE Symp. Series. IAEA.. Series. Refinement of Estimates of the Consequences of Heavy Toxic Vapour Release. Hazardous Materials. Symp. on Nuclear Power Performance and Safety. DENZ—A Computer Program for the Calculation of the Dispersion of Dense Toxic or Explosive Gases in the Atmosphere. No. Indoor fatal effects of outdoor toxic gas clouds. IChE Symp. G. 8. E. P. G. Apr.S. (1987). CARACCIOLO. R.M. Toxic Gas Risk Assessment: The Effects of being Indoors. 6 (3). March. The assessment of major hazards: the density and other characteristics of the exposed population around a hazard source.D. 9. 10. 15. July. LYNSKEY. No. Stern..J & LEES. F. N. (1985). DAVIES.I. May. 93. IChE Symp. & KAISER. A. & PURDY. & DAVIES. (1986). PURDY.. ENEA/DISP Int. P. PETTS. 94. 8 January 1986.. 14. Oct. MARCHIONNE.

There is no better way to test the availability of personnel than to undertake the physical exercise. The whole question of the need to exercise.e. of course. The County Fire Brigade was formed in 1974 as a result of local government re-organisation combining into one authority the former Teesside and Hartlepool . However. This means moving manpower and equipment to the scene and simulating an actual incident.COONEY Cleveland County Fire Brigade. However. Emergency plans of any description are worse than useless if they are prepared and then left on a shelf gathering dust. i. UK 1 INTRODUCTION The success or failure of any on/off-site emergency plan is wholly dependent upon effective communications between the public emergency services. and the industrial/commercial input concerned. discussion and consideration. The best way to ensure that close consultation and coordination take place is by exercising.15 Effective Organisation and Incident Control W. requires a great deal of thought. Considerable importance must be placed on all branches of the plan. this has industrial and commercial financial implications. the industrial complex concerned.C. bordering North Yorkshire and Durham. this can be an extremely costly and time-consuming situation. non-uniformed local government services and. uniformed emergency services. whether it be in respect of an on-site or off-site plan. the local government services in the area. The only way to test a plan is by exercising that plan to ensure that all persons concerned are aware of their duties and responsibilities. Should we exercise or not? 2 BACKGROUND Cleveland County is situated in North East England. Hartlepool.D.

128 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS County Borough Fire Brigades with parts of the North Riding of Yorkshire and Durham Fire Brigades. The present brigade is split into two divisions (Northern and Southern) which are generally divided by the River Tees. on many occasions. on closer examination. the springboard for regulatory control was the Dangerous Substances (Conveyance by Road in Road Tankers and Tank Containers) Regulations 1981. 13% of the total county area is classed as ‘A’ risk. 3 LEGISLATION The last ten years have produced a significant number of regulations which have had a large impact upon the chemical industry nationally and have. has more movements than any other port. any visitor cannot help but notice the very highly concentrated industrial areas which dominate the majority of the county area. By national standards Cleveland County is a small area covering 58 550 ha with a population of approximately 650000. been brought about to assist the Fire Service in its operations and planning. with other areas of special risk. In fact. This regulation came into being after a voluntary scheme was introduced in Cleveland in the early 1970s and a labelling system was drawn up and known as UKHIS (United Kingdom Hazard Information Warning System). This means in effect that Cleveland Fire Brigade responds to the largest concentration of chemical and petro-chemical complexes in Western Europe. In order to discharge its functions. the Brigade has 15 strategically placed fire stations. in terms of hazardous products. A panel was divised which displayed not only the warning diamond but also three other vital sources of information: (a) A simple code to give first strike information to fire service crews attending an incident involving a road tanker or tank container (b) The United Nations number which allowed further information to be sought on the product via chemical information retrieval system (c) A telephone number that could be contacted to speak direct to a source of specialist advice. However. As far as the transportation of hazardous materials is concerned. . The River Tees now rates as the third busiest port in the United Kingdom and.

there are occasions when a number of packages contained within the vehicle are in fact ruptured and the contents escape causing either a toxic hazard or a fire hazard or a combination of both. The voluntary scheme had now become mandatory. It was with this in mind that the pressure put on the Health and Safety Executive resulted in a set of regulations for the marking of . then it is not normal practice for the total load to be involved. or indeed for all the packages to be ruptured and a cocktail situation evolve. However. in March 1979. etc. From experience. glass containers. etc. It is very significant that this set of Regulations came as they did because they mark themselves as the forerunner to the very latest set of Regulations in respect of the Conveyance of Packaged Goods. carrying a mixed load being involved in a road traffic accident on a motorway.) Regulations 1986 which came into force on the 6 April 1987. Later in 1981 this regulation was enveloped into the Dangerous Substances (Conveyance by Road in Road Tankers and Tank Containers) Regulations 1981. difficulties were encountered in the form of drums. Once again this Regulation came about as a result of considerable pressure from the public emergency services to form some type of marking system for vehicles carrying dangerous packaged goods. Packaging and Labelling of Dangerous Substances Regulations 1984 (CPL) which came into full operation on 1 January 1986. Whilst the above regulations were a major breakthrough in providing the fire officer with information upon which to base his attack on an incident involving hazardous materials. These regulations. cardboard cartons. carboys. plastic containers. It is widely appreciated that a simple 3-tonne flatback lorry can carry quite a cocktail of dangerous substances in packages.EXERCISES AND AUDITING OF EMERGENCY PLANNING 129 The pilot scheme which was introduced in Cleveland attracted such favourable comment from both emergency services and industry that. The idea of the CPL regulations was to ensure that a composite label was attached to the package which offered advice and assistance to the Emergency Services when dealing with that particular substance. These Regulations are the Road Traffic (Carriage of Dangerous Substances in Packages. it has been noted that. This problem resulted in the introduction of the Classification. but imagine the 30-tonne containerised vehicle. in conjunction with a list of applicable chemicals. form the basis for legislation in the United Kingdom for the bulk transport of chemicals by road. should any of these vehicles be involved in such an accident. the Hazardous Substances (Labelling of Road Tankers) Regulations 1978 came into force.

the minimum number of plates required is two. Having dealt with regulations designed to assist the emergency services in the handling of incidents involving hazardous materials in transit. outlined in black. 4 SELF-HELP SCHEMES 4. have introduced a procedure called ‘Longstop’. all dancing’ means of solving a problem. However. as a result of that. in all cases it is not possible to have direct access to the manufacturer or his nominee. trader or supplier of the product to offer advice and/or assistance’. do is offer a standard procedure that says to fire authorities. there are other equally important systems which assist the emergency services in bringing any incident to a satisfactory conclusion: Chemsafe and Chemdata. Qne on the front of the vehicle and one on the rear. The reason for this simple marking system is so that the public emergency services will identify the vehicle as having dangerous packaged goods and then use the label on the package as defined under the CPL regulations for subsequent action in dealing with that particular problem.130 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS such vehicles. any vehicle carrying in excess of 500kg of one or more dangerous substance is required to be marked with a rectangular orange plate. Longstop is a procedure that involves the National Chemical Emergency Centre at Harwell. The Chemsafe Scheme goes from there to allow the manufacturer under the standard procedure to make contact with another company who is also a member of Chemsafe to assist on his behalf. and qualified staff with practical experience over a very wide range . for example. ‘Contact the Specialist telephone number if available and make direct request from the manufacturer.1 Chemsafe The Chemical Industry Scheme for Freight Emergencies (Chemsafe) is a voluntary scheme undertaken by the chemical industries in conjunction with the Chemical Industries Association. in fact. should the other manufacturer be closer to the scene of the incident. as a specialist telephone number may not be available and. When the regulations became law. The scheme does not give to the fire service an ‘all singing. the Chemsafe Scheme as devised by the Chemical Industries Association. What it does. which has a massive computer with databanks storing most of the known chemical products available.

and all the other senior persons on the ground. However. the officer in charge of the local police service. there should be a great deal of consultation between the officer in charge of the incident. It was also devised that the print out from the databank would be simple. 5 INCIDENT PROCEDURES There are a number of areas that require to be considered when dealing with this particular subject. should the fire service require additional advice in helping them to deal with an incident. by no means least. Evacuation must be given considerable thought with regard to the welfare needs of the people being evacuated. the protection of property left behind following the evacuation. 4. clear and easy to understand. The prime movers in an evacuation situation would normally be the police force. and simple to access computer databank on chemicals. Nevertheless. in some instances. There is a need in dealing with this type of incident to have a clearly defined role of command. primarily devised for the British fire service. The idea behind Chemdata was to produce an easy to understand. therefore. before this takes place. It is important.2 Chemdata Is a computer based databank. The databank is updated on a regular basis from the National Chemical Emergency Centre at UKAEA Harwell UK. to consider the evacuation procedure that should take place. Evacuation is a simple and straightforward word but can cause chaos and disaster if not handled correctly. incident control and communications. the Chemdata system is being used by 45 fire authorities within the United Kingdom and the current database holds approximately 50000 known chemical substances. simple to operate. that the Fire Service are fully aware of the Chemsafe procedures that are available. available to be called to the incident to offer on the spot advice. documentation and. transportation. whether it be the Chemsafe standard procedure or the Chemsafe Longstop procedure. should we evacuate? Is it necessary? Are we placing the people at risk by exposing them to unnecessary risks out of doors? This subject requires a great deal of discussion in some considerable detail to ensure that all parties involved in the control and . to be available to offer advice and.EXERCISES AND AUDITING OF EMERGENCY PLANNING 131 of chemical substances. After five years of use.

their off-site planning and it is important for the success or failure of any off-site emergency plan that effective communications between the public emergency services and industrial complexes operate. health authority. when all exercises undertaken are analysed in detail. they fit into two main categories: 6.1 Table-top exercises Exercises of this type are used to test the emergency planning procedures by using table-top plans and imaginary situations and a carefully detailed. both from the uniformed point of view and from the non uniformed viewpoint.132 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS implementation of an incident are fully aware of what evacuation means. It is equally important for the needs of other areas of local authority services. etc. the public emergency services. However. The majority of authorities have completed. written scenario. are fully in the picture with regard to the possible problems that could arise following a major incident. The methods of undertaking exercises are many fold and the size and format can vary enormously. or nearly completed. The personnel normally involved are primarily: (i) The principal management from the industry concerned (ii) Principal officers of the public emergency services. exercises need to be undertaken to ensure that site personnel. the local authority. etc. health authorities. There is no better way to test the availability of personnel than to run some form of exercise. The method normally undertaken involves a group of officers using the plans of the premises and simulated situations in a . 6 EXERCISES FOR ON/OFF-SITE EMERGENCIES The Control of Major Industrial Hazards Regulations 1984 (CIMAH) within the United Kingdom place a duty on industrialists and local authorities to prepare emergency plans in respect of major emergencies that may occur and primarily in respect of off-site emergency plans. to have very close liaison with each other and this means only one thing— exercises. In order to ensure that the plans have been drawn up to allow sufficient flexibility. local authorities.

and they should be guided towards certain objectives. It is no good having a haphazard physical exercise. type of product. water supplies. no doubt. You will note that I have used the word ‘required’ above. These exercises must be well planned in advance. By this means. for example. 6. plus a vital debriefing session. etc. the only useful method of testing a plan’s efficiency is to carry out such exercises. You will recall. inventories. This type of exercise is required to take place with site personnel and emergency services. availability of people in the right places to undertake certain tasks. that the regulations do not specifically state that exercises are required. and a whole host of various other activities that can take place on site. A table-top exercise is a useful medium to test out management reaction to receiving telephone calls and being asked specific questions in respect of. It is therefore important that major physical exercises be undertaken on a regular basis.EXERCISES AND AUDITING OF EMERGENCY PLANNING 133 written scenario to alert and actuate policies already laid down in the emergency plans.2 Physical exercises These are by far the most common type of exercise. Clear supervision and assessment. amending and alteration of the plans as required. should take place after an exercise.. A simple. maximum benefit will be gained . without massive disruptions to output. straightforward. as no benefit would be gained from haphazard results. capacities of plants. However. A table-top exercise of the type described above requires firm and close control in order to maintain an element of realism. type of release expected. Telephone contacts and the answering of specific questions and requests by management would be normally all that is required. and soon enough for the exercise still to be fresh in the minds of the people who undertook it. to test the validity of the plan. in conjunction with the local authority etc. since it is necessary on occasions actually to move manpower and equipment to test their availability and capabilities of handling an emergency situation. table-top exercise can have considerable advantages in checking for example the plans in respect of advice to public. with a carefully written scenario. This type of exercise also has the desired effect that it does not interrupt normal industrial activities on the site that are taking place and that combined on-site and off-site planning can take place at the same time.

whether it be on-site or off-site plans of any description. the Director of Education. Difficulties arise in using personnel who are not normally acquainted with a fast response time.2. simply to test a plan with a physical exercise.134 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS from the problems encountered and any mistakes that were seen. Exercises are a way to bring these officers out of their normal environment and place them in an abnormal situation to allow them to expand their horizons and expand the role to be played in an actual situation. require considerable preparation beforehand so that the plans are clearly understood by all parties before the exercise commences. Exercises. Also. is worse than useless if left on a shelf gathering dust. Also. takes place at a very early stage after the emergency plans have been drawn up. 6. The whole question of the need to exercise. This allows all those people concerned with the plan to be aware of their duties and responsibilities. for example fire appliances off normal activities. it could denude available resources away from normal activities. For example. It would be even more difficult to move people in an exercise situation. Most local government chief officers tend to deal with emergency situations on a very small scale.3 Problems associated with the physical type exercise (1) Wholesale evacuation of members of the public cannot be undertaken during a physical exercise as there is a considerable difficulty in evacuating the public in a real life situation. The only way to ensure that a plan is tested is by exercising that plan. It is important that this liaison. (3) Physical exercises undertaken by the public emergency services can only allow limited resources of manpower and equipment so as not to take away the sharp end service to the public. disruption caused to the normal way of life within that particular community would not justify involvement at that level. . no matter in which form they are undertaken. and this type of exercise. County Surveyor and Engineer would normally only deal with small scale emergency situations. Director of Social Services. (2) The actual use of large amounts of personnel has a fairly large cost implication.

Control. this type of exercise gets the dialogue running between all the parties and as the dialogue is in being. CONTROL Dominate—command—exert control over COMMUNICATIONS Practice of transmitting information There is no way that I am attempting to teach the readership English. However. For example. it ensures that consultation continues. You will see from the dictionary definition that to control means to dominate or command. it is important that the definitions of the words noted above are clearly understood. Incident control therefore means a number of things: Hazard definition and identification Hazard effect Effective control of the overall situation Communications Evacuation—yes or no? Let me now look at each of these areas in detail. Also. EXERCISES IN WHATEVER FORM ARE REQUIRED TO TEST EMERGENCY PLANS 7 INCIDENT CONTROL The Oxford English Dictionary definitions in respect of Incident. in fact. A major incident situation within the United Kingdom . due to weather conditions. one area that is not identified is the role of command. let us look at the situation with regard to Incident Control. Whilst that seems simple to state. that cross-flow of information and experience continues. it is in fact difficult to implement. Communications and Evacuation are as follows: INCIDENT Public event causing trouble. The dictionary definition states quite clearly that Incident Control is to dominate or command a public event causing trouble. then the plans will not gather dust. etc.EXERCISES AND AUDITING OF EMERGENCY PLANNING 135 Also. is moving towards an area of population. This. is the subject that is being explored during this particular conference. That release. and that officers and personnel do not meet each other for the first time in a live situation. Firstly. Let us imagine a release of a toxic substance from a particular source. but they have the opportunity in an exercise situation to see their other contemporaries at work.

In the case of an incident not involving fire. it is of paramount importance in any evacuation procedures to find out the effect that would take place on the general population. The command role of such an incident is extremely important. the officer in charge of the operations on the incident ground. i. fire was involved in no matter what form. and of course to the workers on the site. dry. It is very important to ensure that the hazard definition and identification is understood as well as possible to ensure that the local health authorities are provided with sufficient information to provide effective treatment to those persons who may be suffering from the effects of the incident. in practice. .e. the command role does not assume these defined areas. You will see from earlier parts of my paper that within the United Kingdom we have a number of areas for gaining hazard information. cold. 7. We must understand the parts per million definitions as to whether people ought to be evacuated. however. We must. is a combined effort involving all the senior officers of the public emergency services in decision making. then the Senior Police Officer present would have overall command of the situation under United Kingdom legislation. This would appear to be a very haphazard method of operation and. windy. etc. This. consider weather conditions. then the senior fire officer present would have overall command. does not mean that we operate a democracy in command. There is a final decision taken and an autocratic decision must be made. If. must have a clear indication of the hazard. however.2 Hazard effect Whilst it is important to find out the characteristics of the product that is being dealt with. the command function under normal circumstances. that is whether it is wet. of course. be it whatever type of incident.136 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS brings into play all the major emergency services—fire. 7. The hazard effect has a very important part to play in the overall command role. Because of pre-planning and exercises. the definition of the hazard.1 Hazard definition and identification It is extremely important that the senior officer in charge at the ‘sharp end’. left in premises. police and ambulance. etc. and the identification of the problems that that hazard could cause to the population.

g. But it is equally important that there is a continual feed back through the various incident controls. as we are all aware.EXERCISES AND AUDITING OF EMERGENCY PLANNING 137 7. This. in fact.4 Communications The dictionary definition of communications is ‘Practice of transmitting information’ which simply states that we should effectively communicate with each other. i. There is a need to ensure that all messages and orders issued are clearly understood by all persons concerned and acted upon as quickly as possible. in the effect of the release of a toxic substance.3 Effective control of the situation Let me now look at the problems that face a major emergency service. is by no means all the information required by the senior officer present. or by physically sending in Emergency personnel to knock on individual doors? Whilst the methods outlined above are fairly simple. can we be sure that everybody in the possibly affected area has heard the messages or have we created a panic situation in areas outside that designated for possible evacuation? Or. is it unsafe to send officers into a possible affected area to carry out such an evacuation? Could we be placing emergency service personnel in a dangerous situation? Allow me to go back to the question of evacuation later. however. public announcement on radio and television stations. due to a toxic release. communications play an extremely vital role in dealing with any incident and are of paramount importance when dealing with a major incident involving members of the public. albeit slow and time consuming. There are a number of other factors that come into play: Do we tell the public? If so. by what method? Public address from Police vehicles. One of the disadvantages of personal contact is the fact that if you require to evacuate. Let us assume that we have received all the information outlined above which clearly indicates that there could be a serious effect on the population. and therefore to the incident commander. leads to further questions as to Why? What for? Have you spoken to…? etc. a county fire brigade.e. 7. e. to ensure that he is aware that his . However. is it safe to do so or. is the best method of removing people from their homes. as mentioned above. All of this is extremely time consuming and has to be taken into serious consideration. simply to knock on the door and ask them to leave? This. however.

Let me go back to the problem.’ May I continue to say to evacuate or not is also the question? Let us look at the situation that we have been considering before —the release of a toxic substance. be detrimental in that radio interference can occur and cause a number of problems in the transmitting of information. on the face of it. do a number of the major industrial sites within the United Kingdom. if it is not controlled correctly. The question really is ‘Do we evacuate or do we not?’ We are very close to quoting Shakespeare in the act where Hamlet states To be or not to be? that is the question. On the other hand. One of the pitfalls with regard to communications is that a number of the emergency services operate on different radio channels and so incidentally. appears to be simple and easy to undertake but. shops. do we take the decision before it happens to move the population within the area and away to a place of safety? If the explosion or release is imminent. be placing the population at risk by taking them into the open air and. if a release and/or explosion is imminent. offices.5 Evacuation Let me once again quote you the Oxford English Dictionary definition of this simple word: EVACUATION Withdraw from—remove occupants from—place to be considered dangerous to a safer place. messages. 7. To set up communications in post can. would we not. then of course consideration must be taken for the movement of the population.138 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS communications. This is a very simple and straightforward definition which. etc. should the inevitable . therefore. have been carried out. orders. If that release is as a result of an explosion. what appears to be simple can very rapidly turn into chaos and disaster. to some extent. then the moving into that area of emergency personnel to ask the population to move places a number of persons in jeopardy. instructions. if the release is of a prolonged nature. and to allow the normal weather conditions to dissipate the release as quickly as possible and to consider the fact that the public are safer where they are than to move them into the open air. If we have an area that has been affected by a toxic release for a considerable period of time. etc. then it may well be safer to leave the members of the public in their own homes. On the other hand.

Are they fully equipped to take care of the injuries occurring as a direct result of the incident. . additional stocks of equipment and fuel. that serious consideration must be undertaken as to whether to evacuate or not. clogging main roads. Are we prepared for the media? All forms of the media will descend on the incident seeking news. shock. it is a mass evacuation. there are numerous other examples. attempting to gain information from whatever source is available. injuries to firefighters.? (d) Hospitals. Water supplies. to undertake documentation.EXERCISES AND AUDITING OF EMERGENCY PLANNING 139 happen. etc. I do not envisage tens of thousands—it could be two or three hundred—but nevertheless. police personnel etc. strain. expose them to even greater dangers? Even before you reach any of the decisions outlined above. by the word ‘mass’.? (e) Media. relief personnel. etc. involved in mass evacuation and. shelter. as officers controlling the incident. IN MY OPINION. We need to consider. Are there sufficient members of the Police Force available to control traffic. We must be sure that relations with the press are adequate to fulfil the situation. feeding personnel engaged at the incident. EVACUATION SHOULD BE CONSIDERED TO BE THE LAST RESORT WHEN ALL OTHER METHODS OF CONTROLLING A HAZARD TO THE PUBLIC HAVE FAILED. i. etc.e. and are they also prepared to undertake normal medical requirements for the general public who will suffer as a result of evacuation—old persons. Have they manned rest centres? Can they provide sufficient food and shelter? Can they. Can we provide large numbers of public vehicles to transport people to rest centres. You will see. in fact. If the decision is made to evacuate. etc. food centres..? (c) Police. to control looters. there are a number of other factors that come into play. I have outlined only a few of the problems that have to be faced when considering evacuation. informing relatives of those persons evacuated. heart attacks. are the following factors available: (a) Local authorities. therefore. ensure that the people moved from their normal residence can be adequately looked after? (b) Public transport companies.

Its major advantage is that it allows the evaluation of the effectiveness of all the functional components (i. Another approach is that of functional analysis.) these are not necessarily all utilised effectively in an emergency situation.RAMSAY Technica Ltd. The high-fidelity characteristic is important because factors such as time pressure. UK 1 INTRODUCTION In Technica’s experience. Each organisation may well be proficient in its own function. However. By focusing first on the quality factor during the auditing of an emergency plan. One approach to auditing the response capability of an emergency plan is that of high-fidelity simulation. and their suitability for the emergency environment.e. control room hardware.G. etc. this experience may have been obtained in isolation. By high-fidelity it is meant that the emergency can be simulated to match the real situation as closely as possible. London. physical obstruction or danger will significantly influence performance. It is often the case that while emergency plans focus mainly on the capacity of response resources (i. The notion of quality in this context refers both to the quality of the individual component resources and to the dynamic relationships and functions within the emergency plan as a whole. The relationships may include those .MAX-LINO.HARRISON & C. numbers of first-aiders available. emergency organisations) of an emergency plan. having had some training and experience of the relevant requirements. P. an assessment of the efficiency and of the quantity of the resources required can be determined. addressing the response capability of emergency plans involves examining both the quality and the quantity of the response resources.16 Assessing the Response Capability and Vulnerability of an Emergency Plan: Some Important Issues R.e. quantities of fire hydrants. This is an analytical technique for assessing the relationships between the components of any system.

and how these should be effectively coordinated. it was not clear to the on-site officials which of two emergency control rooms they should be communicating with. it will be assumed that the emergency has originated at an industrial site which has its own operational management but can call on regional emergency services for support. It may also be the case that in-house emergency response officials need a clearer definition of their individual roles and responsibilities. They will be discussed in this paper in the light of Technica’s experience in the assessment of emergency plans.EXERCISES AND AUDITING OF EMERGENCY PLANNING 141 between equipment. both on-site and off-site. This paper aims to illustrate how a simulation of the emergency plan. allows an evaluation of a variety of issues. The plan should define clearly the role of the site management with respect to the emergency response professionals. and physical interactions between personnel in the sharing and coordinating of resources. The main advantage of functional analysis is that it can be used to assess on paper the logical flow of events. workload distribution. Auditing the roles can be conducted on paper by documenting the tasks/activities required of the key personnel. coupled with a functional analysis of the data obtained. communications. Equally important is the knowledge of how these two distinct groups must be able to coordinate effectively to conduct their responsibilities successfully. For the purposes of this paper. the coordination of events in time. some tasks were duplicated by each control room. and man-machine interactions. Attempts by the management to take active control could lead to conflicts between themselves and the emergency response teams. As a result. To illustrate from one case study. The assessment team can determine: . 2 ASSIGNMENT OF RESPONSIBILITY This includes the roles and dynamics of the organisations which must be coordinated. and some omissions of functions were made. These include: — Assignment of responsibility — Notification methods and procedures — Communication — Public information These issues are not mutually exclusive.

the following example may be quoted. It is true. in the event of an incident which is likely to escalate beyond the site. An emergency response control room operator became overloaded with communications duties within a few minutes of the commencement of an exercise. logically. that some conflicting activities may not be filtered out of the plan during a functional analysis because there is no access to realtime data. early detection of this likelihood and early notification of the appropriate officials are important for a variety of reasons. It was found that this operator was using two telephones (one at each ear)—trying to brief the Chief Fire Officer and trying to handle several simultaneous communications on the radio. This operator was also required to brief several incoming personnel of the situation. Significant in this issue is the determination of when an emergency situation requires such notification. yet there was no assistance or relief provided for the first 36 minutes of the simulated emergency.142 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — the tasks that can be performed simultaneously without conflict. The functional analysis will then determine the appropriate resources to reduce the possibility of such workload excesses. the potential need to coordinate extra resources beyond the back-up teams already considered. To illustrate the type of operational problems that may be inherent in an emergency plan. In this way those tasks and resources/roles which need to be further developed/altered for the benefit of the emergency plan can be determined. This is where a high-fidelity simulation can be most effectively utilised. These include the reduction in time for determining what resources are needed and. This example illustrates the necessity for the emergency response plan to avoid workload excesses on emergency response officials. however. — the tasks that must be done sequentially. That is. A high-fidelity simulation can then be conducted to ensure operational viability. Another issue to address is that the methods by which early warning and clear instructions are given to the population at risk . 3 NOTIFICATION METHODS AND PROCEDURES It is important for the on-site emergency response officials to be aware of requirements procedures for the notification of government officials.

— notify the relevant emergency officials accordingly. This may be done by considering credible scenarios. In Technical experience. the warning system—auditory alarm—did not notify some of those in potential danger. there appears to be some reluctance on the part of notification officials to warn the population at risk. These include the warning . On the other hand. by making conversation difficult. during the emergency. In other words. to take advantage of the available time from the onset of the incident. in order to determine the potential hazard to the population. Previous work on the evacuation of the public has also identified some of the significant factors in the notification process which will affect population evacuation. determining the shortest time from the onset of a threat to reaching a given population at risk. for fear that this will cause panic. it inhibited the performance of emergency personnel. as far as possible. especially where the release may render sheltering ineffective. — notify the population at risk. A simulation would then provide data to help determine whether the available time allows you to to: — assess. This latter point is discussed further in the section on communication. do officials ensure that the population is given early warning such that any necessary evacuations have the required time for safe and effective conduct? The warning system should be able. Another example is that some on-site personnel during a simulation were unaware than an emergency situation did exist.EXERCISES AND AUDITING OF EMERGENCY PLANNING 143 should be determined. and estimates of the times to reach the population at risk can then be evaluated. Where sheltering is not the appropriate response to a release. to simulate release scenarios by computer modelling techniques. The time taken to complete the above three tasks should ideally be much less than the time taken for the release to reach the population. It is possible. How. however. No high-fidelity simulation has been conducted by Technica which has involved the issuing of orders to the general population for evacuation or sheltering. a functional analysis of the potential for early warning and of the instructions to the population at risk can be conducted. for instance. for instance. the risk to the population. the time taken to evacuate the population must be included in the analysis. Within the approach to auditing the emergency plan.

This information is particularly important for systems in which a queueing system is employed for simultaneous communications to the control room. By this we mean information that must be known as early as possible.144 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS content and the source of the warnings. 4 COMMUNICATION Inherent in previous examples have been some of the effects of communication system and information transfer failures. This information is rather difficult to determine using a functional analysis alone. such as from escaping high-pressure gases. This may result in trivial information delaying critical information. the event itself may be associated with extreme noise levels. where the emergency response operator is described. and which must be transferred to the relevant personnel for action. the effects of external sirens in the vicinity of emergency officials (such as . The identification of critical information can then be conducted within the framework of a functional analysis and. Specifically. another issue is that of the failure to identify/ appreciate what constitutes time-critical information. simply because it is not easy to predict the dynamics of a given incident and its effects on the emergency response personnel. but still on-site. the appropriate factors can be incorporated into the training of the officials responsible for information transfer. In the case of major industrial accidents. Outside of the control room. we refer to situations in which numerous auditory alarms are going off in an emergency control room. and techniques for avoiding this include specific training for those transmitting information and/or separate channels for critical data. once specified. fires or mechanical impacts. These may detrimentally affect the identification of critical information and its subsequent communication beyond the control room. the issue of information transfer and the overload that occurred is well illustrated. and these scenarios may be simulated to provide data on the types of critical information required for the resolution of these emergencies. One other problem within the communication issue is that of the masking effects of background noise on the performance of the operators who are transferring information. Some discussion of these will be conducted in the section on public information. Some assumptions about the emergency scenarios can be made. However. In the first example.

EXERCISES AND AUDITING OF EMERGENCY PLANNING 145

firemen) may also produce masking effects on communication
between them.
In the above situations, it is inadvisable to allow alarms to be
switched off by the incident control room operators because, as
experience shows, this can become a regular reflex and so
undermines the effectiveness of the alarm. The resolution of this
problem will depend very much on the context in which the
auditory alarm is sounded. Whilst the operators in the indicent
control room may wish to silence an alarm to which they have
already responded, this possibility should not be applied in
blanket terms to other locations.
One suggestion from a case study was that, where auditory
alarms were being sounded, the emergency officials could be
allowed to make announcements over the tannoy system. To
ensure that these announcements were audible, the sirens were
attenuated for the duration of the announcement by the official
transmitting the message. Additionally, redundancy and diversity
should be incorporated into the alarm systems such that, where
one mode fails, the other may still be available and functioning.
One solution would be to use visual alarms as well.
5
PUBLIC INFORMATION
It is possible that, whilst an effective emergency plan has been
developed for an on-site emergency, it fails to cope with the many
variables that are evident in the off-site situation. Of most
relevance to this paper is the capability of the emergency plan to
utilise effective methods by which information to the public can be
disseminated.
In Technica’s experience, the requirement of the public for
information is straightforward: unambiguity about what the threat
characteristics are, instructions on how to mitigate these, and the
ability to confirm the threat. It is simply inadequate to expect
sirens to provide the public with such specific information on the
nature of an emergency.
Responsibility for the officials/authority giving such information
must be clearly defined. Research conducted by Technica suggests
that careful consideration needs to be given to those responsible
for information dissemination and to the type of information that
will be given. In other words, a good authority figure, clear
instructions, and preferably two-way communication are
necessary. This would enhance the public’s perception of the
origin and reliability of the information. It is common for the

146 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

public to jam communication lines as they try to confirm the
threat. This should be avoided if possible.
For obvious reasons it is unlikely that a simulation exercise
involving large numbers of the public can be economically
conducted. However, data from a simulation exercise (e.g. threat
characteristics and their relation to the proposed warning
methods and content) could be used to assess the capability of the
warning system to effect the appropriate response in
the population at risk. Data on the efficiency of the warning
process in evoking the appropriate response have been obtained
by Technica, and these can be used in a predictive analysis of the
emergency plan.
6
SUMMARY
This paper has not attempted to discuss all the issues important
to emergency plan assessment. It has, however, documented the
issues that Technica has found, in previous work, to be
important. Additionally, no attempt has been made in this paper
to produce a complete methodology for assessing emergency plans.
The paper does, however, document the use of two approaches.
The first is a high-fidelity simulation, of which we have experience
in implementation for auditing purposes. The second, functional
analysis, is a complement to the first approach, but can be most
useful when a full-scale high-fidelity simulation is not viable. Such
an analysis is more economical than a high-fidelity simulation.
However, the functional analysis, on paper at least, may not have
access to real-time data. We would therefore suggest that the two
approaches be used together. In this way, the benefits of each can
be reaped to the advantage of the emergency plan.

17
Exercises and Auditing: Experience
Gained in the FRG
STEPHAN NEUHOFF
Berufsfeuerwehr Köln/Cologne Fire Brigade, Cologne,
FRG

1
DISASTER RISKS AND PREVENTION
In North-Rhine-Westphalia every city must carry out a disaster
prevention exercise at least twice a year. These exercises must be
directed at specifie local risk factors.
Cologne is a city with an important chemical industry as well as
being an important traffic junction. 25% of the total German
production of chemical materials is produced in Cologne and the
surrounding areas. The city is surrounded by a belt of large
chemical factories as well as many small businesses which
process chemicals or are involved in chemical trade and
transport. Large amounts of dangerous chemicals are transported
by road, rail, ship or pipeline. Cologne is a major road, rail and air
transport junction not only for freight but also for passengers. The
exercises must therefore be aimed at handling these two risk
factors.
In Cologne the city fire brigade is responsible not only for firefighting and emergency services but also for disaster prevention.
As long as a fire or accident can be controlled by the professional
fire brigade, by the volunteer fire brigade and by organizations
involved in emergency services, such as the German Red Cross,
the operation is directed by the chief of the professional fire
brigade. If the situation cannot be brought under control by these
organizations, then units of the Disaster Prevention Service must
be brought in, and a Disaster Alarm is given. The operation is then
directed by the town clerk. An operations staff group, the Disaster
Prevention Management (Katastrophenschutzleitung, KSL), is
available in a specially equipped control centre. This operations
staff group consists of officers of the professional fire brigade
together with the directors of various city offices such as the
Health Office or the Press Office. It also includes members of state

148 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

agencies such as the police and military services, and (if
necessary) members of private companies.
Depending on the situation at the site of the emergency, one or
several Technical Operations Management Groups (Technische
Einsatzleitung, TEL) are established. They consist of an officer of
the professional fire brigade together with a small operations staff.
The TELs direct the units which have been put under their
command.
2
THE AIM AND THE EXECUTION OF EXERCISES
The exercises should enable a realistic simulation of a disaster
and also the testing of disaster prevention measures. They should
include the following procedures:
1. Notification of the management and units of the Disaster
Prevention Service and other agencies
2. Communications between the KSLs, the subordinate TELs,
the supervisory authority and other state agencies and (if
necessary) private companies
3. Situation assessment and decision making by the operations
staff
4. The execution of measures such as warning, evacuating and
assisting large numbers of injured people
Three types of exercise can be performed depending on the specific
aim: alarm exercise, staff exercise, or a complete exercise. Any
combination of these exercise types is also possible. The alarm
exercise is only performed in order to check the time between the
alarm and the ‘ready for action’ state. A staff exercise is only
performed by the operations staff; the units at the disaster locality
are simulated. During a complete exercise the KSLs, TELs and all
units perform the exercise in a simulated emergency situation.
The most frequent exercises are the staff exercises. Alarm
exercises for the 4500 assistants of the Disaster Prevention
Service in Cologne only take place every 3 years. Complete
exercises are only performed approximately every 5 years due to
the extensive preparations which are necessary.

EXERCISES AND AUDITING OF EMERGENCY PLANNING 149

3
ALARM EXERCISES
The most recent unexpected alarm exercise in Cologne showed
some rather sobering results. The exercise was performed on a
weekday at 18.00 hours (6 pm). The members of the Disaster
Prevention Management (KSL) were notified by telephone or, if
available, by radio receivers. In cases where a simultaneous
notification of all radio receivers could take place using a collect
call (i.e. just by pressing a single button) the staff members
arrived at their positions after approximately 30 minutes. Giving
the alarm by telephone required much more time: an average of 3
minutes for each call. In addition, the telephone lines at the fire
brigade control centre were blocked, and staff members whose
presence would have been essential, especially during the early
stages of a disaster, were not available. This problem will be
tackled by the installation of a computer which automatically dials
the stored telephone numbers and plays a prerecorded tape, after
which the staff members who have been notified are registered
and a list is printed. The fact remains, though, that at least 1
hour is required before the management staff members can
commence their work. This means that in the meantime the
situation must be managed by leading staff members of the
professional fire brigade, the emergency services and the police
who are on duty at the time of the incident.
The units of the Disaster Prevention Service were alerted by
sirens and by telephone calls. In the units which were notified by
sirens, 20% of the members were available after the first halfhour, 40% after 1 hour, and 60% after 2–5 hours. In the units
which were notified by telephone calls, approximately 10% arrived
during each half-hour, resulting in a total of 46% after 2–5 hours.
The conclusions are that notification of leading staff members
and units of the Disaster Prevention Service must be done by
sirens or by using a collect call for radio receivers. Plans must also
be made for employing units with only 50% of their personnel.
4
COMMUNICATION PROBLEMS
All technical means must be employed together to ensure efficient
communications between the Technical Operations Management
(TEL) at the site of the emergency and the Disaster Prevention
Management (KSL).

150 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

The radio communications network is often overloaded or even
jammed during the early stages of an emergency. It is also
frequently used by reporters. Therefore communications using
mobile telephones in vehicles should be established as soon as
possible. This can then be replaced by connecting the Technical
Operations Management (TEL) with existing private telephone
lines. A prerequisite for the use of telephone communications is
the availability of secret telephone numbers for the Disaster
Prevention Management (KSL) which cannot be blocked by calls
from the public, from town councillors or from the press. Longer
written reports can be transmitted by radio telex, and plans and
maps can be forwarded by using telefax machines.
Not only during exercises but also during real emergencies does
it take a long time for the Technical Operations Management (TEL)
at the emergency location to forward a first detailed report of the
situation to the Disaster Prevention Management (KSL). As the
Technical Operations Management (TEL) is nearly always
overloaded, especially during the early stages of an emergency, it
can take 45 minutes or more. The Disaster Prevention
Management (KSL) must therefore have their own scouts, who can
use motorcycles or 4WD vehicles to reach the emergency area and
to report from there. The use of video cameras has turned out to be
a successful measure in Cologne. The city of Bonn even has
facilities which allow the direct transmission of video recordings
from a helicopter into the control centre of the Disaster Prevention
Management (KSL).
The long time needed by the Technical Operations Management
to investigate and report the situation has an additional
consequence: it must not affect protection measures for civilians,
for example if an accident occurs in a chemical factory! In Cologne,
chemical factories must therefore report each incident using the
code numbers D1-D4 depending on the suspected amount of
danger to the public. If the fire brigade is notified of a chemical
accident with the code number D3, the endangered area is
immediately warned by sirens, radio stations and vehicles with
loudspeakers, and closed off by the police.
5
ASSESSING THE SITUATION AND MAKING
DECISIONS
We have noticed, during all staff exercises, that the management
groups can only work efficiently if they are kept as small as
possible. The positions and tasks within the group must be

EXERCISES AND AUDITING OF EMERGENCY PLANNING 151

defined beforehand, so that the group members can prepare
themselves and practise properly. Every position must be occupied
by two people in order to enable shift work during longer
emergencies.
The performance of a management group depends very much on
whether the individual group members are capable of working
together. The most common problem is the incorrect behaviour of
superiors at the highest level. These problems increase in
significance if the exercise is being prepared, supervised and
evaluated by staff members of this superior. Therefore exercises in
Cologne are regularly planned and supervised by groups from
other cities or by disaster prevention schools. This has resulted in
a less restricted viewpoint and a more critical approach. It also
reduces the probability that information about an exercise is
released too early.
So far, in Cologne, 4 staff exercises have been undertaken
together with chemical factories. Cooperation was good as staff
members were exchanged with members of the factory
management as liaison officers. The emergency situation during
the exercise consisted of a transport accident inside or in the
direct vicinity of the factory area. The factories were not prepared
to simulate an accident in a production area or at a storage area.
Possibly they did not want to discuss the number of fatalities or
injured persons which could result from an accident in these
areas.
Two of these factories are directly on the boundary between
Cologne and neighbouring city areas. Exercises and real accidents
have indicated the problems which can occur during the required
cooperation. Planning and protection for the entire factory area
must be the responsibility of the city which has the larger
potential and facilities for protection. Tasks, information exchange
and responsibilities must be defined precisely in advance.
6
EXECUTION OF PROTECTION MEASURES
The protection measures which are required during a chemical
accident, such as
— warning,
— closing off areas,
— investigation,
— care of injured persons,
— evacuation, or

however. radio broadcasts and loudspeaker vehicles is not known. in which a fire threatens a filled LPG . The second problem area was the hospitals. were equipped with loudspeakers. Use of loudspeaker-equipped vehicles for alarms has been extensively tested in the Cologne area and the results have been used as a base for planning purposes. Unfortunately it has not yet been possible to test the treatment of a large number of poison cases in an exercise in order to gain some knowledge of the requirements for an antidote storage system.152 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — decontamination. depending on their size. The handling of large numbers of injured persons was practised in Cologne by assuming the case of a crash-landing of a JumboJet At the simulated crash site the main problem was the management of large numbers of doctors and ambulance personnel in order to ensure rapid examination and treatment depending on the gravity of the sustained injuries. past accidents should be evaluated and future hypothetical accidents should be simulated. In order to do this. The test case is a railway accident. or they can be planned and then actually executed by units. the position of which depends on the prevailing wind direction but not on the type and amount of toxic material released or on other prevailing weather conditions. even had problems when confronted with two badly injured patients at the same time. Other cities use several templates or simple empirical formulae for a calculation. cassette tape decks and prerecorded cassette tapes. It is therefore necessary to install systems which can rapidly give the alarm to additional operating teams. or complete preplanned measures can be tested by the units. or which can be employed immediately. National or EC recommendations are necessary. At night and on weekends they are understaffed and. A staff exercise is to be undertaken in Cologne in 1988. Eighty-four fire brigade vehicles which are either permanently occupied. can be planned during an exercise by the management group without being executed. The success of a combined alarm using sirens. This exercise would require a large amount of participation from the general public. as it has not yet been tested. In Cologne the endangered area is defined by superimposing a template or pattern on a map of the city. The storage of antidotes depends less on the number of possibly poisoned persons than on the time available for the administration of the antidotes and the number of available doctors.

The planning of an evacuation on this scale. which is extremely complex and requires a large number of individual measures. The participation of hundreds of people can be organized. can be practised by the operations staff. but the involvement of thousands exceeds our present organizing capacities. but this would reach the limits of this kind of exercise. According to a state recommendation. As the incident in Mississauga is difficult to compare with the situation in a large European city.EXERCISES AND AUDITING OF EMERGENCY PLANNING 153 tank car. the area within a 1 km radius of the accident site is to be evacuated. it would only be possible to indicate the problems involved in evacuating 10000 or more people by a practical exercise. In the Cologne city area this would require the evacuation of between 10000 and 100000 inhabitants. .

18 Auditing and Exercising of Emergency Plans for the Danish Oil and Natural Gas Transmission System. The systems. Denmark 1 INTRODUCTION After the discovery of oil and natural gas fields in the Danish part of the North Sea. today include 3 fixed installations covered by Article 5 of the EEC Major Hazard Directive 82/501/EEC. The specific requirements for the internal emergency plan are part of the safety requirements laid down by the Ministry of Labour in Order 406/1979 regarding the safety of natural gas facilities. the concession for import. the US ASME Code for Natural Gas Transmission and Distribution systems was used. an oil storage facility. of which the first part was put into service in 1983. oil and gas transmission and storage systems including fixed installations have been built. Based on this concession. sale. transmission and storage of natural gas was given to Dansk Olie & Naturgas A/ S (DONG) in 1979. Including Fixed Installations HANS HAGEN & PETER JOHANSEN Danish National Fire Service. 2 NATIONAL REGULATIONS COVERING OIL AND NATURAL GAS TRANSMISSION SYSTEMS The above-mentioned systems were established in accordance with the Acts giving DONG the concession. As safety code. This order also requires that the operator must obtain construction as well as operation . supplemented with specific Danish requirements. The safety regulations for the systems were established jointly with the Natural Gas Coordinating Committee headed by the Energy Agency comprising all planning and safety authorities. This concession was later extended to include crude oil. Copenhagen. and a natural gas cavern storage facility. The facilities are a gas treatment plant.

As an example of the EEP. which has approximately 20000 inhabitants and a fire brigade consisting of 2 first attendances. alarm plans. The emergency plan also includes the necessary staff training and exercises. the technical procedures. A condition for such authorization is that the internal emergency plan has been prepared and implemented. Besides the fire brigades mentioned. This means that the authorities in approving the project have to take this into consideration and if necessary set up requirements for the company’s own fire brigade and/or fixed installations for fire-fighting. As regards the external emergency plan. the emergency control centres. Within a short range of the plant there are two small communities. The emergency plan consists of a plan covering all DONG oil and gas systems together with specific plans for individual facilities. The EEP is prepared by the local fire authorities in coordination with the police and sent to the Danish National Fire Inspectorate (DNFI) for approval. This Ministry lays down the general requirements whereas the implementation rests with the police. DONG has prepared an emergency plan covering the organizational structure. The exercises specified in the emergency plans include local exercises and exercises involving the whole system and the external emergency services. The nearest big town is Esbjerg with approximately 80 000 inhabitants and a fire brigade consisting of 3 first attendances. and the equipment to minimize the consequences of accidents. 3 INTERNAL EMERGENCY PLAN PREPARED BY DONG In accordance with the requirements of the order by the Ministry of Labour. each with 1 first attendance. the Civil Defence Corps and the hospitals. the fire authorities. and liaison with the external emergency authorities. 4 EXTERNAL EMERGENCY PLAN (EEP) The main principle in making the EEP in Denmark is that the fire brigades available in the area in which the plant is situated must be adequate. the communication systems. the treatment plant for natural gas in Nybro will be considered. the Civil Defence Corps send the following appliances to the plant in a case . The plant is located in the community of Varde. the responsibility rests with the Ministry of Justice.EXERCISES AND AUDITING OF EMERGENCY PLANNING 155 authorization.

the facilities. and within 50 minutes a further 36 firemen and 8 vehicles. with the following objectives: (a) Testing the alarm procedures (b) Testing the collaboration between DONG and the external emergency services and the availability of special equipment (c) Making on-site and off-site personnel familiar with their equipment. These reports normally cover the following items: (a) Tasks carried out during the exercise (b) Communications with other participants and the press (c) Lessons learnt . and possible accidents 6 AUDITING OF THE EMERGENCY PLAN After each exercise a short report is made by the participating DONG units and the external emergency services. 5 PLANNING OF EXERCISES The exercises are planned by DONG in collaboration with the external emergency services.156 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS of ‘great alarm’: within 15–20 minutes there will be 31 firemen and 9 vehicles.

VERSTEEG Ministry of Housing. FRG .SCHNADT TÜV Rhineland. Physical Planning and The Environment. The Netherlands Rapporteur: H.F.SESSION IV Techniques for Emergency Plans Chairman: M.

In order to accomplish its objectives. Barcelona.SIGAÉLSb aDepartamento de Ingeniria Nuclear. Barcelona. TIGRE uses mathematical models to calculate the values of some variables representative of the physical phenomena due to major accidents. The second data base contains the information needed to estimate physical and chemical properties of hazardous substances. Universitat Politécnica de Catalunya.TRUJILLO BRAIN Ingenieros SA. In this way it is possible to alter the meteorological conditions and also the source term adopted in the Response Guide. In addition. So the program TIGRE allows the operator to determine the real consequences of the accident and also the best counter-measures to be taken. Spain The computer code TIGRE [1] is an expert system. Spain & A. These two provinces contain approximately 70% of the Spanish chemical industry. b Departamento de Termotecnia.a B. The models used are derived from those proposed by the Netherlands Organization for Applied Scientific Research (TNO) and described in the so-called Yellow Book [3]. It contains a complete catalogue of the accidents taken into account in the Response Guide [2]. It has been applied to the plans of the chemical complexes of the provinces of Tarragona and Huelva (Spain). known respectively as PEE/ PLASEQTA and PEQHU. The first deals with major accidents. specially designed to fill the needs of off-site Emergency Plans prepared under the requirements of the 82/501/CEE directive. . TIGRE contains two data bases.SENYÉ.19 The Computer Program TIGRE and its Application to the Planning of Chemical Emergencies A.

The simplicity of TIGRE is improved by a very detailed user’s manual and by short courses given to the personnel responsible for its use. as also does the code TIGRE. it is also possible to use TIGRE to predict its consequences. When the Chemical Security Group has assessed the accident. Finally. It has been deduced from a meticulous analysis of the installations. TIGRE stands for Tratamiento de la Information y Guía de Respuesta en la Emergencia (information treatment and response guide during the emergency). (2) If the accident is not similar to one of those contained in the Response Guide. 2. 3. NOTES 1. If the substance involved in the accident is not one on the data base [6] it is also necessary to supply properties data. At this moment the Emergency Plan becomes active. In addition. They also have to be complete enough to provide reliable results. Due to that lack of precision they have to be supported by the Chemical Security Group [4]. The Response Guide contains a great number of hypothetical major accidents covering a wider range of severity. it contains a great number of safeguards to minimize erroneous operation. it is possible to obtain a forecast of the consequences of the accident in a few seconds. The sole difference is that the operator has to provide some more information about the characteristics of the accident and the place where it has taken place. The option chosen is a compromise between the two needs. Methods for the Calculation of the Physical Effects of the Escape of Dangerous Material (Liquids and Gases). The only additional information needed is meteorological data.TECHNIQUES FOR EMERGENCY PLANS 159 It is worth noting that the models used have to be simple enough to be capable of offering responses during the emergency. Report of the Committee for the Prevention of Disasters. one can use TIGRE in two different ways: (1) If the accident is similar to one of those contained in the Response Guide. emergency simulacra are prepared periodically so as to maintain the operability of the whole system. Directorate General of . The emergency starts when a chemical factory notifies the CECOP [5] of an incident. The use of TIGRE has been simplified as much as possible in order to make it suitable for use during emergencies by unqualified personnel. TNO.

It is an operations centre from where the emergency procedures are directed. Labour. The substances data base contains more than 60 substances. The Netherlands. Ministry of Social Affairs. 6. Among them are the most commonly used of the classified substances under the 82/501/CEE directive. Voorburg. . 1979. CECOP stands for CEntro de Coordinación OPerativa. 5.160 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 4. The Chemical Security Group is one of the four task groups that are included in the Emergency Plan. It has to assess the accident and provide all the information needed to the Direction of the Plan.

We will then provide a conceptual framework for decision making under crisis. AI has its roots in the mathematical logic systems of Frege. Troy. Whitehead. These theorists addressed thinking by formalizing some aspects of reasoning into a relatively simple framework. 2 EXPERT SYSTEMS [2] The origins of artificial intelligence (AI) date back several decades. The crucial advance was the recognition that computers were not limited to numeric calculations but could process symbols. Russell and Tarski. The formal systems of logic and newborn computers were then linked by these systems of logic. New York. We will first review a relatively new facet of computer technology — expert systems. during. Albany. We follow with a discussion of a prototype expert system for response to an accident at a nuclear power generation facility. among others.WALLACE Rensselaer Polytechnic Institute. USA & WILLIAM A. USA 1 INTRODUCTION The capabilities for computer technologies to provide decision support in emergency response are now well recognized [1]. and after potentially catastrophic events must be managed in order to have effective response. . a situation typified by emergency response. The information flow prior to.20 Expert System Technology to Support Emergency Response: Its Prospects and Limitations SALVATORE BELARDO State University of New York. Our final section discusses the potential advantages and limitations of expert system technology in emergency response. and in the theories of computation developed by Church and Turing. New York. We feel strongly that computer technology can be a crucial component in this management process.

much as he or she would approach a human expert on the same problem. The user explains his or her problem. Expert systems have recently emerged as the leading practical application of AI. the searching for possibilities and alternatives. (3) and (4). rather it is ‘hunch-like’. perhaps performing some tests. Domain information is often stored in tables or matrices. Many systems use production rules of the form ‘IF A THEN B’ to store information types (2). Much of the knowledge that is used by human experts does not constitute definite decision sequences. and (5) conceptual models of the domain. AI researchers have developed techniques that attempt to represent human decision making. and then asks questions about the computer’s proposed solution.1 Representing expertise Representing the various types of knowledge that characterize expertise constitutes one of the main themes of expert systems research.162 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The introduction of symbolic processing opened the door for attempts to mimic the human mind. (2) hard rules or procedures. Expert systems are often designed with knowledge concerning: (1) facts about the domain. Much of this information must be stored in the program of the expert system using special techniques for knowledge representation. while the designers’ conceptual model of the problem is usually built into the program logic. Knowledge representation. and learning processes are all ingredients in AI research. The clearest distinction between expert systems and conventional computer programs is the flexibility of the artificial intelligence design. (3) problem situations and potential solutions. . The user of an expert system interacts with the computer in a ‘consultation dialogue’. (4) general strategies. Studying the ways humans solve problems. 2.

3 System processing Expert systems attack problems by feeding all the available information concerning the problem into the knowledge base that makes up the heart of the system.TECHNIQUES FOR EMERGENCY PLANS 163 2. This ability is one of the distinguishing features of consulting with human experts and is implemented on computer systems to improve the user’s confidence in the system’s judgment. The issue of knowledge presentation is also an open research question [4]. In systems involving diagnostic problems.2 Transfer of expertise The primary bottleneck in the development of expert systems is the acquisition of the knowledge of the expert. this process is often arduous and inexact and can lead to long delays in producing a working expert system [3]. With the availability of the exact reasoning process followed by the system. the program uses its conceptual model to suggest tests to be performed or questions to be answered.4 Explanation of knowledge A key feature of many expert systems is their capability to explain their reasoning in understandable terms. differing philosophies are used to explain whether the system can actually mimic the reasoning processes of the expert. it is easier to convince users that the solutions are valid and reliable. . Often this consists of production rules that generate possible hypotheses or solutions to the problems. However. 2. Typically the system designers consult with one or several ‘experts’ for long periods of time during the development stages. These narrow the solution range in order to enable the system to reach a valid conclusion. 2. Since intuition and opinion are part of this knowledge. Several systems use an elaborate solution technique but then attempt to explain their solutions in conventional ways. the process of confirming or narrowing the solution begins. Once initial possibilities have been determined.

Communications and control are difficult at best. they ignore critical information (which they are unable to process or relate to the events facing them) and resort to rule-based behavior. managers face a unique decision making process. The latter concept is related to the psychological phenomenon known as . will result in a reduced quality of performance. The need to provide decision support to emergency managers is readily apparent from theories of decision making. crisis managers cannot analyze options available to them. As a result. The sequence of events and decisions at Chernobyl. individuals try to compensate for their deficiencies by constructing a simplified representation of the problem and by behaving rationally within this representation [7]. Feeling intense time pressure and operating under the stress created by dealing with life-threatening events. Bhopal. Successful emergency managers are people who have a valid mental set of rules or can instantly determine upon which ‘experts’ they can rely. One theory suggests that a decision maker has an optimal band of information processing capability that. Knowledge-based support and control systems may be used to evaluate and determine correct courses of action.164 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3 DECISION SUPPORT FOR EMERGENCY RESPONSE [5] In an emergency. and in Switzerland are evidence that the heuristics of the first-line responders and crisis managers may not be appropriate to the scale of the crisis that they face. Another theory suggests that. or by calling attention to exceptional conditions. appropriately used. The success or failure of the response operation is dependent upon the validity of the rules selected. The probability that information-free. Dynes and Quarantelli [6] state that decision making during crises is marked by a rapid increase in the number of decisions made and the volume of information that must be processed. to perform functions automatically (thereby diminishing the volume of decision making that must be done). when information handling capabilities of an individual are overwhelmed. Information technology. when breached. can support the judgment of crisis managers and can also aid in the actual management of a crisis response. and these organizations are likely to include many members with inadequate expertise. trust-based decision making will succeed is diminished by the fact that emergency management involves rapidly changing ad hoc organizations.

FIG. 1. TECHNIQUES FOR EMERGENCY PLANS 165 . Flow of action in emergency management operations.

A decision aid (box F) can dampen feedback in the loop (and thus lessen the impact of stress and surprise on the decision maker) by reducing the decision maker’s information overload. and information processing abilities (box B). in particular. The closed-system nature of the conceptual model indicates the large degree of dependence among the factors. specifically with regard to surprise (box D). The decisions made in one stage of the disaster are constrained by the time available and they. information technology decision aids. in turn. which results in a more efficient use of the decision maker’s cognitive processes. caused by the inability of the decision maker to match his or her information processing capabilities to the information demands. a reduction in stress.e. his or her cognitive capabilities (box A). and an assessment and evaluation of the tradeoffs associated with the various alternatives (box G). to reduce data flows to prevent information overload (box C). Stress occurs and resultant dysfunctions are apparent in the decision maker’s performance. in the selection and design of expert systems to support emergency response decision making. but also a product of the individual’s role in an organizational system and of the availability of appropriate resources to support the decision making process. as well as the ability of the system to insure that the data are in a form meaningful to the user and of value in the decision process. The quality of the decision made by an individual in a crisis setting (box I) is a function of the quality of the information received (box H).166 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS ‘cognitive strain’. when necessary. The flow of actions in emergency management operations follows the dynamics of the event. Figures 1 and 2 illustrate the flow of decisions in a crisis event. real or perceived. The major difference between emergency response situations and routine decision situations is highlighted by the ‘loop’ in Fig. The quality of the information used in the decision process depends on the ability of the information system to monitor and. constrain the decisions made in later stages. 2 (from box A to box E). Decision making can be improved. and improved information processing. Decision making performance is not only a function of the psychological attributes of the individual. therefore. Reducing the information overload results in an increase in available response time. . Figure 2 is based upon Smart and Vertinsky’s [8] conceptualization of the decision process during crisis. stress (box E). by selecting people with the proper psychological prerequisites. Our interest is in this latter category and. or by engineering the environment so that better decisions result. i.

(2) a data analysis capability. (3) normative models. For example. Normative models can assist the decision maker by providing solutions that are not obvious. but data on current conditions can also be stored and processed as needed. In many circumstances the data are processed statistically to provide specific types of information that may be useful in obtaining appropriate decisions. A conceptual model of the role of a decision aid in an individual crisis decision process. This information is typically obtained prior to the decision situation. 2. (4) expert systems.TECHNIQUES FOR EMERGENCY PLANS 167 FIG. This system would interact with the decision maker and collect data from the environment. and (5) an interactive technology for display and use of the data and models. 4 EXPERT SYSTEMS AS A DECISION AID FOR EMERGENCY RESPONSE We have identified the components of a decision support system for emergency management in Fig. 3. evaluating tradeoffs . The five subsystems are (1) a data bank. projections of human resource requirements and staffing constraints can be combined to yield a forecast of future recruitment and selection goals. either directly to the data bank or from the user. The data bank stores information obtained from the operating environment. The data are then presented to the decision maker in their original configuration or after transformation according to one or more models.

as we indicated in the previous section. It then provides a recommendation to the decision maker using the system rules. . The decision aid must take into account the differences in how individuals approach the problem solving process. Components of a decision support system. 3. the technology required for display and interactive use. Thus the interface technology must be flexible and provide several display and retrieval alternatives. 3. Conceptually the system could manage the interaction between the models and provide advice to the decision maker on the appropriate model(s) for a decision situation. The expert system technology. a model management system could act as an interface between the decision maker and the various models. takes the data from the data bank and assesses their applicability to the decision process in question.168 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. and providing recommendations on a specifie course of action to be taken. Although not explicitly shown in Fig. may be the most important part of the system. The last subsystem. between alternative solutions. Even the most ‘optimally’ designed system may go unused if the information is not presented in a form that supports the decision maker.

public information. Typical response proceedings during a radiological emergency are depicted in Fig. of which 57 are AND/OR trees and 7 are IF/THEN tables. In the first one. accident assessment. It includes activities such as direction and control. whereas in the second a data base or file is interrogated and the user is only required to confirm or supply ‘judgmental’ answers. public notification. 4. communication. Typical response proceedings during a radiological emergency. the nuclear industry has undertaken a major initiative to improve nuclear plant emergency response capabilities. The objective of this prototype system is to aid in the decision making process. evacuation. The system queries the user for identification and then prints a list of procedures that have to be followed—matched to the responsibilities of the user. state and the private sectors as well as the coordination among them.TECHNIQUES FOR EMERGENCY PLANS 169 5 AN EXAMPLE IN NUCLEAR PLANT EMERGENCY RESPONSE [9] Since the TMI-2 and Chernobyl accidents. FIG. questions are asked of the user at each step of the inference process. Progress has been made in developing computerized data systems to support the emergency response facilities. . The knowledge base was taken from the emergency level classification for the Indian Point Nuclear Power Plant and the procedures from the State of New York Radiological Emergency Preparedness Group in the USA. It consists of 64 modules. A prototype expert system was built to provide decision support for emergency response. The system can run in two modes: interactive or semi-automated. response management must include federal. etc. The system was built using the GEN-X expert system shell product of General Electric. However. protective response action. 4.

170 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

TECHNIQUES FOR EMERGENCY PLANS 171 .FIG. 5. Logical structure of the expert system.

Artificial Intelligence as a discipline is just in its infancy. The example we used was the case where these managers are faced with an incident at a nuclear power generator facility that has the potential for catastrophic consequences. It is designed using a hierarchical top-down approach. site area emergency. 6 DISCUSSION We have attempted to demonstrate the role of expert system technology in providing support to emergency managers.e. interrogating a fact/data base and asking the user only a few questions. It has two running modes: interactive. Of particular note is the use of mobile cellular telephone systems with expert system technology [11]. simulated the events surrounding a nuclear accident. unusual event. and semiautomated. i. local or private sector. and have. it is used operationally. alert state. This can be accomplished by taking advantage of its modular design as well as its ability to perform forward and/or backward chaining. a CAN’T ANSWER response at one stage will produce a call to an inner module that asks for more specific parameters. The user can be any individual involved at the federal. asking questions at every step.172 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The system has several levels of nesting. A schematic of this process is shown in Fig. We have only discussed one area in this rapidly expanding (and often confusing) field. general emergency. state. and comparing the answers with those obtained in the field.e. We can. This process goes on for several levels until an answer is obtained. We might also consider other new advanced information technologies as possible candidates for supporting emergency response. This prototype system is being validated by implementing various plant scenarios from past drills. The difficulty with its use in emergency response is the lack of routine in the activities required to manage the emergency. but have not tested this technology even in that environment [10]. 5. radiological. i. natural hazards or other type. power loss. The events that determine the above conditions are classified as: thermohydraulic. The plant conditions can be one of the following: normal operation. i. The area of expert systems seems to hold promise for implementation in a meaningful way.e. going from general to specifie questions until an answer is obtained or can be inferred. safety systems actuation. The availability of .

NY.A. Public Administration Review. Other references used are BARR. Italy (in press). while a US government report by the Congressional Research Service of the Library of Congress. Disasters. Expert systems as decision aids for public managers: an assessment of the technology and prototyping as a design strategy. 30–9. The result would be recommendations for emergency response. A Guide to Expert Systems.e. M. (1976).A. how we instil confidence in the user of expert systems. M. Expert systems as decision aids for disaster management. W. Decision Sciences and Engineering Systems Department. & FEIGENBAVIN. S. Washington. and MICK. W. & WALLACE. Reading.W. & WALLACE. accurate digital information will complement the judgment capturing and processing capabilities of expert systems. Stanford University Press. CA. Technical Report No. 1. (1987). i. J. Very realistic and useful graphics and voice interactive capabilities will be available within the next few years. Interfaces. 563–71. Information and Management (in press). W. DOD/ARPAN00014– . Information Technology for Emergency Management. Handbook of Artificial Intelligence. & WALLACE.R.A. S. NOTES AND REFERENCES 1. 1987 World Conf. A. & WALLACE. E. and WATERMAN. The problem of transfer of expertise or knowledge acquisition has yet to be resolved. S. See BELARDO. 14(2). & WALLACE.. (1985). MA. AddisonWesley. Vol. surveys the topic.A. 4.A. Organizational Communication and Decision Making in Crises. 2. on Chemical Emergencies. (1985). Special Issue. W. D. Presenting uncertainty in expert systems: an issue in information portrayal.. Managing the response to disasters using microcomputers. becomes the key to user acceptance and understanding. (1986). (Eds) (1979). R. displayed graphically in ‘real time’. Rome. one example of ongoing research is GRABOWSKI. 6. HARRALD. and BELARDO. W. see LAMBERTI.A. W. 9. for examples. Knowledge Acquisition Using Prototyping: An Empirical Investigation. (1984). DC.TECHNIQUES FOR EMERGENCY PLANS 173 rapid. This section draws heavily upon HURLEY. Troy. Stanford. US Government Printing Office. & QUARANTELLI. The impact of advanced technology will be most evident in the way the user interacts with computing and communication technology. Knowledge based decision support systems for responding to chemical accidents. & WALLACE. KARWAN. D.A. C. DYNES. 3. E.A. 98–101. in time to support the decisions that have to be made to protect life and property. A more detailed discussion may be found in BELARDO et al. The issue of knowledge presentation. (see Note 1). Proc. 1984. 5. 37–87– 120. Rensselaer Polytechnic Institute.

H. 14(4).M.. . and to be published in Proc.A. Troy.L.. Simulation of a crisis management information network: A serendipitous evaluation.A. W. 9. W. C.. Boulder. PAZER. 14(6).R. (1980). Advanced Research Projects Agency. On Managing Disasters: The Use of Decision Aid Technologies. (1987). & DANKO. see BELARDO. (1983). TAYLOR.A. An expert system for improving nuclear emergency response.. 75–C–0458. USA. W. & VERTINSKY.A. & YEATER. R. This diagram is based upon SMART.. (1984). WALLACE. Utah. Artificial Intelligence and Other Innovation Computer Applications in the Nuclear Industry: Present and Future..A. M. A. WALLACE. (in press). S. 6(3). Department of Defense. Systems Man and Cybernetics. WALLACE. IEEE Trans. DC. R. 588–606. NSF Workshop on Natural and Technological Hazards. NY. We have assessed (fortuitously) simulation as a basis for training in BELARDO. 11. Snowbird. 640–57. K. Decision Sciences. Technical Report No. University of Colorado. 8. 22 (4). RYAN. An investigation of system design considerations for emergency management decision support. Psychological determinants of branded rationality: implications for decision-making strategies. Administrative Science Quarterly. W.L. Decision Sciences.174 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 7. KARWAN. 37–87–118. (1975). I. Washington. GOLDBOGEN. 31 August-2 September. Rensselaer Polytechnic Institute. and have used gaming to test decision aids.D. Designs for crisis decision units.H. 10. & WALLACE. 795–804. W. CO. SALAME. (1987). 409–29. G. S.

To date the instrumentation and the tools to achieve this are very simple and unreliable. * SMART: System for Measurement-based Assessments of Released Toxicants.G.21 Improved Emergency Response after Release of Toxic Substances: Application of the System SMART* D. Once they have an idea of which material has been released.KNAUP Brenk Systems Planning. it can be caused by a fire or it can follow a transport accident with hazardous materials. they require information on emission rates and meteorological data to estimate the consequences of the release to people in the vicinity of the accident site. This important decision in off-site emergency management as to which protection measure must be taken for the general public is generally left to the first on the scene. Measuring is carried out with absorption tubes. If this is compared with the actual course of a released cloud (Fig. H. the fire brigades. FRG An accidental atmospheric release of toxic substances can take place in chemical factories. Prediction of the propagation of the released material is effected with the aid of some precalculated plume models (Fig.BRENK & A. 2) we can see that the prediction made using Fig. Identification of the material released is achieved by UN number. Emergency response forces have to be able to act properly in all these cases.D. a quick and accurate estimate of the toxic air concentrations in consequence of the accident must be made and a decision on protective measures must be deduced from this estimate. 1). FRG & H. or by questioning the manufacturer. Cologne. if it is carried out at all.DE WITT.HESEL TÜV Rheinland. Whatever the reason for the contingency may be. Aachen. 1 can lead to a severe misjudgement of the real conditions. .

measuring air concentrations and of course making a decision on which protective measure to take. a personal computer with relevant software. the lack of versatility of the devices and the costs. Figure 3 presents a rough sketch of the unit and its equipment. assessing the released quantities. dosimeters for individual substances. are not (or not yet) suitable for use by fire brigades. portable GC/MS devices and so on. The key to the developed system is the computer software. To improve this situation. the German Ministry for Research and Technology has sponsored us to develop a mobile unit for use by the fire brigades. FIG. The developed unit consists of a command vehicle. Actual shape of a propagating cloud under realistic weather conditions. wind speed). following the released and dispersing cloud. which was designed to help the fire-fighter perform his tasks of identifying the released substance. a meteorological mast and three very simple measuring units equipped with absorption tubes. which guide the operating fire-fighter through . we found out that all other devices. such as infrared spectrometers. The main reasons for this are the expertise required to operate and to maintain the systems. 1. input and output devices. This disappointing result of our study forced us to stay with the absorption tubes and to incorporate them into the system.176 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. Example of a precalculated plume model (variables: wind direction. The software provides masks on the screen of the PC. 2. Concerning the air concentration measurements.

All these modules are controlled and initiated from a main programme. This module SMART permits best possible elimination of uncertainties connected with each of the two factors dispersion calculation and measurement. by means of two key elements. which also takes care of the input and output. In short. Figure 4 summarizes the modules which make up the programme. the programme. see . the modules perform the following tasks: — Identification of the released material — Estimation of the released quantity using plant-specific data — Recording and processing of meteorological data — Processing of air concentration data — Estimation of source parameters — Dispersion calculation with adjustment to the measured air concentration data — Output of prediction and recommendations While most of these modules can be found in similar approaches.TECHNIQUES FOR EMERGENCY PLANS 177 FIG. the module which carries out the dispersion calculation and adjusts it to the actually measured air concentration data is unique. 3. The mobile unit and its basic equipment.

4.178 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS FIG. Depending on the specific question to be answered. Subsequently these parameters are used as input data for the second key element of SMART which is real-time projection of the dispersion situation. which is then initiated. 5. In mathematical terms this is realized with the aid of the following equation which is a modified least-squares fit formula: . After the release of a toxic substance a first approximation of the dispersion situation is made as usual. The fundamental calculation procedure. In the second step. e. for which measurements are provided. the adjustment may be accomplished by evaluating 5 model parameters such that the procedure results in a best fit within the scope of the prediction accuracy of the dispersion model used. 10 minutes. This results in a more realistic set of model parameters.g. without knowledge of measured air concentrations. Fig. the computer module SMART incorporates the current concentration values measured at different locations and compares them with the modelled concentration values. This is repeated for each time interval. The first element is the feed-back of current air concentration measurements into diagnostic calculations in order to adapt the calculations to the measurements. Modules of the computer code applied in the emergency response system. can be described as quasi-continuous adjustment of the calculated to the measured concentrations. including the source term. The feed-back of current air concentration measurements into diagnostic calculations is realized by the following adaptation sequence.

TECHNIQUES FOR EMERGENCY PLANS 179 FIG. N is the total number of absorption tubes and k is the number of trials to find the best fit. 5. . where CC and CM are the calculated and measured off-site air concentrations of each measurement location i at the end of the time interval n. Logical scheme of the corrective/predictive procedure of the System for Measurement-based Assessments of Released Toxicants (SMART). based on Monte Carlo techniques and guided by their sensitivity with respect to the dispersion calculations. The model parameters are: — Wind direction — Wind velocity — Standard deviation of horizontal and vertical wind direction fluctuations — Source term These parameters are evaluated according to a particular ‘trial and error’ strategy.

This is a restriction which is not necessarily true. This set is then used to project pollutant concentrations for subsequent time periods. particularly for long time periods. It can be opened.180 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The adjustment procedure provides a most realistic and accurate set of effective model parameters to describe the current dispersion situation. This allows a proper adaptation to both changing weather conditions and varying releases of toxicants. It helps the emergency forces to identify the released substance. . however. the system is in its test phase in a joint effort with the Cologne fire department. it recommends measuring points and measuring equipment. it measures and records meteorological data. Test applications with experimental dispersion data and first implementation in the emergency response system have revealed good operational performance of SMART. easy to maintain and reasonably priced. The complete system which is installed in the mobile unit performs a number of tasks. above all. by repeating the measurements and the adaptation procedure within sufficiently short time interval such as 10 minutes. The projection is based on the assumption that the meteorological conditions and the release rates are constant during the prediction time. Therefore it can be stated that it provides very realistic and accurate information on the dispersion situation after an accidental release of toxicants to the atmosphere. It thus enables the first on the scene to make a quick decision about necessary protection measures. At the moment. it registers every step of the accident. The equipment used is still relatively simple. it acts as a decision aid.

again show the need for good preparation for such situations. like the LPG escape in Mexico and the methyl isocyanate escape in Bhopal (both at the end of 1984).22 Emergency Management of a Gas Escape C. One of the means to reduce the risk that people on-site as well as off-site will suffer from the consequences is to draw up an emergency management plan. Several elements can be distinguished: (a) Risk analysis of the hazardous activity. In this paper the TNO analysis of the Mexico and the Bhopal disasters will be used to illustrate the necessity for all three elements. number of people involved) — development of the disaster in time (b) Analysis of the type of action which will be the most effective during an emergency (and what type of organization should be set up) (c) Development of the tools to be used in an actual emergency An analysis of accidents that actually occurred will be of great benefit for all three steps mentioned.PIETERSEN Division of Technology for Society TNO.M. what types of accident will happen: — magnitude of the escape — possible consequences (area. The Netherlands 1 INTRODUCTION Large industrial disasters that happened recently. Apeldoorn. but that does not mean that industrial disasters will cease to take place. It is the responsibility of the industry as well as the authorities to reduce risks to a minimum. Such an emergency response plan should be drawn up after a thorough analysis of the particular situation. These accidents represent .

also available in software packages for personal computers. to be used in actual emergencies. Quite often the ‘maximum credible accident’ approach is followed. magnitude. The judgement of what is and what is not credible is of course a subjective one. 2 RISK ANALYSIS OF HAZARDOUS ACTIVITIES FOR EMERGENCY RESPONSE PLANNING In order to be able to set up an emergency response plan for a particular industrial activity (installation. TNO is currently developing such a system on behalf of the Dutch Ministry of Internal Affairs (Fire Inspectorate). The importance of preparing an on-site emergency schedule (for the workers) is based on risk assessment techniques and is illustrated with TNO wind tunnel modelling of NH3 escapes at a Dutch chemical plant.182 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS completely different situations. in order to set priorities. (2) A calculation of what will happen to the surroundings of the activity in the case of an escape: (a) Effect calculations: — concentrations in vapour cloud as a function of distance — passage time of a vapour cloud — heat radiation as a function of distance — overpressures upon explosion etc. for which completely different emergency responses will be necessary. transport or handling).). location and time requires an automated decision support system. It also leads to large differences in requirements with respect to the type of organization (number of people. These effects can easily be calculated via existing models [1]. The diversity of possible disaster situations by type. (b) Consequence calculations: . delay in being present at the site etc. a risk analysis of the activity should be drawn up: (1) Identification of representative accident scenarios: What can go wrong? A certain indication of probability of the scenarios will be useful. This paper gives a short description of the possibilities of that system.

19 November 1984 In November 1984 a disaster involving an LPG installation in Mexico City resulted in the death of over 500 people and 7000 people were injured. it is worthwhile to analyse some accidents in more detail. also with regard to the emergency response aspect. — damage to different types of building These calculations need introduction of a specific environment. An analysis of comparable accidents that actually occurred will be useful in order to verify the models that are used to estimate the development in time and to learn from already taken emergency response actions. 3 THE GREAT DIFFERENCE IN EMERGENCY RESPONSE REQUIREMENTS OF THE MEXICO AND BHOPAL ACCIDENTS 3. approximately 900 had to stay there for further treatment. By 25 February 1985. Accident data are available for instance in the TNO databank FACTS. A TNO team visited Mexico shortly afterwards to carry out an investigation [2]. They handled 7231 wounded. The Mexico and Bhopal accidents are examples of such an analysis. of which 5262 were treated in provisional emergency centres. However. During the disaster 985 medics. . heat radiation dose etc. 1780 para-medics and 1332 volunteers were giving medical assistance. Of the 1969 wounded taken to 33 hospitals. 1.TECHNIQUES FOR EMERGENCY PLANS 183 — number of injured and killed people as a function of toxic dose. The area in question is given in Fig. This investigation was mainly directed to a check of the existing damage and effect models and to the emergency relief that had taken place. Some figures in relation to the emergency relief are given below. 710 patients had recovered. 32 were still in the hospitals and 144 people had died there. TNO has currently developed software for personal computers that will make these calculations possible [4].1 Mexico LPG disaster.

with a total of approximately 125000 hot meals. the emergency services used 363 ambulances and 5 helicopters.) — Evacuation of people — Stoppage of leak — Some time can be gained by effectively cooling (water spray) the tanks.184 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS For transportation purposes. . A vapour cloud formed and was ignited shortly after the first leak (±10 minutes). The Mexico disaster developed very rapidly. 1. due to the rapid sequence of events this should be an automatic system — The emergency relief organization for this type of disaster should be able to act very fast and to be present within a very short time FIG. however. Each day 35 000 hot meals were provided. Reproduction of the area in which the damage occurred. Eleven provisional shelters were established for 39 000 homeless and evacuated people. some time may be available to take certain actions: — Removal of ignition sources (traffic etc. The escalation to a complete disaster took place in the next few minutes. Generally it can be stated that in this type of accident the warning time is very short.

In Bhopal hardly any action was taken with respect to emergency response in the sense described above [3]. a massive escape of methyl isocyanate caused the death of at least 2500 people and over 100000 people were injured in Bhopal. — The development of the disaster allows time for important mitigating actions: evacuation etc. At the request of the official Indian scientific investigation team. India. Compared with the Mexico accident this type of accident shows important differences with respect to emergency relief. FIG. In Fig. In fact this running worsened the situation. A toxic cloud developed during the night. Important points are: — A warning time (even before any gas escape) may be available. In a TNO report [3] the accident and its emergency management are extensively described. the Union Carbide plant itself was also visited.2 Bhopal disaster. 4 December 1984 Shortly after the Mexico disaster. . Movement of people to the hospitals during the Bhopal disaster.TECHNIQUES FOR EMERGENCY PLANS 185 3. in Bhopal it might have been obvious 1 hour prior to the escape that a very dangerous situation was rapidly developing. 2. 2 is illustrated how people started to run into the direction of the hospitals without any guidance. they were running in the wind direction and therefore stayed within the cloud.

to be used in actual disaster situations. people should run to these rooms. This can be achieved by studying accidents from the past and/or via risk assessment techniques. More people outside or on their way outside were killed (see Fig. From descriptions of several accidents involving NH3 there is clear evidence that staying inside a building in the event of a release has an advantage. 3). it has been reported [6] that 10 men survived in a control room 80 m from the release. Adequate protective means and communication equipment should be present at the spot. depending on the circumstances. the calculation of concentrations of toxic or inflammable/ explosive gas clouds. It is also mainly for these accidents that TNO developed an automated information and calculation system. For instance. may be rather difficult. in built-up areas like chemical plants. Running in the up-wind direction is not necessarily valid.) may be such that at present only wind tunnel modelling can predict the shape and concentrations of the cloud. However. for the relatively small distances involved.186 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The scope of an emergency response organization to take effective actions is quite large for this type of accident. 4 ON-SITE EMERGENCY MANAGEMENT It is important to base on-site emergency management schemes on realistic accident scenarios. two of them were killed. Five men left the control room. The influence of obstacles (buildings etc. not only the down-wind area is covered but also the area up-wind and in the lateral direction. This should be realized in the case of an accidental release of a heavy gas. Gas-tight rooms are to be created and clearly indicated. For emergency management it is important that. . The dispersion of the gas in all directions and the protection by (relatively) gas-tight rooms has set clear requirements for on-site emergency management in the case of toxic heavy gas releases. they put wet cloths over their faces.

3.TECHNIQUES FOR EMERGENCY PLANS 187 FIG. . General layout of Potchefstroom plat.

The user can interact with the system. TNO developed an Information and Calculation System to support decision making in FIG. The system also gives an insight into the development of the situation as a function of time. Physical Planning and Environment (Directorate General for Nuclear Accidents). It also calculates the number of people that will survive and the number of people with several degrees of injury. .188 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 5 INFORMATION AND CALCULATION SYSTEM FOR EMERGENCY RESPONSE Commissioned by the Dutch Ministry of Internal Affairs (Fire Inspectorate) and the Ministry of Housing. in order to get informed: — The system calculates and shows the size of the threatened area and the number of people involved. The structure of the program is given in Fig. — The system is capable of analysing the effect of possible measures relating to a reduction of numbers of victims. 4. The system is still in a demonstration phase and will shortly be implemented in fire brigade regions in the Netherlands. Information and calculation system for emergency response. 4. disaster situations with toxic and nuclear material.

19 November 1984. Bhopal: Risk Assessment and Emergency Management. Ammonia Plant Safety. 1979. H. 6. TNO report. 5. TNO report. LONSDALE. 4. (1987). GULDEMOND. PIETERSEN.P. TNO. 126–31. (1975). The behavior of denser than air ammonia in the presence of obstacles: wind tunnel experiments. Analysis of the LPG Incident in San Juan Ixhuatepec. the Netherlands.M. PIETERSEN. Software Package RISKCURVE for Personal Computers. Also available as a software package (EFFECTS). C. Calculation of the physical effects of the escape of hazardous material (gases and liquids)—‘The Yellow Book’. 3. 2. 1987. . C. Plant/Operations Progress. Voorburg. C. 5(2). Directorate General of Labour. 17.M. TNO.TECHNIQUES FOR EMERGENCY PLANS 189 REFERENCES 1. (1985). Ammonia tank failure: South Africa. Mexico City.

event trees.SICILIANO & E. This phase identifies the accident scenarios presenting the most significant risks for which an emergency procedure is required. to identify the sequence of events/failures leading to an undesired outcome. L. .LAMBARDI. Genoa. to quantify their effectiveness in risk reduction This approach. models of accident analysis (fire. toxic release. for deciding upon the reference accident scenarios.23 Effective Emergency Planning Design by Means of Risk Analysis Models A. both in terms of their consequences and their probability — Detailed representation and modelling of emergency plans. V. This phase implements system models.DONATI. Italy 1 INTRODUCTION This paper discusses the analysis and design of emergency planning in an integrated framework of risk analysis. or whatever is relevant for the installation being considered). Then each emergency procedure combining to form the emergency plan is evaluated with a systems analysis approach. like any other system or physical process already modelled. Two aspects are taken into consideration: — Risk analysis of the installation (industrial or civil). considers emergency plans as virtual ‘safety systems’ and quantifies their effectiveness and reliability as such. like fault trees. 2 THE OVERALL MODEL The initial phase of the analysis consists of the risk analysis of the installation.SILVESTRI Ansaldo SpA. and introduced into the overall plant model. in effect.

reaction time to the alarm. two computerized models. a room. passages. . a building) and routes (doors. 3 SYSTEM MODELS FOR AN EMERGENCY PLAN In order to evaluate the efficiency of a designed emergency plan. a floor. During this time the plant or site situation may change (e. toxic release propagation) A complete model will take into account all the parameters and should evaluate the possible interaction between them. obstacles. bottlenecks) — Modelling the dynamics of people involved (e. smoke. Interactions with the phenomenology of the accident will be developed in a successive stage. behaviour in stress conditions) — Introducing the use of possible means of conveyance (e. Ansaldo has carried out a study of the first two steps and has developed.g. ambulances) and their availability — Analysing the phenomenology of the accident (fire.g. one for the dynamics of people’s movements and another for people’s egress probabilities. stairs. and the pre-established safety course may then be unfit for use. The construction of the model passes through many phases. To evaluate the adequacy of an emergency plan for a building it is necessary to be able to estimate the egress times of individuals (or homogeneous groups of people).TECHNIQUES FOR EMERGENCY PLANS 191 This approach requires the emergency procedures to be evaluated in terms of their physical effectiveness and of their probabilities to be effective (a standard probabilistic risk assessment approach). lifts. it is necessary to develop a model of the egress flow of people which points out the main parameters that may affect the success of the emergency procedure. vehicles. and the probability that this time is less than a predetermined value. the main ones are: — Modelling the structure of the plant (e. open spaces.g. in an initial stage. because of the accident or the environmental situation).g. Egress time is here considered to be that elapsing between the alarm signal and the moment the group reaches a pre-assigned zone. speed.

It is characterized by a mean speed and by various factors that may change it. the group is to be considered lost. A route may be interrupted at a certain time. The code takes into account the topological situation. this speed changes depending on the number of people present. Three different egress modes are envisaged: — Independent groups. the presence of obstacles on the escape routes. the ‘zone’ and the ‘route’: Group: homogeneous composition of people who.g. positions of people. consequently the traffic model permits march inversions and route changes. forbidden routes. the variation of the environmental conditions (e. Route: ordered succession of zones that each group crosses. Fz is the product of zone factors. if no alternative route is found. alone and in optimum conditions. Basic entities of the model are the ‘group’. depending on the varying plant conditions. and non-evacuable zones. It is characterized by a direction. It is also capable of searching routes alternative to the standard ones. Zone: each area into which the plant is subdivided. With each zone is associated a typical speed which corresponds to that of a person crossing it. V is evaluated through the equation: where Fg is the product of group factors. the code provides extension of the accident-related zones. At each time. and this may force the group to go back to look for a junction with a free route. amount of smoke). in emergency situations. It is characterized by a length and a surface. ‘to reach safe zones. Factors and speeds are . depending on the situation as known by single persons. on the environmental conditions and on the shape of the exits. Vz is the typical speed in the zone. run along a preset route to reach a safe zone.192 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The code EPDES (Emergency Planning DESign) is able to model the emergency traffic with the goal of evaluating the single person or group position. with no exchange of information among people about route practicability — Groups exchanging information about route practicability — ‘Panicking’ groups The group speed.

for example (see Refs 3 and 4) a state transition (Markov) model (the latter is applicable if nonlinear effects. This graphic representation allows identification. it is thus linked both with codes of accident analysis and. such as a direct Monte Carlo type simulation of the dynamic model by assigning appropriate probability distributions to input variables (in this approach response-surface methodologies may represent an aid) or by using an appropriate logical model. The EPDES code is part of a code family for the design of emergency plans relating to a reference scenario. thus determining the accident situations that are most relevant. with account taken of the emergency plan. i. verified according to pre-defined acceptability criteria. Refs 1 and 2).TECHNIQUES FOR EMERGENCY PLANS 193 obtained from experimental correlations (see. This can be accomplished in various ways.e. a stochastic simulation is performed on it. for example. The transition matrix can be automatically constructed and reduced by checking the consistency of its states and its topological characteristics. in the future. To predict the probabilistic outcome of this traffic model. of the consequences of events forced by the code operator. At the present stage of development. The EGRESS program models the transient states of occupancy of a building by means of a Markov model. where the values of state transition probabilities are based on the physical model just discussed. also. for design purposes. and therefore different emergency plans. So no arbitrary choice of route is allowed. This resulting risk can be then. even the alternative route in case of obstacles is fixed. the modelled emergency plans can only consist of a ‘rigid’ procedure. The code provides a 3D graphic representation that shows the state of the system at different time steps. with codes of probabilistic evaluation (presented by Ansaldo on the general organization of the methods and codes). In this way it is possible to compare different egress paths. which means that each group of people has an a priori fixed and known route to exit. the dependence of state transition probabilities on the visited states is shown not to be important). Implementation of these models with the appropriate input data will produce the various degrees of success of the emergency procedure at various times with associated probabilities. subcases of different egress paths can be studied by means of the restart option. The results of this analysis can be coupled with the risk analysis of the installation. in an interactive way. checking the people’s time to .

This can in general be obtained by introducing or modifying engineered or procedural features with the general effects of: — increasing the effectiveness of the emergency plan through improved performance (e. 4 DESIGN OF AN EMERGENCY PLAN The overall model just described can be used for design purposes with the aim of reducing the risk generated by the process that takes places in the installation. — increasing the probability of reaching a successful terminal state under various conditions by reducing the effects of such factors of variability that may cause a large dispersion of the probability distribution of the parameters of merit like the time for people to leave the site). The basic entities of the model in this case are groups. etc.g. Groups and routes have the same physical interpretation as in the EPDES code. . states. type of exit.194 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS egress at a given level of probability (or vice versa). egress rates. or reduction of dependence of its success on the success of single items or factors). if the installation is already in the operational state. this can be used to find the areas of the problem that need more details on data or on the model. by modifying the process intrinsic or active safety or. door. and routes. even if their expected or mean value is adequate. the rate depends on various factors: speed of people. and are defined as: State: possible configuration (occupation by people) of the rooms or spaces Rate: the mean flow of people from one zone to another. by improving the emergency procedures to be foreseen for its safe operation. At present no search for optimal routing is performed. In this way the probability of success/failure of an emergency plan may enter into the plant system event trees as a nodal probability like that of a protection system. A sensitivity analysis can easily be performed to check the relative influence of the parameters. whereas states and egress rates are physically correlated with the concepts of zone and of group speed. in particular. a reduction of expected times of execution of the procedure.

Dover. 1979. must then take cost-effectiveness and cost-benefit considerations into account. and other authors (1982–1987). N. . Fire and Flammability Handbook. REFERENCES 1. La Sicurezza contro Vlncendio degli Edifici a Struttura di Acciaio. T. 3. (1961). SCHULTZ. Elements of Queueing Theory with Applications.L. CASCARINO. Van Nostrand. Various articles. SAATY. (1985). 2. Antincendio (in Italian). Monografia 6 della Ricerca: II comportamento delle strutture portanti de acciaio alle azioni sismiche Parte II. A.TECHNIQUES FOR EMERGENCY PLANS 195 The choice between various alternatives at all levels. 4. if it exists.

large-scale threats to man and his environment had a natural origin. UK 1 INTRODUCTION There is nothing new about major hazards.CASSIDY Technology Division. by a succession of incidents causing widespread damage and death. Nor is the catalogue confined to damage to humans. Flixborough. storage and accidental ignition of black powder may well provide the first examples of large-scale damage from manufactured substances and artefacts. it is only their character that has changed over the years. Many major hazard risks have an element (which may indeed predominate) of environmental damage where effects may persist long-term. most would include catastrophes such as those at Oppau. Seveso and Manfredonia. San Carlos. first of risk analysts. Bantry Bay. it is even more recently that the threats posed by large-scale chemical engineering and energy processes have attracted the attention. The entry of chemistry. mainly storm.24 Major Industrial Risks: Examples of a Technical and Predictive Basis for Onand Off-Site Emergency Planning in the Context of UK Legislation K. of course. In the last couple of centuries. and Bhopal. Bootle. as a response to growing public concern. although the archaeological record amply demonstrates the potential for disaster that arose from mankind’s early attempts to harness the potential of fire. It may well be that the recent . Texas City. Many lists of such incidents could be produced. The justification for such concern has been demonstrated at regular intervals. all of which have been seminal in terms of public and regulatory response. In the Middle Ages the manufacture. large-scale water dam failure and boiler explosions accompanied the onset of the Industrial Revolution. Intially. flood and fire. and particularly of chemical engineering. San Juan Ixhautepec. and then of legislators. on to this stage has been relatively recent. Health and Safety Executive.

published by HSE [17]. and to emergency services. many of the latter in particular presenting environmental as well as humandamage risks. 7] This approach is very much an interdependent package of controls and responses. We have a system. which is centred in the following concepts [3]: — Identification: via the Notification of Installations Handling Hazardous Substances (NIHHS) Regulations [4] — Assessment and control: via the Control of Major Industrial Accident Hazard (CIMAH) Regulations [5]* — Mitigation: via the CIMAH Regulations (involving emergency planning and information to the public and land-use planning control [6. 4 ASSESSMENT AND CONTROL The general requirements of the CIMAH Regulations apply to sites which store or use hazardous substances which satisfy criteria * Extensive guidance on and interpretation of the CIMAH Regulations can be found in a guide to the Regulations.TECHNIQUES FOR EMERGENCY PLANS 197 pollution incident at Basle will prove to be a watershed of equal significance to the incidents listed above. it stimulates greater onsite awareness of the hazards and risks. it permits identification of such sites to land-use and emergency planners. and. hopefully. . appropriately tailored to the relevant risks. based mainly on the advice of the Advisory Committee on Major Hazards [1] and confirming European standards [2]. 3 IDENTIFICATION There are in the UK some 1750 installations subject to NIHHS. 2 THE UK APPROACH The UK has been a leading architect in the framing of legislative control for major industrial hazards. and several hundred more now notified under CIMAH. The requirement for statutory notification has a number of effects: it gives priority to HSE attention.

to intervene in the escalation process. which store large quantities of flammable toxic or explosive materials. it is . in the case of some aspects of emergency planning. This latter approach apart. Further. and small inventory top tier sites (SITTS). however. Several thousand such substances have been identified as being in regular use in UK industry.1 The safety case Emergency planning and information to the public are measures primarily designed to mitigate the consequences of any major incident. be remote). on request (and produce documentary evidence as appropriate).and off-site emergency plans — Provision of appropriate information to the public — Submission to HSE of a ‘safety case’ 4. that he has considered the potential for major accidents from his operations. reasonably practicable precautions have been taken. and for which much lower thresholds (1 tonne or less) are prescribed. and has taken all appropriate steps both to prevent their occurrence and to mitigate the consequences of any which may occur. should it occur (the probability of a major accident should. In the case of hazardous installations. however. The additional duties which fall to the occupiers of such sites are: — Preparation of on. the operator of the site must: — notify the HSE of any major accident which has occurred on his site. This is a general requirement of UK law [8]. such questions should be concerned with the residual risk after all appropriate. more specific duties under the Regulations apply to sites on which are stored or used certain substances in excess of specified thresholds. or. In the UK there are over 200 LITTS and several hundred SITTS notified to HSE. In such cases. — be prepared to demonstrate to an inspector. reactivity or explosiveness. with details of steps taken to prevent its recurrence (to be a ‘major’ accident it need have no more than the potential for harm). which store or use materials that are considered particularly toxic.198 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS related to toxicity. These sites are known as large inventory top tier sites (LITTS). flammability.

TECHNIQUES FOR EMERGENCY PLANS 199 reinforced by a specific requirement to present to HSE a written report (the safety case) which: — describes the installation. whilst identifying any remedial action. It identifies the critical areas. Additionally. and information to the public. This is not a process of approval. — analyses the effectiveness of the safeguards (both hardware and software) that have been applied. The analysis is. — reaches conclusions about the risks presented by the installation. a written demonstration of the application of good management techniques to major hazard control. 3 MITIGATION The main elements of mitigation are planning. and places it in its geographical and social context. emergency 5. and — on the basis of the above analyses. on a formal basis. Many of our hazardous installations are not ideally located with respect to adjacent developments. location. and the hazard ranges of some of our industrial processes may be very great. since 1972. and the hazard analysis carried out at an early stage of the assessment process highlights inter alia the relevant areas for potential mitigation. there is an existing legacy of previously permitted and continuing incompatible development. however. which can then be addressed on a concentrated and continuing basis.1 Location Adequate mitigation of major hazard risks is best achieved by planning control of incompatible land uses. Such controls have been applied in the UK. We are. Neither is it a once and for all exercise. as there are revision and updating requirements. — identifies any relevant major accident hazards. however. or one of licensing. This most powerful tool of control is therefore only partially applicable . the inhabitants of a small island (where intensive landuse is at a premium). effectively justifies the continuation of the operation. including that provided by adequate emergency planning.

if promptly detected and dealt with. may have virtually no ill effects.200 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS to the existing situation.g. despite continuing developments in UK planning law [9] which will shortly introduce a ‘consent’ procedure for hazardous installations. when necessary. The spectrum of possible incidents may be very wide. The recent SIESO booklet is an important addition to this corpus of advice [13]. operational measures are therefore required. however. The objectives of emergency plans are to contain and control incidents. 12]. to safeguard employees (and anyone nearby who might be affected). 5. If allowed to escalate.and off-site as a result of foreseeable emergencies and what the effects of an incident could be on the environment. Additional. They should then assess what dangers could arise to people on. the following criteria will apply. 5. the combined resources of the organisation concerned and the public emergency services. any incident may have serious consequences both on and off the site. As a minimum. vulnerability. involving close cooperation between the site operator.2. The smallest. the county authority and the emergency services. .and off-site emergency plans. The following principles are relevant in the production of emergency plans. the USA [14]. the local authority.2 Emergency planning CIMAH requires effective arrangements for on. e. Any relevant analysis will therefore involve an investigation of hazard. Similar guidance is being produced in other countries. and to minimise damage to property or the environment.1 Assessment of the hazards and risks Manufacturers need to assess their activities to ensure that all that is reasonably practicable is being done to avoid or reduce danger. General advice on emergency planning has been published in the UK by both the HSE [10] and industry [11. and risk. This should be followed by consideration of how these could be mitigated by preplanned remedial and rescue measures using.

permanent) — Indirect hazards/risks There are many uncertainties in the predictive modelling of all the above issues. — Extent of the vulnerable zone (the significantly affected area) and the conditions that can influence the impact (e. sensitive populations—hospitals. For these reasons. in terms of size and types (residents. This will assess the probability of damage (or injury) to individuals or to the community due to a hazardous materials release.g. — Types and quantities of hazardous materials located in (or transported through) a community — Location of hazardous materials facilities (and routes) — Nature of the hazard most likely to accompany hazardous materials spills or releases. similar uncertainties occur in real situations. and may be misleading. old folk’s homes etc. topography) — Population. and the actual damage which might occur.TECHNIQUES FOR EMERGENCY PLANS 201 (a) Hazard identification. size of release. in the light of the vulnerability analysis. Assessments purporting to give closely defined accuracy are suspect. (b) Vulnerability analysis. expected to be at risk within the vulnerable zone — Essential support systems which may be affected by any incident — Any particular risks to the environment (c) Risk analysis. wind direction. in any event. delayed or chronic — Types of damage to property (temporary. where a degree of realistic . permanent) — Types of damage to the environment (repairable. such precision is more relevant (if achievable) to a developing incident than to a preplanning protocol. It will include information on: — Event probability — Relevant environmental phenomena — ‘Domino’ effects — Types of harm to people (including high risk groups).). employees. schools. a relative simple broad-brush approach is the preferred option. repairable. whether acute.

there is no doubt that. For example. but detailed planning should concentrate on those events that are more probable. If such a disaster occurred. Provision will need to be made for the call-out of the other key personnel when they are absent from the site. and the Site Main Controller. in practice. The manufacturer has to . and the action by the emergency team. but more minor events may need to be considered. the risks of aircraft crashing on to an installation that is not within a few miles of an airfield are remote and the consequences need not be considered in detail. Examples of hazard ranges from substances presenting explosive. Seismic effects are unlikely to result in major vessel failure in the UK. The plan will also set out the arrangements for initiation of the plan.202 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS conservatism and pragmatism is desirable. flammable and toxic hazards will be presented at the seminar. These will include the Incident Controller. arrangements will be made to ensure adequate emergency response. Key personnel will be identified. 5. Nominated key personnel having immediate tasks to perform will always be available. (b) Off-site plan. safeguarding those on-site. whose primary task is to take charge at the scene. the emergency control centre or centres. the existing peace-time emergency plans would form the basis of the response by the emergency services. as well as the Incident Controller or deputy and an emergency team. Flexibility will also allow the response to be extended and increased if extremely remote combinations of adverse events and circumstances lead to consequences larger or more severe than those that formed the basis for the emergency plans.2.2 Content of the emergency plans (a) On-site plan. Where the level of manning does not give cover round the clock. Flexibility of response is a paramount requirement. The off-site emergency plan will be based on events identified by the manufacturer that could affect people and the environment outside his premises. with some indication of the likelihood of each type of event and its relevance for emergency planning. Emergency plans must be capable of dealing with the largest incidents that can reasonably be foreseen. A selection of these examples is given in the Appendices. Plans must also have sufficient flexibility so that the response is tailored to the severity of the incident. with overall responsibility for directing operations from the Emergency Control Centre. raising the alarm.

so that a command and response structure is in place before the event. This is particularly so in the case of ‘mixed’ hazard sites. 16]. again. but it must be sufficiently flexible to allow for the remedial measures to be extended and increased to deal with extremely adverse combinations of circumstances and consequences or with an escalating situation. The plan will need to cater in detail for those events identified as being most likely. Potential hazard ranges may be very great. Where environmental risks are present (and especially where they predominate). special arrangements will be necessary. depending on the size and characteristics of potential incidents. and potable water and food supplies may be at risk. ambulance. or where there is significant risk of escalation. It is essential that any arrangements include a suitable off-site Emergency Control Centre. have duties to deal with emergencies and accidents of all sorts. The plan will identify and detail immediate action to be taken to protect those in danger and arrangements for caring for those affected by an incident. shut doors and windows. . etc. Such risks may. he will take overall command of the off-site activities. being indoors may provide initial protection. Several different responses may be necessary at a single site. The plans will ensure coordination of existing services targetted to hazards specific to the industrial installation. In many cases the advice on immediate action may be to stay or go indoors. The recovery phase should also be preplanned. The plans should set out a command structure and identify the respective roles and responsibilities of the senior personnel involved. with control on the ‘fire ground’ by the fire authority. An Emergency Coordinating Officer may be designated. special problems may be associated with environmental risks. Indeed. be delayed rather than immediate. tune in to the local radio and await further instructions (normally from the police). fire authorities. In the UK. the police will have overall control of an incident. The Chief Executive of each Emergency Planning Authority will normally designate an Emergency Planning Officer to oversee the plan. Evacuation may present special problems. police. The emergency services. extent and likely effects of such incidents. but on the longer term it could increase the risk [15. however.TECHNIQUES FOR EMERGENCY PLANS 203 provide the emergency planners with information on the nature.

as appropriate) as part of the preparation and realisation of an emergency plan. — a description of the operations on site. however. Off-site. and thereafter to be rehearsed at suitable intervals. For this reason the UK CIMAH Regulations impose an additional duty to inform persons who are within an area that it is for the HSE to define (usually the land-use planning consultation distance). preplanned procedures are essential. such detailed briefing and preparation will rarely be possible. 5. as was clearly shown at San Juan Ixuatepec and Bhopal After each rehearsal or practice.3 Rehearsals and training Both on-site and off-site emergency plans need to be tested when first devised. In addition.3 Information to the public Any emergency planning depends for its success on an appropriate response from those covered by the plan. in the initial shock and confusion of a real incident. and of the hazards and risks that might affect the recipient of the information. its effectiveness should be reviewed every time it is used to deal with a real emergency. for a number of reasons: (a) They familiarise on-site personnel with their roles. The minimum information to be given is: — that the hazardous installation is notifiable.2. and this necessitates adequate briefing of those liable to be affected. and . they also familiarise them with the special hazards (c) They prove the current accuracy of the details of the plan (d) They give experience and build confidence in the team members.204 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 5. plans should be reviewed to take account of any shortcomings highlighted by the exercises. their equipment and the details of the plans (b) They allow the professional emergency services to test their parts of the plan and the coordination of all the different organisations. and has been notified to HSE. Onsite personnel should receive this briefing (and training.

TECHNIQUES FOR EMERGENCY PLANS 205 — any emergency measures (including appropriate personal behaviour) to be taken in the event of an incident Methods of giving information will vary. . Guide to Emergency Planning. Further guidance on emergency plans HS (G)25. Advance and regular information can be given to those resident or working in the area. Housing and Planning Act 1986. 4. CIA. 14. 6. 1986. Notification of Installations Handling Hazardous Substances Regulations 1982. 1 March 1987.E.Hazardous installations and the law. Health and Safety at Work etc. 82/501/EEC CASSIDY. 5. 6 CONCLUSION The UK has set in place a system of major hazard controls which should help to prevent major accidents and minimise the effects of any that may occur. 8. 12. HMSO. OJEC Directive No. 1984. Hazardous Materials Emergency Planning Guide. HMSO. those in control of public amenities can be similarly informed. CIA. HMSO. 3. HMSO. Advisory Committee on Major Hazards. 2. However. 1984. 11. K. National Response Team Report NRT. 9. Eurochem’ 86. REFERENCES 1. 1977. SIESO. as will frequency. 7.Chem. On. Adequate and relevant information is therefore a prerequisite for control and response in an emergency situation. Control of Industrial Major Accident Hazard Regulations 1984. I. General guidance on emergency planning within the CIMAH Regulations for chlorine installations 1986. HMSO. 1979. CIMAH Regulations 1984. 13. Department of Environment Circular 9/84. transients may well receive the information only in an emergency situation. Three Reports: 1976. Recommended procedures for handling major emergencies (and supplement). HMSO. Act 1974. The controls are a combination of discrete but interdependent elements. HMSO. Town and Country Planning (General Development) and (Use Classes) (Amendment) Orders.and off-site emergency planning are essential parts of the overall system of control. 10.

chlorine: 100ppm/10min: severe effects C1·667 /(min)=104 (Dicken ‘fatal’ level) C2·75 t=3·2×106 (HSE criterion) Consultation distances For example: . 16. HMSO. buildings repairable e. & DAVIES. PURDY.g. Eng. G.C. Chem.. P. A guide to the CIMAH Regulations 1984 HS(R)21.I.206 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 15. G. APPENDIX 1 Criteria Heat Blast Toxic 300 kJ/m2 (e.Chem. Toxic gas risk assessments: the effects of being indoors.g. 30 kW/m2 for 10s): severe burns 200 kJ/m2: burns 2 psi (14 kN/m2): some severe injuries 1 psi (7 kN/m2): injuries. 17. I.Toxic gas incidents: some important considerations for emergency planning. DAVIES. & PURDY.E. (1986). January.C. 1986. P..

300 kJ/m2 is dangerous (e. delayed local.TECHNIQUES FOR EMERGENCY PLANS 207 APPENDIX 2: FLAMMABLES (a) Criterion: effects of thermal radiation For a few seconds exposure.g. 20 kW/ m2 for 15s) (b) Application: Basic assessment (LPG) (c) Other variables Ignition may be immediate. torch: immediate and escalation Fireball: immediate hazard (and domino effect) . delayed remote. or none Immediate ignition Small leak Medium leak Vessel burst Local fire: escalation hazard Jet.

208 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (d) Examples (i) Fireball effects (propane) (ii) Large cloud Ranges to LFL (m): (iii) Vapour Cloud Explosion (VCE) Violent. large release. partial confinement: explosion . short delay to ignition.

g.): like TNT M(TNT)=(Expl) × Efficiency e. i.TECHNIQUES FOR EMERGENCY PLANS 209 (vi) VCE (Propane) For hydrocarbons: M(TNT)=M(half contents)×0·42 M(H/C)=2×flash fraction—vessels (v) Other explosions Solids (ammonium nitrate. 300 ppm Fatal cnt=A (or probit) . sodium chlorate. 14 kPa at 850m (e) Mitigation? Escape Shelter APPENDIX 3: TOXICS (a) Criterion: Chlorine toxicity Immediately fatal. distance/(m) 1/3) e. 50 te sodium chlorate gives 14kPa (severe damage) at 400 m 4000 te ammonium nitrate gives 80 kPa (devastation) at 300 m. 500 ppm Very quickly fatal. etc.e.g. Sodium chlorate: Efficiency=(l/4)Ammonium nitrate=(1/8) TNT Blast: related to scaled range.

210 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (b) Application: Basic assessment (chlorine) (c) Mitigation? Shelter Escape Wind direction (d) Other substances .

1 Management information and decision support systems Development of managerial decision support systems has been pursued separately in two schools. there is a dramatic development within electronic information technology and. decision support systems for operating crews during plant disturbances and accident control. and for support of the general emergency management organization. widespread efforts to exploit this technology in the design of systems for support of systematic risk analysis. 2. and consequently there is an increasing potential for severe accidents. normative decision making strategies based on objective economic .ANDERSEN & J. At the same time.RASMUSSEN Risø National Laboratory. we will briefly view the general development of decision support systems. focusing on the formulation of rational. 2 REVIEW OF THE STATE OF THE ART OF DECISION SUPPORT SYSTEMS As a basis for discussing the use of information technology in support of emergency management. Denmark 1 INTRODUCTION The current trend in the industrial development is towards large. This in turn creates an increasing demand on methods for systematic risk analysis and. one based on a management science perspective. 4000 Roskilde. centralized production units. in the case of release of the accident potential. quite naturally.25 Decision Support Systems for Emergency Management V. means for effective emergency management.

A major class of proposals for decision support systems has been based on decision making research rooted in economic theories. System science approach. His conclusion is that both approaches are too schematic and drawn to unacceptable extremes. A general criticism of this approach has been that the formal models based on economic or decision theories fail to appreciate the complexity of the challenges under which real-world decision makers must operate. a more integrated view of the system has recently been evolving. Recently. and that a more balanced view should be taken. However. top-down approach to the design of management decision support systems has been taken by system scientists. respectively. another based on a social science perspective and focusing on the social system and considering the roles and needs of the people in the system. and there will be no structured way to plan a functional system design. considering what are generally labelled ‘management information systems’ and ‘decision support systems’. in particular the expected utility theory developed by economists and mathematicians. He compares the approaches taken by the two schools based upon management science and social science. His discussions relate to business decision making. Social science approach. an attitude which has caused considerable controversy. These lines of development will be briefly reviewed. but on what they should do. Critics of decision theory also argue that it is not useful as a guide because human beings do not behave in accordance with the fundamental assumptions of the theory. An illustrative example is the discussion presented by Sutherland [1].212 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS analysis of the problem domain. Management science approach. have been considered alternative paths to a solution. This means that there will be no formal basis for evaluating the performance of such a system. the perspective of the social science is primarily concerned with the characteristics of the decision makers and their social roles. The two approaches. the only basis for judgement will be user-acceptance. a more integrated. respectively. The approach focuses on decision making from a prescriptive point of view only. It is a logical structure for decisions and makes no claim that it represents or describes the information processing of human decision makers. due to increasing understanding of the cognitive aspects of decision making. The emphasis is not on what they do. but the conclusions are well related in the present context. Whereas the management science approach is focused on the problem characteristics. In doing so it . which therefore will be based on bottomup integration of the requirements of the individual activities.

(2) Strategic analysis at the next lower level includes contingency planning related to stochastic-state techniques to provide for deductive techniques for problems the ‘state’ outcomes of which are variable. etc. attention shifts to the instrumental capabilities they imply in terms of a collection of decision aids.g. ideal-type or categorical) as well as context specific properties’.e. This technique underlies most classic military contingency planning. Support in this function is essential for executives who are responsible for development of the policy over the long run. whether or not it is requested by the existing management authorities. and will require different tools for effective support Four levels of decision types are identified and correlated with decision processes and support models: (1) Goal programming and long range planning at the highest level are related to the sequential state model for heuristic problem solving procedures or structured decision making procedures. Then the set of all primitives is reduced into a prime set. Given this prime set. The basic idea of this system theoretic approach is that any properly conceived management support system should include tools for all of these levels. All . Sutherland emphasizes the need for a structured design methodology: 1. 3. 4. in order to remove redundancies.TECHNIQUES FOR EMERGENCY PLANS 213 is necessary to take into consideration that decision making in the different levels of organization cannot be covered by one theoretical model. i. The next step is an attempt to reduce a population of functionally abstracted decision requirements to their most fundamental constituents. (4) The lowest levels concerned with the operations management. This is so. such as econometric methods. parametric decision theory. (3) The tactical programming. includes ‘equilibrium maintenance’ mainly based on statistics-based decision and control instruments for dealing with probabilistic problems. 2. to decompose into elementary operations or primitives. This is the domain of methods of industrial engineering and operations research. The first step is to identify ‘a population of decision requirements that is derived by examining organizations in aggregate in terms of universalistic (e. one level further down. based on discrete-state instruments which are primarily algorithmic and analytical methods that allow optimal solutions of deterministic problems. such as game-theoretic models or logical analysis programs.

correlation. probabilistic. and indeterminate problems. Now a prime set of system facilities is generated. Therefore the tools for the different levels in an organization will be different. The rationale for this solution will be to ensure that organizational decision problems get all the precision and discipline they deserve. equifinal. Four levels of problems are considered: deterministic. 2. or AI algorithms). and sequential state. The prerequisite for this concept will be that the analytical procedures or techniques underlying a support system are congruent with the nature of the problem at hand. Expert systems. Any of the higher-order decision requirements should thus be able to be met by synthesizing in effective real-time the functions pertinent to the integral decision aid.214 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS integral decision aids or model base components are now decomposed into their lowest order transformational components— the microfunctions which are the basic elements of ‘any structured model-base’. Optimal tools are then to be found in the diagonal of the representation.2 Expert systems: artificial intelligence approaches While the approaches to decision support systems mentioned above are predominantly problem driven. This congruence is discussed with reference to a generic problem/ instrument domain. Also four instrument categories are used: discrete state (operations research. to have a mutually exclusive set which in aggregate should be able to perform all the functions associated with the set of decision aids from which they were derived. stochastic state (contingency planning). but no more. while choice outside the diagonal will be either ineffective (insufficient) or inefficient (too sophisticated) for the purpose. 5. finite state (statistical decision theory. 6. . In the present context the term ‘expert system’ is used in the ‘classical’ sense to characterize a decision support system based on heuristic rules derived from experts and intended to support a well defined decision maker having a uniform set of decision tasks within a bounded information context. industrial engineering. regression). the solutions based on artificial intelligence approaches are by nature tool driven.

use of ‘expert systems’ for support of the decision making process ‘on-line’ seems to be premature. Hayes-Roth et al. In order to be accepted by a user. nuclear magnetic resonance. of which only few are in actual. More differentiated approaches have been taken to the design of decision support systems. and other such data to infer molecular structures — Mycin: for medical diagnosis of infectious blood disease — Expert. advice from an expert system in a risky decision context will require a more elaborate explanation capability than is presently available (see.TECHNIQUES FOR EMERGENCY PLANS 215 Recent reviews of the historical development of expert systems [2] focus on expert systems for application in domains of very uniform characteristics. The present expert systems are laboratory ‘demonstration’ systems. However. as well as location and identification of hazardous . Likewise. air traffic control. when AI techniques have been considered tools in a design effort based on analysis of the problem requirements. has been proposed by Thorndyke [4]. This proposal will be reviewed in some detail because distributed problem solving appears to be an important feature of emergency management. Thorndyke describes a system for model-based situation assessment and planning based on expert system architecture. — Explanation of their reasoning process is frequently silent on fundamental issues. such as: — Dendral: for analysing mass spectroscopic. A system oriented approach to design of a system for support of distributed decision making. for instance. etc. Applications are described for military strategic planning. Rasmussen and Goodstein [3]. based on the tools made available by artificial intelligence research. From this review. Other artificial intelligence approaches. — They have no independent means of checking whether their conclusions are reasonable. Caduceus: for other domains of medical diagnosis — Prospector: for geologic survey support. serious use. AI tools for organization of the distributed data base available to emergency management may be feasible. [2] have formulated that today’s expert systems typically show up badly when measuring along a number of dimensions: — They are unable to recognize or deal with problems for which their own knowledge is inapplicable or insufficient.

and should be considered in more detail for future developments. the ‘world model’: a sensor. — The information needed for decisions may stem from a large variety of sources. and decision making will have the nature of a cooperative effort in a distributed system. such as engineering textbooks. requirements for analysis supplying data in order to have proper data attributes and formats compatible with user needs . and instructions. analysis of prior accidents. — Several organizations and technical services may be involved. a plan generator.3 Decision support in emergency management The present problem of information systems for emergency management appears to be characteristic in the following respects: — The problem domain is poorly denned. For the organization of these activities the Hearsay-II paradigm is used [5]. the ‘cooperating experts paradigm’ is used. — Support from the system may be relevant during dynamic emergency situations. To model the organization of time stressed situation analysis and planning. information retrieval. and a controller. The resources to consider in emergency control may belong to different technical service fields. A number of experts are organized around a common data base. The conclusion of this review is that the structure offered by the Hearsay system concept for communication and coordination in a distributed group of decision makers appears to match the needs for data base support in emergency management. a communicator. inhomogeneous data bases. being dependent on the size and nature of the actual case. risk analysis. procedures. an evaluator. — The decision maker(s) are difficult to identify in advance.216 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS chemical spills. laws and regulations. as well as for planning purposes. 2. caused by very different physical processes. The system should support decision making related to a large variety of emergencies. Key problems for system development will therefore be to consider: — Organization of large.

The first domain of an analysis which will serve to bridge the gap between the purely technical description of the work content and the psychological analysis of user resources should represent the functional properties of the system in a way which makes it possible to identify the control requirements of the system underlying the supervisory task. and should include an analysis of the decision task and the information processing requirements in terms referring to human cognitive functions. of the control and decision task. In general. when designing systems for support of decision making. . This is an analysis in technical systems terms and will result in a systematic and consistent representation of the problem space. will be an important area of development for advanced information technology. By use of a multi-facet description system it is possible to represent a great variety of conditions by a rather low number of categories in each domain. the problem is to design systems which are also effective during situations which have not been foreseen during design. The problem domain. related to general features. the following dimensions of a conceptual framework for description of a cognitive task have proved useful for the analysis of cognitive tasks. and which are not familiar to the user. For design it is necessary to structure the great variety of real life work conditions into domains which correspond to design decisions. and the structure of the communication network involved — The nature. in general terms (covering typical situation scenarios). the approach to the design of a decision support system based on new technology should be taken from a cognitive point of view. 3 A FRAMEWORK FOR ANALYSIS AND DESIGN OF DECISION SUPPORT SYSTEMS In consequence of the discussion in the previous section.TECHNIQUES FOR EMERGENCY PLANS 217 — Analysis of the organization of the cooperative decision making. as well as of the requirement for the information formats used by the information sources. and hence for design of decision support systems. and the related information needs At present it appears very plausible that a coordinated data base and a consistent specification of the information needs of the various decision makers. From this point of view.

an appropriate representation of the problem space should reflect the varying span of attention in the part/whole dimension. An analysis in this problem domain can serve to identify the information processing strategies which are effective for the different phases of the decision sequence in order to identify the required data. It is generally found that a given cognitive task can be solved by several different strategies varying widely in their requirements as to the kind of mental model and the type or amount of observations required. Change in representation along both dimensions is normally used by decision makers in order to cope with the complexity of a decision task [6. The next domain of analysis to consider is related to the decision process which has to be applied for operation upon the problem space. 7]. finally. Mental strategies and heuristics. planning of resources and. . control structures. is a resource management problem. evaluation and choice of goal priority. The decision sequence. and the varying level of abstraction in the means/and dimension. and processing capacities.218 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS Since decision making in emergency management. 4 IMPLEMENTATION FOR EMERGENCY MANAGEMENT SUPPORT It follows from the preceding section that the most important domain of analysis for emergency management will be the problem domain and the decision task. Cognitive control domain. execution and monitoring. It is generally accepted that the decision process can be structured into a fairly small number of typical decision processes representing the various phases of problem analysis and diagnosis. as in many other contexts. including the role and cooperation of several decision makers. the form of the displays should be selected from consideration of human cognitive control mechanisms. While the information content should be included in the messages from a decision support system from an analysis of problem space and mental strategies.

the change in system properties represented is not merely removal of details of information on the physical or material properties. When moving from one level of abstraction to the next higher level. In other words.TECHNIQUES FOR EMERGENCY PLANS 219 4. At the lower levels. these properties are represented by concepts which belong to several levels of abstraction. suitable for transfer . and so forth. the representation of the relationships controlling the state of affairs in the emergency management context.e. i. i. chemical. Models at higher levels of abstraction are closely related to a specific purpose which can be met by several physical arrangements. Change of level of abstraction involves a shift in concepts and structure for representation. Emergency management can be considered a resource management problem in a means-end hierarchy representing the functional properties of the environment. models at low levels of abstraction are related to a specific physical world which can serve several purposes or violate different goals. More fundamentally. The lowest level of abstraction represents only the physical form of the system. In this hierarchy. In man-made systems these higher level principles are naturally derived from the purpose of the system. Above this.e. its material configuration. Thus an observer asks different questions to the environment depending on the nature of the currently active internal representation. information is added on higher level principles governing the cofunction of the various functions or elements at the lower level. the functional properties are represented in more general concepts without reference to the physical process or equipment by which the functions are implemented. from the reasons for the configurations at the level considered. The next higher level represents the physical processes or functions of the various components and systems in a language related to their specific electrical. as well as a change in the information suitable to characterize the state of the function or operation at the various levels of abstraction. elements in the process description match the component configuration of the physical implementation.1 Problem domain The first aspect to consider will be the problem domain. Therefore shifts in the level of abstraction can be used to change the direction of paths. or mechanical properties.

Domain of potential risk. 1.FIG. 220 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

FIG. Domain of emergency management resources. 2. TECHNIQUES FOR EMERGENCY PLANS 221 .

e. and decision making in a specific situation will be a resource management task aiming at a proper relationship in the potential many-to-many mapping between the levels. the domain of the potential risk.2 Domain of potential risk This part of the representation includes information identifying the potential risk sources. A property of the total emergency management system considered at an individual level can be characterized in three different ways: (1) ‘what’ it is. its causal properties in interactions at that particular level. and equipment/personnel which is available to form the counteracting and mitigating force. 1. and the domain of mitigation resources. and (3) ‘how’ it may be implemented by resources at the next lower level. processes. its role at the next higher level. 2. It represents the problem space for the planning part of the representation. For the emergency management systems. i. This means that . and analysis of the technical features of prior cases.222 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS of knowledge from previous cases and problems. technical manuals. the information related to the decision space will be discussed for two separate categories. (2) ‘why’ it may be chosen. i. This part of the data base will supply the basis for the analytical part of the representation.4 Use of problem representation This representation of the problem space will be a multi-level representation in terms of available/required equipment-processfunction-purpose elements.3 Mitigation resource domain This domain includes the information about functions. Examples of the information at the various levels may be seen in Fig.e. The information included at the various levels can. and the information will be available from risk analysis. 4. be as shown in Fig. and the possible higher level consequences in relation to social norms and legal rules. 4. for instance. 4. their functional physical properties making it possible to predict the accidental propagation of effects of accident releasing mechanisms.

This advisor can be a human domain expert or an ‘expert system inference machine’ attached to the data base. If procedural transformations are incorporated in the data base. and the use of information technology should be considered not only for advice giving to the expert system. 6). This. unless very specifie information can be supplied. Decision making in a particular situation will be an iterative consideration of the resources at the various levels until a satisfactory relationship through the levels has been identified. and the boundaries of his information needs in terms of location in the problem space chart (see Fig. Pejtersen’s work on information retrieval in libraries [8]).TECHNIQUES FOR EMERGENCY PLANS 223 the data element in a data base should be characterized from at least three different points of view. depends on the identity of the actual decision maker. it will have to be rather general rules. The form in which the information should be stored in the data base depends entirely upon the users’ formulation of their problems and needs (cf. in turn. 5). or information on the actual state of affairs will have to be transmitted to the advisor in possession of the necessary general background knowledge or the intermediary working on the available data bases (see Fig. goals and constraints with the available physical resources. If the procedural information has to be generated on-site. This will involve the task of keeping track of a many-to-many mapping in a complex net. as well as upon the hierarchical structure of the operating organization. The data base representing the problem domain in terms of risk potential and emergency management resources will include structural information about functional properties and causal relationships which must be transformed into procedural information in order to be operational in the actual accident situation. possibly conflicting. . The nature and the related sources of information to be included in a data base should be specified for each of the cells in the domain abstraction/decomposition matrix (Figs 3 and 4). connecting the various. it will either have to be done by the decision maker himself. but also for support of the decision process itself (for instance by alerting the user to consider other relevant means-end mappings than the one behind an actual information request). This transformation can be based on heuristics derived from prior experience or deductions based on state information from the case actually present.

FIG. Information sources. domain of potential risk. 3. 224 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

domain of emergency management resources. TECHNIQUES FOR EMERGENCY PLANS 225 .Fig. Information sources. 4.

Information users.FIG. 5. domain of emergency management resources. 226 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS .

6. and to specify the format in which information should be supplied by .TECHNIQUES FOR EMERGENCY PLANS 227 FIG. The conclusion of this preliminary analysis is that the meansend hierarchy is well suited to structure the information content of the data base which is underlying emergency management decisions. during preplanning as well as during the actual situations. Thus structured. it will be possible in a consistent way to identify the proper search terms to use for retrieval design. Problem domain in emergency management.

and must be able to cope with a wide variety of accidents. plant design. One is to analyse the present emergency management organizations . smaller scale accidents related to fires. where a top-down approach has been taken by analysing the requirements needed to satisfy the specified goals for an emergency management system. A short description of the content and status of this programme will now be given. a state-of-the-art review of models and methods available for construction of a conceptual system [10]. The other group includes emergency management related to accidents in hazardous industrial installations for which emergency organizations have been carefully planned and for which risk analysis typically has been made. The subjects addressed in this project led to a preliminary description of accident and emergency scenarios [9]. etc. e. while most nuclear power-plant emergency situations will be located in the lower right of the diagram. Figure 7 shows a schematic representation of different types of emergency situations. A decision support system for the latter is being developed as a joint Nordic programme NKA/INF.1 NKA/INF project content The basis for the study of the potential use of advanced information technology for accident and emergency management was established in a pilot project undertaken in 1985. Most non-nuclear emergency situations will be located in the upper left of the diagram. has two lines of development. The programme. toxic spills. such as risk analysis. 5. expert systems [11]. Outside the diagonal the task will be either insufficient or ineffective.228 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS the numerous data sources. One group includes the rather frequent. incident analysis. and a review of available tools from artificial intelligence. The emergency management organization is established ad hoc. Nuclear power installations are typical for this category.g. at present. where the focus of optimal support for a given situation is found in the diagonal of the representation. and inspections. operations planning. 5 EMERGENCY MANAGEMENT IN NUCLEAR AND NON-NUCLEAR INDUSTRIES The problems involved in industrial emergency management appear to fall into two rather distinct groups.

to evaluate the problems perceived and the possibility of remedy by means of modern information technology. 7. concurrent line of approach has been to establish models of the distributed decision making involved in operations like emergency management. For concerted activity communication between the decision makers is necessary.TECHNIQUES FOR EMERGENCY PLANS 229 FIG. in order to evaluate whether advanced information technology will influence the effective way of organizing. The programme consists of five main activities: (1) The study and detailed analysis of accident and emergency scenarios based on records from incidents and drills in nuclear installations (2) Development of a conceptual understanding of accident and emergency management with emphasis on distributed decision making. Another. Different types of emergency situation. The approach taken to such a model may be to consider decision making a control task involving a number of decision makers each controlling only part of a loosely coupled problem space. and control structures that are involved (3) Development of a general experimental methodology for evaluating the effects of different kinds of decision aids and forms of organization for emergency management systems with distributed decision making . and procedures. information flow.

and it must be limited to the extent where the prototype development is possible using the available resources. The general point of departure for the conceptual work has been to design a framework for analysing different kinds of emergencies. Such organizations were found to present problems under certain conditions because [12]: — all kinds of emergencies cannot be foreseen. The prototype system will experience a dynamic development throughout the major part of the project. a ‘vertical slice’ is identified dependent primarily on two criteria: it must be able to display the major features of the conceptual system. Conceptual work. This will be reflected in the recommendations and guidelines developed in the final phase of the project. In the later phases of the project the scenario descriptions will gradually change to data and knowledge acquisition. and applications of expert systems and methods from artificial intelligence (5) Production of guidelines for the introduction of advanced information technology in the organizations based on evaluation and validation of the prototype system. . In the first stage. 5. but in the present context only the status of the conceptual work will be described.2 Status of the NKA/INF programme The programme has developed conceptually.230 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (4) Development and test of a prototype system for a limited part of an accident and emergency organization to demonstrate the potential use of computer and communication systems. and this may create a need for a more flexible structure with the capacity to reconfigure itself. The keyword for the project is system studies with emphasis on system integration. and the conceptual work will be followed by development of a general experimental methodology and by experimental work using the prototype as test bed. In an early stage of the project a limited target area must be defined. and in prototype implementation. in data acquisition and specification of data and knowledge base. we have been concerned with the problems of hierarchical organizations in emergency management. data base and knowledge base technology. Based on the scenario descriptions.

which [13]: — provides a clear specification of the goals of an emergency management system. we have tried to create a general framework for analysing emergency management based on the view of emergency management as a control system. 2. be controlled in emergency management. The problems here are those of coordination in a system characterized by distributed decision making. some first thoughts on how the decision support system should be evaluated have been looked into. and (c) the knowledge required for understanding these displays . Here a distinction between two forms of evaluation has been discussed: analytical evaluation and empirical evaluation. — specifies the information needs. — provides a specification of what the components of such a system should be. This comprises two steps: — Mapping the decision support system onto a set of general decision tasks — Assessing the extent to which these tasks are supported by analysing (a) the nature of the situation. and — specifies what can. and what cannot. — some aspects of emergency management cannot be modelled hierarchically but require a different control structure. This is done in an attempt to test the general usefulness of these diagrams as an analytical tool for analysing information needs in emergency management. In the second stage. In addition. and — hierarchical command and control systems are not needed for all kinds of emergencies. Further work is now directed towards solving two problems: 1. It is recommended that an analytical evaluation be performed first.TECHNIQUES FOR EMERGENCY PLANS 231 — information delays would make it hard to exercise control by means of a hierarchical system that would be too slow. To develop a conceptual framework for those aspects of emergency management that cannot be controlled hierarchically. Using the time-area diagrams developed as part of the analysis of emergency management as a control system to analyse a variety of emergencies. (b) the kind of displays that are provided.

232 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

It is also recommended that the empirical evaluation be directed
towards limited and well defined functions of the decision support
system. DESSY-D, a general interactive program for simulating
dynamic systems, is being developed for this purpose in Uppsala.
The methodological problems in using this system for the
evaluation of a decision support system are now being analysed.
6
CONCLUSIONS AND RECOMMENDATIONS
The conclusion of the present feasibility study will be that the
recent development of advanced information technology, together
with the trend towards more cognitively oriented approaches to
studies of decision making, offer promising lines of development of
improved tools for emergency management. Such improvements
will be necessary in order to cope with the increasing potential for
unacceptable consequences of accidents which is the result of
industrial centralization together with the widespread use of
hazardous substances. In addition, a reconsideration of the
information basis of emergency management will be relevant now
because much information of great importance for emergency
management will be generated or collected from activities such as
systematic risk analysis, safety inspections, quality assurance
programmes, etc.
REFERENCES
1.

2.
3.

4.
5.

6.

SUTHERLAND, J.W. (1983). Normative predicates of next generation
management support systems. IEEE Trans. Syst. Man. Cybern., 13
(3), 279–97.
HAYES-ROTH, F., WATERMAN, D. & LENAT, D. (1983). Building
Expert Systems, Addison-Wesley, Reading, MA.
Rasmussen, J. & GOODSTEIN, L.P. (1985). Decision support in
supervisory control. 2ndIFAC/IFIP/IFORS/IEA Conf, on Analysis,
Design, and Evaluation of M an-Machine Systems, Varese, Italy, 10–
12 September.
THORNDYKE, P.W. (1982). A Rule-based Approach to Cognitive
Modelling of Real-Time Decision Making, ORNL/TM-8614.
Erman, L.F., HAYES-ROTH, V., LESSER, V. & REDDY, D. (1980).
The Hearsay-II speech understanding system: integrating knowledge
to reduce uncertainty. Computing Surveys, June, 213–52.
RASMUSSEN, J. (1985). The role of hierarchical knowledge
representation in decision making and system management. IEEE
Trans. Syst. Man. Cybern., 15(2), 234–43.

TECHNIQUES FOR EMERGENCY PLANS 233

7.

8.

9.

10.
11.

12.
13.

RASMUSSEN, J. (1985). A Framework for Cognitive Task Analysis,
Risø-M-2519. Also in: Hollnagel, E., Mancini, G. & Woods, D. (Eds),
Intelligent Decision Support Systems in Process Environment, Springer
Verlag, Berlin, in press.
PEJTERSEN, A.M. (1980). Design of a classification scheme for
fiction
based
on
an
analysis
of
actual
user-libarian
communications, and use of the scheme for control of librarians’
search strategies. In Theory and Application of Information Research,
O.Harboe & L.Kajberg (Eds), Mansell, London, pp. 146–59.
JOHANSSON, R., ANDERSSON, H. & HOLMSTROM, C. (1986). A
Descriptive Analysis of the Management of Nuclear Power Plant
Emergencies, Studsvik Technical Note NI-86/7.
RASMUSSEN, J. (1986). A Cognitive Engineering Approach to the
Modelling of Decision Making and Its Organization, Risø-M-2589.
BERG, Ø. & YOKOBAYASHI, M. (1985). Review of Expert System
Techniques Relevance to Computerised Support Systems in
Emergency Management, Institutt for Energiteknikk, INF-630(85)1.
BREHMER, B. (1986). Organization for decision making in complex
systems. Unpublished note.
BREHMER, B. (1987). Emergency management as a control system.
Unpublished note.

SESSION V
Lessons Learnt from Emergency
Management of Major Incidents
Chairmen: M.VASSILOPOULOS
Ministry of the Environment, Greece
U.POLI
Istituto Superiore per la Prevenzione e la Sicurezza del lavoro, Italy
Rapporteur: E.L.QUARANTELLI
Disaster Research Center. USA

26
Experience Gained from Recent Major
Accidents in the Federal Republic of
Germany
STEPHAN NEUHOFF
Berufsfeuerwehr Köln/Cologne Fire Brigade, Cologne,
FRG
1
FIRE IN THE CHEMICAL FACTORY KALK IN
COLOGNE
On 8 September 1982 at 11.02 a.m. the factory fire brigade of the
Chemical Factory Kalk (CFK) notified the professional fire brigade
in Cologne by radio of a fire in the factory.
The CFK is a chemical factory which produces and stores
nitrogencontaining fertilizers. Several smaller plants also produce
organic and inorganic bromine compounds. The factory is located
in the central city area in the direct vicinity of residential areas.
The bromine plant No. II is a steel-framed skeleton structure
with thin asbestos-cement walls. It covers an area of 450 m2 and
is 15 m high. It contains apparatus for producing brominecontaining flame-protecting substances for plastics. Methanol and
other alcohols are used as cleaning fluids.
During the refilling of a 2000 litre plastic container, about 150
litres of methanol leaked from a faulty pipeline. The methanol,
which is electrically conductive, caused the generation of sparks
on an electricity cable which led to an ignition process.
The fire rapidly spread through the entire production plant and
threatened a neighbouring storage area with barrels of solvents
and other chemicals as well as an additional production plant.
The professional fire brigade and the CFK factory fire brigade,
comprising a total of 109 firemen, used 6 water cannon and 5
lines to restrict the fire to the bromine plant and to extinguish it
after 2 hours.
Dense smoke moved over the neighbouring residential areas.
The area was closed off by the police. Radio broadcasts were used
to request that civilians remain inside and close all windows and

236 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

doors. As bromine in the smoke was suspected, an evacuation
seemed to be necessary.
A meeting of the Disaster Prevention Management, directed by
the town clerk, was held in order to prepare first measures.
Measurements then indicated however, that no halogens were
present in the smoke.
The bromine plant was completely destroyed. The damage was
estimated at approximately 20 million DM. The plant was not
rebuilt.
The following conclusions were arrived at:
1. The plant had been approved 6 months before the incident.
The plastic container, however, was not approved.
2. Cooperation between the professional fire brigade and the
CFK was excellent. A definite advantage was the fact that a
joint exercise had been held in the factory area 5 months
before the incident, in which a similar scenario had been
assumed.
3. The police as well as the fire brigade had set up separate onsite Technical Operation Management Groups. As liaison
officers were not exchanged immediately, an optimum
coordination of the protection measures was not possible. In
future a joint command centre should be established.
4. Measurements and analyses were made by the fire brigade,
the industrial inspection board and by the factory. There was
no centralized control and evaluation of the measurements.
2
LEAKAGE FROM AN LPG TANK IN
WILHELMSHAVEN
On 23 January 1985 a gas tank in the Mobil Oil AG refinery in
Wilhelmshaven was being filled with pressurized liquified butane.
During the filling procedure it was noticed that liquified gas was
being emitted at the top of a support brace and a visible cloud
drifted past several other storage tanks towards a nearby river.
The spherical tank had been constructed in 1975. Its diameter
was 19m and its volume was 3500 m3. The authorized operating
pressure was 6·6 bar. The wall of the container was 18 mm thick
in the support brace area.
The construction of the tank was officially supervised. It was
then subjected to construction and pressure tests as well as a final
acceptance test. External inspections were made every 2 years. The

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 237

inside of the tank was inspected every 3 months by refinery
personnel. No faults were detected.
The refinery area was closed off immediately. Shipping
movements on the river were also stopped. The liquified gas was
pumped into a nearby tank. Fortunately ignition of the gas did not
occur.
The incident was caused by a crack in the tank. During the
construction of the tank, the tops of the hollow support braces
were at first covered by welding on a cover sheet, in order to
prevent rainwater from entering the support brace. This cover
sheet was also welded directly onto the spherical tank. This
however, led to strong tensions so that the welded joint was
separated again. Instead of being welded, the cover was bolted on
with clamps, and a rubber gasket was inserted between the cover
sheet and the support braces.
This gasket obviously did not seal properly and rainwater
entered the support braces. Due to a separating sheet inside the
support braces, 1 m3 of water was sufficient to fill the tops of the
support braces. This water then froze during a frost period. The
pressure of the ice pushed in the tank wall at the tops of the
support braces. At one of the tops an 860 mm long and up to 15
mm wide crack was formed.
The crack was not noticed earlier due to the intense cold in
January. The pressure in the tank was only slightly higher than
atmospheric pressure. Gas was only emitted when the tank was
filled with liquified gas.
The incident could have been avoided by drilling holes in the
separating sheets and in the lower ends of the support braces, in
order to let trapped rainwater flow out.
3
EXPLOSION IN THE RHEINISCHE OLEFINWERKE
IN COLOGNE
On 18 January 1985 at 15.47 hours (3.47p.m.) the control centre
of the professional fire brigade was notified that an explosion had
occurred in the southern area of the city. At first it was assumed
that a tanker had exploded and caught fire in the harbour.
The control centre sent several fire-fighting vehicles to the
harbour. Telephone calls to a refinery and investigations from a
helicopter led to the conclusion that a fire was burning at the
Rheinische Olefinwerde (ROW). At 16.07 hours (4.07 p.m.), 26
minutes after the explosion, the ROW finally called to report the
explosion.

238 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

The ROW is a large chemical factory, which uses the products of
a nearby refinery to produce ethylene and plastics such as
polyethylene and polypropylene.
The accident occurred in the distillation section of an ethylene
plant; 4t of propylene leaked from a cracked pipe. The gas
emission was registered by the gas warning system, but the
immediate emergency shut-down of the entire plant could not
prevent the ignition of the gas cloud by a heater. The pressure
wave caused extensive damage and several fires started in the
plant. In the area surrounding the factory, windows were
destroyed up to 9km away, 43 employees were injured by
splinters, but in a nearby residential area only one person was
slightly injured.
The ROW factory fire brigade and the professional fire brigade,
with a total of 104 firemen, used 24 water cannon and up to
60000 litres of water per minute to cool the plant. The fire could
not be extinguished completely. Remaining gas was allowed to
burn itself out. The fire brigade operation was only completed 9
days later. The plant has since been rebuilt. The damage
amounted to 100 million DM.
The gas had escaped through a crack in a 100mm diameter
pipe. During the investigations an identical pipe was filled with
water and cooled. The ice which formed did not lead to any cracks
in the pipe. Only after a second test with the same pipe did it
crack due to being over-stressed in the first test.
Before the accident it had been generally accepted that the
propylene did not contain any water. The investigations, however,
revealed a water content of 1 ppm. At a production rate of 130
000t per year, 70 litres of water accumulated in the pipe. Two
periods of severe frost during the winter of 1984/1985 then
caused the damage.
The following conclusions were drawn:
1. The ROW took much too long to report the accident. After
this incident the city of Cologne reached an agreement with 7
chemical factories, with the aim of much quicker reporting of
fires and accidents.
2. The extensive water supply network with 257 installed water
cannon permitted rapid cooling of the plant.
3. A world-wide investigation of comparable plants was
undertaken to check for possible water accumulation in lowlying pipeline sections.
4. The suspended ceiling in the plant control station collapsed.
However, it was still possible to shut down the plant. Special

When the truckmounted pump was switched on. A safety analysis was undertaken during the reconstruction of the plant. it was snowing. The truck driver. . immediately switched off the engine and called the fire brigade. Dut to the dense fog and ice formation. 5. so that water had to be transported by fire brigade tank trucks. with great presence of mind. but still surrounded by houses. The sewage system was flushed with large amounts of water. at 13. The operations chief of the fire brigade immediately evacuated the nearby houses and had the central heating systems switched off. The gas cloud spread out and was visible as a 1 m thick fog which covered an area of 500m2. and it could not be sealed. the drive shaft to the secondary drive came loose and damaged the connecting pipe between the tank and the pump.m. In the field into which the gas had been moved. Finally a water curtain from 4 lines was used to push the gas cloud into a low-lying field. At the edge of the village. The propane which was pressurized at 9–5 bar immediately started escaping through the leak. and the temperature was approximately 0°C. 52 hours (1. On 23 December 1986.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 239 explosion protection measures were included in the new control station. the driver of a liquified gas tanker truck was filling a fixed liquified gas tank. the lower explosion limit could only be measured at a distance of 60m from the tank truck.52 p. An attempt to seal the leak by spraying water on it failed due to the high pressure. it was at first difficult to determine the exact position of the leak. The vehicle contained 4–81 of propane. In the meantime the entire electricity supply to the village had been cut off by the electric supply company. After 1 hour the gas concentration in the centre of the village dropped to values below the detection limit. The factory concluded that the worst possible damage was the failure of a support brace which would result in a 2 m long flame.) a slight east wind was blowing. This also resulted in a shut-down of the water supply system. The danger of igniting the gas stopped any attempt to tow the truck out of the village. 4 LEAKAGE FROM A LIQUIFIED GAS TANKER IN DIELHEIM The village of Balzfeld with a population of approximately 1000 is part of the municipality of Dielheim.

The pressure shock wave caused extensive damage to the exterior of the three-storey hotel and to nearby buildings. Flushing the air in the houses was therefore felt to be unnecessary and the inhabitants could return to their homes. At 15. opened and large amounts of liquified gas escaped. connected the gas tank with the heating apparatus in the heater room in the cellar of the hotel.00 hours (4.) on 27 December 1986 the safety valve on the gas tank. Gas had invaded both underground floors through windows and ventilation openings. had escaped without being ignited. A strong smell of gas was reported to the hotel reception and the hotel personnel tried to evacuate the hotel. For environmental protection reasons it was heated with liquified gas which was stored in a covered underground pressure tank. a stronger pipe. The tank heating system was constructed as a closed secondary system. . Two over-pressure safety valves were installed on the tank. The tank heater was regulated by a pressure controller which switched off the pump for the warm water cycle as soon as the tank reached an over-pressure value of 4 bar. or some type of protection for the pipe between the tank and the pump could have prevented the incident. the gas pipe and the inlet and outlet pipe for the warm water heating system.m. swimming pool). The highest value reached only 10% of the lower explosion limit. Three pipes. The lower part of the tank could be heated by a warm water pipe system. The tank contained 64·1 m3 and was constructed for a permitted filling of 85% at an operating pressure of 15·6 bar. which was 54% full.) ignition occurred. approximately 2400 m3. in the heating room and in the underground garage.m.00 p. squash courts.240 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS As soon as the over-pressure in the tank truck had disappeared. Measurements were then made in all nearby houses. the tank was flushed with nitrogen until gas could no longer be detected at the leak. 5 LIQUIFIED GAS EXPLOSION IN A HOTEL IN GARMISCH-PARTENKIRCHEN The Hotel Riessersee with 350 beds had been opened in May 1985.45 p. The police and the fire brigade had not yet arrived at this time. At 16.45 hours (3. A different pipe location. Most of the damage occurred in the fitness centre (sauna. A total of 4·8 t of propane.

. During manual operation. The public prosecutor prohibited any alterations to the system. Six tanks had exactly the same configuration as the one in Garmish-Partenkirchen. A total of 23 heating systems were shut down immediately. Eight people were injured. the pressure controller was switched off. The gas was left to burn for 4 days. A total ban of heating systems on liquefied gas tanks is therefore being considered. as increased amounts of gas can be removed through an evaporator. The remaining liquefied gas was then pumped out by a tank truck. however.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 241 Seven people were killed instantly in the fitness centre. The investigation of the heating system for the tank showed that the pump for the warm water cycle could be controlled either automatically or manually. The explosion had ignited the gas escaping from the safety valve. The dome shaft was therefore flooded in order to cool the fittings. and 4 died during the following days. The warm water could continue heating up the gas tank until the safety valve opened and gas escaped. An additional 26 tanks were found to have either no pressure limiter or no temperature limiter. There is no technical need for a heating system for liquefied gas tanks. When the system was operating automatically the pressure controller on the tank was switched on and controlled the pump. Only 56 out of a total of 40 000 liquefied gas tanks were heated with warm water. After this accident a special investigation was undertaken in Bavaria.

USA 1 INTRODUCTION There appears to be general agreement that the number of accidents. explosions.e. Delaware. University of Delaware. the Disaster Research Centre (DRC) in 1977 began a 4year study of socio-behavioral preparations for and managing of chemical disasters. as well as the management of. little attention has been given to the behavioral features of the problem. incident was a public manifestation of what many observers have known has been a growing increase of problematical risky events in the chemical area. However.QUARANTELLI Disaster Research Center. i. . To begin to close this gap in knowledge.L. fires. systematic and comparative data on preparedness were obtained from 19 communities in the United States that had varying degrees of risk due to dangerous chemicals. spills. The Bhopal. In the first phase of the study. In the second phase of the research. DRC examined organizational and community preparedness planning for. or other acute chemical threats. response to sudden dangers resulting from hazardous chemicals. Newark. In 45 field studies. disasters and catastrophes involving dangerous chemicals has been increasing in the last decade or so. This study was the first systematic and largescale effort of its kind undertaken by social scientists. India.27 Community and Organizational Preparations for and Responses to Acute Chemical Emergencies and Disasters in the United States: Research Findings and Their Wider Applicability E. DRC studied 26 managements of responses to major emergencies or disasters that resulted from toxic releases. the human and group aspects. Considerable technical research has been undertaken on the handling of hazardous chemical occasions.

g. as part of a series of field studies on organizational functioning in crisis occasions. In this paper a general overview is presented of these findings. it does take later work into account. with special attention being given to emergencies and disasters resulting from transportation accidents. Currently. while this paper is primarily a summary presentation of the first systematic research. including the phosphorous spill from a train derailment in Dayton. There later field studies and analyses have been used to test and to extend some of the observations and conclusions that were drawn from the initial large-scale research. as well as different substantive foci of the study. in 1986 and another similar spill in the Pittsburgh metropolitan area. A general report on the full study has been given [8]. is contained in the publications previously cited. Thus. the specifics of which have been reported in publications elsewhere [1–7]. Louisiana [9]. Since that initial research. However. obtained primarily through intensive interviewing of key personnel and collection of documents. Two explosions were separately studied in field studies: (a) a chemical tank explosion in 1982 in Taft. a series of official reports on chemical incidents was recently systematically examined (e. More specific information about the methodology and theory. the bulk of the paper reports what DRC found in its studies of response management. and (b) a major catastrophe outside of the United States: the liquified petroleum explosion in November 1984 in the Mexico City metropolitan area.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 243 The on-site data in both phases of the study. for other purposes. which was concluded in 1981. were subjected to a variety of quantitative and qualitative analyses. We also recently undertook a comparative analysis of transportation accidents that involve phosgene gas versus those that involve dangerous nuclear wastes [10]. Massachusetts. In addition. Ohio. DRC has done additional work on chemical disasters. We will first briefly summarize what we learned about preparedness planning for chemical threats. DRC has also looked at seven more chemical incidents. 418 people were injured and there was a forced evacuation of a 1·5 square mile area which contained 23 000 inhabitants). the report on an incident in Somerville. where in 1980 a cloud resulted from a spill of phosphorus trichloride as a consequence of a train accident. .

2 Availability and mobilization of resources In principle. That is. This variability in perception may partially be the result of role expectations as they apply to these different sectors of the community. particularly in the public sector. Many tangible resources either are unknown. are unrecognized as such. therefore.1 Threat perceptions There is a degree of perception that chemical agents. there are noticeable differences between threat perceptions of public and private groups. There is little collective mobilization of resources except in a minority of communities with local comprehensive mutual aid systems (i. with the latter seeing chemically based disasters as less likely than the former. or are the property of private groups. 2.e. However. sectors. and organizations selectively vary in their perceptions of chemical threats [11]. have more potential as disaster agents. compared with other agents. This type of role expectation can sensitize these groups to the various demands of their domains. networks of relevant organizations from both the public and private sectors that form for the express purpose of sharing resources in disaster preparedness and response). Such systems are particularly strong with respect to resource sharing . many public sector groups (such as fire departments) have official responsibility for emergency preparedness and are expected by the community to carry out these responsibilities. and a lack of leadership and responsibility for their availability prevails. there are many potential resources available to prepare for chemical emergencies and disasters. and. different communities. More intangible resources are also undependably and unevenly available. but not in fact. On the other hand. In particular. even when available tend to be segregated inefficiently from other kinds of community disaster resources.244 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 2 RESEARCH FINDINGS ABOUT DISASTER PREPAREDNESS PLANNING 2. fewer private sector groups (with the exception of chemical companies) have formal responsibility for preparedness planning and. are less likely to be aware of disaster threats in general.

This reinforces a disinclination to disturb local economic benefits from chemical plants or to argue against what is seen as a public unwillingness to spend governmental funds for almost anything. most do not. in providing a role for the medical area.e. The general pattern. and in addressing the problem of evacuation [12]. . There is a tendency to believe that communities could respond to emergencies and disasters better than they probably do. formal or informal contacts between and among organizations and groups) for chemical preparedness planning in most of the communities we studied. even though the collective resources of the latter sources are extensive in nature. 2. but overall evidence shows a pattern of weak community social organization for chemical emergencies and disasters. values. is one of weak vertical rather than horizontal linkages within communities. although they are usually weak in risk assessment. the structure tends to be hierarchical in nature. That is. 2. however. There is also an almost total absence of local extra-community linkages. including disaster preparedness planning. More integrated linkages are slowly evolving. the social climate in most local communities in the United States is not favorable to preparedness planning. and beliefs provide incentives for planning.3 Patterns of community social organization A variety of social linkages were found (i. Extra-community resources are seldom part of any individual or collective preparedness planning for the mobilization of resources for chemical disasters. While some of the existing norms. there tend to be links between local fire departments and the chemical companies in their areas.4 Social climate As a whole. In particular.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 245 and communication. with authority vested in the uppermost levels and with few provisions for effective crosscommunication among the various disaster relevant groups.

However. both the chemical industry and United States governmental agencies have initiated a variety of programs aimed at improving local community preparedness for chemical accidents and disasters. practices. documents. much is happening with respect to chemical disaster preparedness planning in the United States in the last few years. in most localities. We should observe that.e. over the long or short term. Community disaster preparedness for chemical problems is generally poor.5 The planning process and preparedness Only a low degree of preparedness planning for chemical emergencies and disasters exists in most communities in the United States. Nonetheless. formal and informal agreements. Extra-community groups that do have resources for chemical crises are seldom incorporated into local planning. the private sector is relatively well prepared. However. it comprises all the activities. An additional impediment to local planning efforts is the fact that the most relevant resources rest in the hands of extra-community groups (i. In fact. it tends to make for a better response to chemical emergencies and disasters. if not nonexistent. are intended to reduce the probability of disaster and/or the severity of the community disruption occasioned by its occurrence.246 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 2. The effectiveness of this planning and its contribution to the better management of hazardous chemical incidents has not yet been documented. to the extent that preparedness planning of any kind exists. we would . an end product of the planning process. realistic process stressing general principles aimed at reducing the unknowns in a problematical situation. state and federal level organizations) rather than with the local community organizations that invariably are confronted with problems associated with the immediate post-incident response. or viewed as an extension of everyday operations. Preparations for chemical disasters are especially handicapped by the public-private sector split in the United States. with the exception of some fire departments. Almost certainly this preparedness planning will make the situation better than it was. good preparedness is actually a knowledge-based. Preparedness is often incorrectly equated with formal disaster plans. while the above observations reflect our field studies in the last decade. especially for in-plant accidents. As such. Partly triggered by the Bhopal catastrophe. and associated social arrangements that. such planning is frequently non-existent among public emergency organizations. however.

the threat from developing. Also. In the disaster area. such as the plant fire squad. Fixed-site incidents. differ somewhat in the two types of situations. barges. in-transit accidents will. not massively and quickly. The differences in the managing of the two types of crises are the result of a variety of factors. There were some major differences in the patterns of response to hazardous chemical incidents that occurred at fixed sites compared with those that resulted from an accident that occurred while a vehicle was in transit. Which organizations participate in the response to the crisis and what they do. such as a plant. as well as the difficulties that emerge. and that occur on publicly accessible lands.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 247 suspect that what we report from the past will generally be what will be found in future studies. there are enough differences in the responses to make them worthwhile noting. rather than the fire department of the local community. as most other areas in life. if not to prevent. in-plant chemical emergencies tend to lead to actions to contain. Fixed-site situations generally are those that occur in chemical plants or on their property. trains. or aircraft carrying hazardous chemicals. usually generate responses that are specific to the particular chemical hazard involved. such as the local police and fire units. In . are likely to involve only company-related groups. In contrast. Chemical plant incidents in the United States almost always occur on private property.1 Fixed and in-transit sites. Although there are many common elements between the two types of crises in the United States. on the other hand. For example. emergencies that occur at a fixed site. usually quickly. 3 RESEARCH FINDINGS ABOUT MANAGING RESPONSES TO CHEMICAL DISASTERS 3. improvements tend to occur incrementally and slowly. often initially trigger general accident response measures rather than specific chemical disaster responses. In contrast. such as those that occur at a plant. In-transit accidents. such as those that involve trucks. evoke the appearance of community emergency agencies. In-transit incidents are the result of transportation accidents. many of the initial activities in in-transit accidents are devoted to measures to protect the community.

but most efforts of this kind were unsuccessful. There is a tendency to equate accident preparedness with disaster preparedness. this does not always occur. not only is there likely to be less preparedness planning for intransit accidents. between responses to fixed-site accidents and responses to in-transit accidents probably are the result of other factors. Moreover. responses to chemically threatening incidents are better when the accident occurs in a fixed facility than when the accident occurs in transit. Also. ships.248 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS contrast. In summary. For example. even if an incident is an accident that is not a disaster. Unless the accident is of major magnitude. In contrast. Larger companies are more likely to have detailed and extensive preparedness planning for chemical mishaps. Although incidents beyond a certain level of impact are supposed to be reported to the public authorities. Chemical companies generally have good emergency preparedness programs. or planes carrying dangerous chemicals are involved in a transportation accident. any incident in the United States that may lead to the pollution of any body of water could lead to the activation of the national contingency plan for such events and the active participation of the US Coast Guard. the mobilization of resources to alleviate the accident will probably help alleviate the potential for a disaster occurring. tank trucks. regardless of local and state plans and the activities of community and state agencies. when level of risk for an accident to . but there are more problems that must be coped with in transportation related events. usually it is difficult to prevent the community from finding out about the accident. however. This is related to the low social visibility of incidents that occur at plants. There are often complicated jurisdictional questions and multi-level organizational issues when trains. most (although not all) in-transit accidents are more socially visible. The major differences. In our study we discovered some attempts to maintain secrecy about hazardous incidents in railroad yards. in-transit accidents. only the workers and officials immediately present in the plant may know that there has been a chemical mishap. Often minor mishaps in chemical plants are so well handled that they never develop a potential for becoming a disaster. however. and the extent of preparedness is usually related to the size of the company. even though they may involve a private carrier. especially if the plant is part of a nationwide or international corporation. usually occur in what normally is viewed as a public setting.

according to our study it appears that the potential for the occurrence of catastrophic chemical disasters compared with the potential for occurrence of non-catastrophic incidents is greatest in fixed installations. that determine whether there will be a minor nonchemical mishap or the threat of. it appears that the locations that have the greatest risk of occurrence of a chemical catastrophe or major disaster are those where better preparedness and response measures are likely to be found.2 First responders The importance of the initial response in a chemical emergency is widely recognized. That is. In part this results from the relatively little systematic chemical disaster preparedness planning for accidents that occur on roads or highways. In incidents that occur inside chemical plants there usually is no lack of understanding that a hazardous chemical is involved. One major American chemical manufacturer has produced a safety training film entitled Those Vital First Minutes’ to emphasize the necessity of proper and quick actions during the period immediately following a chemical mishap or an accident that involves chemical substances. Motor vehicle incidents are least likely to result in catastrophic accidents. Our study did not obtain enough information to form a conclusion about the potential for the occurrence of chemical catastrophes as a result of barge-ship and airplane accidents. The next most vulnerable type of accident is that involving railroads. However. better preparedness for accidents generally exists in plants that produce the most dangerous and greatest volume of hazardous chemicals. It is often the actions taken in the first few minutes. In general. On the other hand. a chemical disaster.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 249 occur is considered for different modes of transportation. There are many factors that can affect the magnitude of the possible danger in an incident. Thus. our study found that motor vehicle incidents are generally handled less efficiently and effectively than those occurring on railroads. railroads have undertaken far more elaborate planning for chemical threats. or actual occurrence of. 3. just before a release or just following a spill. it is in such locations that the quickest and most efficient initial responses to a chemical mishap are likely to occur in the United States. We observed in the study . a far more problematical situation usually exists in the early stages of an in-transit mishap.

The problem with misperception of the initial situation is compounded in that organizational and community disaster plans rarely discuss the combination of a transportation accident and a hazardous chemical incident. In principle. there is a tendency to consider all cues in terms of normal or expected events. they seldom initially activate the disaster plans of their organizations. that its own study showed that required placards were in place on only 77% of the railcars. Motor vehicle or train accidents are initially seen only as transportation accidents or wrecks. responders are acting in a way that has long been observed in the disaster literature. (This excludes situations in which placards and symbols had either been destroyed or were made illegible as a result of the transportation . and even more rarely do they activate the plans specifically for chemical disasters. Unfortunately. The view that placarding requirements are often widely ignored is supported by the observations of our study. A DRC content analysis of plans determined that separate consideration of the two types of events was almost universal. in another unpublished report from a railroad. they are not automatically recognized. even when placards and symbols are in place and readable after an accident. first responders should be aware of the various placards and symbols that are mandated by law in the United States to be carried on tanks and other containers of hazardous materials. One systematic study of trucks in Virginia found that 41 % of the trucks stopped for inspection were violating placard requirements for hazardous materials [13].250 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS that in transportation accidents first responders seldom initially perceive a dangerous chemical threat unless there are obvious sensory cues. they do not always fully understand their meaning. If an occurrence appears to be a transportation accident. it will be perceived and defined as a transportation accident. various studies have determined that the legal requirements are not always followed. It is stated. that is. This is true even when first responders are from emergency organizations such as fire or police departments. such as a strong pungent odor or eye and skin irritations. Our study revealed that first responders do not always note the signs that identify hazardous materials. In doing this. One consequence is a tendency for responding groups in transportation accidents to initially use their routine accident standard operating procedures. The general tendency of first responders is to define the situation as it appears to be on the surface. even if aware of them. namely a transportation incident. However.

injured.) Also. one survey found that 23% of trucks carrying hazardous materials failed to carry required shipping papers [13]. the relevant papers are not always carried on the vehicle. If the papers are found. Also. and waybills did not fulfill their informational needs. even if a search is initiated. Thus. seldom have the knowledge to read technical papers correctly. In a US National Transport-ation Safety Board hearing. for all these reasons. Also. that is involved. responders learned about the hazards long after the incident was over. first responders seldom have easily accessible manuals or booklets that would define the symbols or indicate how they should respond to the incident according to the type of dangerous chemical substance. Some of the DRC observations on these matters have also been reported by others. Of course. and frequently in accidents that involved multiple dangerous chemicals. or disappear from the accident scene.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 251 accident. thus precluding questioning by first responders. However. they are not always understandable to people without an appropriate technical background. Sometimes first responders to transportation incidents do initiate searches for invoices or other relevant papers. computer printouts. placards. Moreover. It was rare in the chemical emergencies that resulted from a transportation accident for first responders to learn quickly what they had to face. They stated that all too often placards located on hazardous materials tank cars were destroyed. especially operational personnel. There have been cases in which first responders have been unintentionally misinformed by truck or train personnel about the dangerous cargoes that were being carried. such personnel do not always know exactly what type of goods the vehicle had been carrying. Personnel from the transporting carrier are sometimes killed. usually the first responders to transportation accidents. Personnel from law enforcement agencies. in some instances. and in immediate . the knowledge of the train crew was limited as to the exact placement of tank cars and the materials carried. first responders are frequently uncertain about the specific nature of the chemical threat even after they suspect that the incident is more than a routine accident. it is sometimes difficult to find the invoices or shipping bills for the material that is being transported. witnesses from the fire service area indicated that reliance on technical manuals. it was observed in the study that personnel from the carriers were sometimes reluctant (if not actually uncooperative) to provide relevant information to first responders. identified by the placard.

They may recognize that the community is possibly endangered and that some chemicals may be involved. We observed more than once that company personnel often failed to report promptly to outside authorities fixed-site accidents that involved chemicals. such as a toxic cloud. this is the general principle stated in the disaster literature [15]: faced with responding or not responding to an uncertain threat. Not infrequently. there seldom is a problem of identifying the chemical threat. the function of rumor behavior is to provide some definition of a situation when none is otherwise readily or officially available [16]. However. Incorrect identification may be diffused to many others through rumor among local officials outside of a plant or near the site of a transportation accident. This failure to communicate existed even when the threat expanded or continued to develop outside of the plant grounds. the outside community agencies did not find out about a chemical threat until there were obvious sensory cues.252 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS emergency conditions there was not adequate time to search for waybills and cross-reference materials with an emergency manual to determine general emergency actions [14]. A few situations were observed in which an evacuation was initiated even though the community did not officially know the nature of the danger from which people were being evacuated. there are other kinds of problems that result from the typical behavior of first responders to fixed-site accidents that occur in plants. . but they have no specific knowledge beyond these impressions. As students of rumor phenomena have stated. All efforts by first responders to identify the exact nature of the chemical threat in transportation accidents are beset by a number of difficulties. correct identification of the chemical involved by the first or early responders sometimes does not occur. although in one case it took company officials hours after an explosion to realize they had a poisonous gas episode potentially present in the situation. in contrast to in-transit accidents. the latter course of action is most likely to be followed. it is understandable that the responders from outside of plants often remain unclear for some time about the specific nature of the chemical threat. In the face of a very unclear and uncertain threat there is likely to be a delay in doing anything. Because it is known that a danger exists does not necessarily mean that the exact nature of the danger is understood. As previously noted. Given such circumstances. In accidents that occur in chemical plants in the United States. We noticed in our study that community emergency officials often learned by chance about the possible danger to their localities.

To identify something as a threat does not automatically mean that there is knowledge about the specific nature of the threat or how to handle it. They usually lack the appropriate equipment. the volatile reaction that will occur if water is combined with calcium carbide. Trained teams normally do what should be done. fire departments are not well prepared to respond to most sudden chemical incidents. it was observed in some chemical incidents that.g. but explosive. In general. or corrosive threats that might result from other chemicals involved in the transportation accident are overlooked. Their traditional routine of quickly putting water on a blaze tends to be done automatically. Our study also found that first responders to transportation accidents tend to overlook two important and dangerous possibilities. if a fire is perceived or if one chemical is identified as capable of burning. even when a chemical threat is correctly identified. perhaps surprisingly. unfortunately. they often do not know where to turn for . this is focused on. even if they correctly identify the dangerous chemical and know its effects. Thus. First responders to chemical incidents often literally do not know what to do. responders to on-site accidents generally do not recognize the different and various kinds of multiple hazards that might be present because of a variety of dangerous chemicals on the same train or truckload. Thus. an equivalent recognition of the specific dangerous nature of the threat was not always known. e. fire department personnel (most likely the first responders to the danger) may not act appropriately. In the DRC field work. in some instances this can be one of the worst things to do. Thus. it is possible for mistakes in judgment to be made. First responders tend to be oriented to the existence of a single chemical agent rather than a multiple chemical agent. Moreover. materials. asphyxiating. however. even when the identification of the chemical substance was correct. In almost all cases there is an initial overlooking of possible synergistic effects. and protective gear. direct observations were made of trained company emergency response teams who acted incorrectly and endangered themselves and others. Trained personnel also may act inappropriately. In addition.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 253 Hazardous chemicals may have varied and multiple effects on human beings and on the ecology of the environment. with the exception of some in large communities and other special cases. given the complex nature of dangerous chemicals and the various contingencies involved. The lack of widespread knowledge about correct stabilization and neutralization procedures is especially significant at the local community level.

the nationwide chemical emergency reporting center. A discussion of the possible contingencies is presented in the next section. DRC discovered more than one fire department that had personnel who had never heard of CHEMTREC. but they are not necessarily doing something relevant to the problem at hand. especially in transportation accidents. Many of these weaknesses in coping with chemical incidents result from the primarily volunteer nature of the staffs of the nearly 30 000 fire departments in the United States. These contingencies can be . As the nature of the chemical threat becomes clearer. as well as the degree to which they respond. This generally facilitates action being taken by the organizations. We observed in field work during our study that some emergency organizations have relevant technical manuals available. There is often a delay in defining a transportation accident as one that has the potential to be a chemical disaster. and frequently. there is considerable variation in the use of such manuals.3 Impact and situational contingencies Different types of contingencies can influence the way in which a community will respond to a particular chemical threat. Therefore. Defining what is happening and what can and should be done is a large part of the early response. Moreover. Although the situation has been changing rapidly in recent years. the manuals are not consulted at the height of the emergency.254 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS information. A major observation of the DRC study was that the initial responding activities of emergency organizations usually follow standard operating procedures. A vast majority of first responders do not have experience from a similar situation that they can rely on. but such definitions are not always correct. This is in part because there can be many contingencies present in a potential disaster situation. there usually is a tendency to try to adjust to the newly recognized situation. Yet it is these volunteer groups that are often among the first responders and that usually are the lead organizations in fighting hazardous chemical threats in transportation accidents. they are often inaccessible to the first responders. Trying to clarify the situation is often a prime activity. There is an ad lib quality to the pattern of the first response. as mentioned earlier in this paper. For example. 3. relatively few local personnel have had training in dealing with hazardous chemicals. experience in responding to any unusual emergency in the past is likely to influence the response to the current situation. however.

but most are not. a few chemicals can explode. 3. the opposite is stressed in this paper. However. To a considerable extent.3. Given the variety of characteristics that might be involved. in that responders to the crisis may not correctly perceive either the damaging and destructive potential or the controllability of the chemical threat. some chemicals are toxic. Thus. Different chemical agents generate different risks and threats. Nevertheless. certain chemicals only become dangerous when they combine with other chemical substances. The situation is complicated. myriad possibilities of risk could be present. . Both of these characteristics will have implications for the manner in which responders to an incident can and will attempt to neutralize the threat.1 Impact contingencies Impact contingencies include those characteristics of the chemical agent that can affect the organized response. of course. the potential consequences of the risk still remain. we do not argue for the importance of idiosyncratic factors. even though the managing of a chemical incident and its effectiveness will be affected by differences in the chemical agent’s impact characteristics as well as by variations in the social aspects of the particular situation. there are dimensions of risk that are inherent to the chemical agent. many of these variations can be reduced to one of two types of possible consequences: the damaging or destructive potential of the chemical or chemicals. other chemicals remain inert. what we shall be discussing as the tactical problems posed by contingencies are what often appear to an unsophisticated disaster planner or operational emergency worker as idiosyncratic or unique in a specific hazardous chemical threat incident. others cannot.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 255 divided into two categories: impact variables (or chemical agent variables) and situational variables. aspects which appear to be idiosyncratic when observations are made of only one or a few cases turn out to be more general features or happenings when enough incidents are observed. In fact. However. the specific characteristics of the chemical agent or agents involved in a major accident will influence the risk and threat to a particular environment. For example. even if they are incorrectly perceived. and the ability to control the chemical or chemicals. While risk assessment essentially involves a perceptual component.

which are both highly reactive and have a tendency to polymerize. there can be a tremendous difference in threat or impact of a chemical accident. Unlike in many natural disasters. In another emergency the risk may be extremely high. This great variation in possible damaging destructive potential is an inherent agent contingency in a threatening chemical situation. all of which can be further affected by meteorological . again there is often substantial variation in the damaging or destructive consequences. toxicity. In both of the situations previously noted. In general. responders may be presented with different contingencies that are primarily dependent on the inherent properties of the type of chemicals that are involved in the accident. in one emergency the responders might be faced with a relatively low-risk situation. Chemicals that have a high-risk potential are exemplified by the inherent dangers of compressed gases or the hazards posed by gases such as butadiene and vinyl chloride. be more than a threat of impact—there can be actual impact. and synergistic possibilities. unless it occurs within the confines of a chemical plant. however. There can. One result is that multiple exposures to chemical risks may not provide a good learning experience that can be used in other emergency situation. Thus. Certain agents have a greater potential for damaging results than others. Those managing a chemical threat can be faced with widely differing dangers depending on which chemical or chemicals happen to be involved.256 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The damaging destructive potential of any chemical agent is the amount of damage and destruction it can do to people and to the ecological environment. DRC studied some actual chemical incidents in which populations that were dozens of miles away from the actual disaster site were endangered. usually has little idea of the destructive potential of such substances. experience in one chemical disaster does not necessarily transfer well to the next incident. Chemical properties of an agent include flash point. Those managing a localized disaster are presented with operational and response problems different from those faced by the responders to a diffused disaster. depending partly on inherent qualities of different substances. vapor density. The typical first responder (whether police or fireman) to a chemical accident. the high-risk chemicals are those that are extremely volatile or that exhibit an unstable molecular structure. Thus. Yet other chemical disasters were examined in which the actual destructive impact was confined to the part of the truck or railroad tank car involved in the accident.

however. Usually. these representatives will often attempt to exercise authority and control in the situation. independent of the perceptual factors. the contingency of the damaging destructive potential of any chemical agent may influence the coordination of inter-organizational response. While the same perception exists for most natural disaster agents. Thus. If a disaster is large enough to necessitate a response from state. While both destructiveness potential and uncontrollability of the agent are inherent to the properties of the chemical. insofar as response is concerned. there is also the factor of the uncontrollability of chemical agents. Finally. This often complicates jurisdictional problems because there are often discrepancies in responsibilities among different governmental sectors. the magnitude of which partly depends on inherent properties of the chemical or chemicals. or federal level of government.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 257 conditions such as precipitation. a chemical’s controllability is only partly dependent on the properties of the chemical agents. regional. different dangerous chemicals provide different threat of actual impact contingencies to which those managing the disaster must react The magnitude of a disaster can also complicate the response pattern. The results of our study suggest that there is . the belief is sometimes expressed that this should not be the case for chemical substances. everything else being equal. This is in addition to the possibility that responders may have incorrectly perceived the chemical danger or even not perceived any threat at all. a number of representatives of agencies from different jurisdictional levels will respond to the event. controllability is partly dependent on the community’s ability to perform certain initial response tasks. We usually have more involvement of state and federal organizations. and other similar factors. the greater the volume. Here too there may be considerable variation between the inherent uncontrollability of a chemical agent and the responder’s perception of this uncontrollability. wind velocity. Perceptual differences aside. or some combination of levels. In actuality. In addition to potential or actual destructiveness. Controllability also depends on the amount or volume of the chemicals. sometimes over the opposition of local officials. Our study determined that most community officials are likely to assume there is a high degree of uncontrollability in most chemical agents. they are not. as well as on the capability of the community to respond appropriately in the critical period of time immediately following the onset of an accident that has a potential to be a disaster. In a large-scale disaster. the greater the uncontrollability.

258 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS misunderstanding with respect to both destructiveness potential and uncontrollability. In turn. That is. if any. and there may be resultant fires and explosions due to the derailment. They are non-specific in this respect. the risks posed by dangerous chemicals are not restricted to certain localities or regions of the country. or tornadoes are specific to certain localities. the threat presumed to exist due to a chemical emergency often exceeds the inherent possibilities of most chemical substances. they are relatively random in their manifestations of hazard. in the derailment of a train carrying dangerous chemicals. Probably one reason for a general misunderstanding of the potential effects of chemical agents is that. the derailment is a problem that must be solved. Although chemical agents are widespread throughout American society. it is unlikely that any given population group will have had much. . few people have any experience in viewing chemicals and certain risks associated with technological accidents. these may create a chemical spill or toxic cloud that might not otherwise have occurred from the derailment alone. the image of the risk presented by chemical agents is vague and tends to be exaggerated. most natural disaster agents such as earthquakes. but they are seldom major threats acrossthe-board. In general. Therefore. Consequently. In some actual chemical disasters. but our study showed that the reverse is often the common view. hurricanes. In some acute chemical cases there are often multiple elements of a disaster occurring either concurrently or sequentially. except within the chemical industry. In contrast. Impact contingencies add to the possible variation and complexity of the response in chemical incidents. the perception that chemicals are involved in an accident often leads to a perception of danger. Chemicals can present major risks and result in major consequences. For example. Most chemicals are not inherently dangerous. the situation is further compounded for those managing the event by the multiplicity and variety of hazardous aspects that may be involved. direct experience with dangerous chemicals. As in projections of risks at nuclear plants. community officials and the public tend to overestimate the damaging and destructive potential of dangerous chemicals.

For example.2 Situational contingencies Situational contingencies include those specific characteristics of the particular social context in which a chemical mishap first occurs. when chemical accidents occurred inside plants or chemical company property. the response was delayed and confused because no local agency believed it had exclusive responsibility for. time or circumstances affecting the response to. (a) Variations in location. A chemical problem also occurs at a specific point in time—more accurately. the incident. at some social time in the community life. The location at which a chemical threat or disaster occurs significantly affects the response. ranging from the degree of knowledge the public will have about the event to the possible courses of action that responding organizations can take. we observed during our research that. there are particular circumstances associated with each chemical emergency. situational contingencies will be discussed that can be classified as variations in location. the overturned truck carrying a dangerous chemical cargo may or may not have displayed the required warning placards or signs. Situations were studied in which. A chemical incident does not just happen. public. a chemical incident. A chemical incident.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 259 3. in a place with distinctive features. can occur on private property. it happens in a particular locality. for example. There were also situations in which local fire departments were denied entry onto private property on which a chemical emergency was occurring. Thus. and jurisdiction over. or a public location. These possibilities have implications for a variety of factors. for instance. Such a lack of clarity over response initiative would not occur in a private setting. a mixed public-private setting. the larger community seldom found out quickly about such events unless there were immediate casualties.3. the location (actually property responsibility) and whether that property is a private. Another locational contingency involves the geographic and demographic setting of incidents. An obvious possibility that may . In the following subsections. Likewise. or private-public responsibility (which is a contingency) have an effect on the patterns of managing chemical emergencies. or the managing of. because the chemical emergency was in a public setting. In nearly every case there was a delay between the time that the accident on private property was turning into a potential disaster and when this happening became public knowledge.

and domain are unclear and often overlapping. (b) Variations in time. manifestation of the complexities that can be generated for responding organizations by impact contingencies. the day of the week. if not competition or conflict. although the same difficulties also frequently surface outside of city boundaries. Furthermore. Many rural or quasi-rural areas in the United States are locales where organizational responsibility. These two types of time are not equivalent. ambiguities can surface about who has the major responsibility for managing the disaster. These patterned activities vary (and not always directly) in relation to the time of day. the contingency of the location in which a chemical emergency occurs can have a major impact on the managing of the response. Thus. there are community social phenomena such as the rush hour. The inherent destructiveness of the chemical agent might not differ. authority. An accident that might have only minor consequences in a rural area could have potentially catastrophic consequences in an urban area with high population density and heavy concentrations of buildings. Thus. because many jurisdictional boundaries and domains are often vague. there is a rhythm to social life. chemical disasters that occur in port areas or that involve bodies of water appear to generate jurisdictional problems in the response. A chemical incident in such a location is certain to elicit interagency confusion. and the season. For example. In every community. However. if an emergency occurs near the uncertain boundaries of two or more separate jurisdictions. and thus a single situation may involve multiple disaster potentials that generate different demands to which the affected community must respond. we frequently noted in our research that interjurisdictional and interagency problems may arise. but it could vary depending on the geographic setting in which the destructive agent manifests itself. it is not chronological time but social time that creates an effect. Moreover. Therefore. the water needed to douse the fire might actually trigger a dangerous chemical reaction that otherwise would not occur.260 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS affect the pattern of response is whether the incident occurs in a rural or urban setting. In particular. depending on the location in which the chemical incident occurs. but not uncommon. The time when a chemical threat or disaster occurs also has an important effect on the response. Each of these events creates different demands. with certain activities ebbing and increasing in particular patterns and cycles. the incident may generate different emergencyrelated tasks that are incompatible with each other. This example represents an extreme. major .

Canada) chemical incident. certain material resources could not be easily located and used because the organizations owning them were closed and it was difficult to find any personnel with relevant information on how the resources could be obtained or the authority to do so. do not have either the same quantity or quality of personnel available at all times. Some chemical incidents were studied in which the response developed slowly because higher level emergency officials were not immediately available because the incident occurred outside of regular weekday working hours. and most emergency groups are on a 24-hour basis. In a few cases. variations in time can create different contingencies. evacuation is easier to carry out when it is light than when it is dark. (c) Variations in circumstances. In addition to contingencies due to location and time.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 261 sports events. For example. In our research. days. Thus. The chemical risks might be identical in two chemical emergencies. depending on the time at which the incident occurred. the rhythms of community life (or social time) can create significantly different situations with which responders must cope. as well as the state of readiness of emergency organizations and how quickly resources can be mobilized. There may be other circumstances affecting the situation. Even organizations that operate on a shift basis. and holiday weekends [17. a railroad derailment may produce no chemical toxic release for several hours. or perhaps not at all. 18]. responding organizations must maintain site security and mobilize certain resources for the duration of the episode. but because of the time at which the accidents occur there could be somewhat different situations for the responders and managers to face in the two cases. Thus. As indicated earlier in this paper. However. . some events that eventually become chemical emergencies may initially be no more than a transportation accident or a plant mishap. massive evacuation was partly delayed. At the Mississauga (Ontario. there are still other possible variations. two of these factors will be illustrated here: the duration of the threat and the speed of onset. With respect to time. Such social times affect where people will be concentrated and what they will be doing. We noted in our study that there was a significant variation in response. because of a reluctance to try to move a large number of people at night [19]. similar to variations in the location of an accident. according to police reports. chemical incidents were observed in which the response activities ranged from a few hours to nearly a week.

This realization should encourage general tactical planning that takes contingencies into account. In such cases. in many transportation accidents the initial accident does not always produce an immediate chemical emergency. circumstances can create different types of situation. For example. and information outward from the disaster site.262 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The residual polluting effects of a dangerous chemical neutralized can likewise extend the duration of an incident. Depending on many factors. and in that way circumstances partially structure the managing of the response which should occur. the response can be directed primarily at preventing a chemical emergency from developing. What flows out is even more erratic than . Speed of onset is another situational variable that may affect response patterns. Although an urban area might suffer more in absolute terms. in almost any chemical incident. the circumstances are such that the threat is over quickly. A chemical emergency of the same duration would not have the same consequences in a metropolitan area. we noted that smaller communities were more adversely affected by a prolonged emergency. things. Both the outflow and the convergence patterns are marked by much uncertainty and un-evenness of knowledge of the situation by selectively involved organizations. preventive efforts cannot be taken and the response management generally focuses on recovery efforts. including impactrelated contingencies. and rapid depletion of certain kinds of resources. However. It is easy to think of impact contingencies in individualistic or idiosyncratic terms. substantial losses to the local economy because of closed businesses. including properties of the chemical agents as well as how the potentially dangerous substances are initially treated. Much of what happens after the arrival of the first responders and their initial definition of the situation can generally be visualized as convergence and outflow patterns. depending on the kind of community in which it occurs. Among the negative consequences noted in the study were lost wages for volunteers in emergency organizations. In other cases. This can create greatly differing consequences. However. there may be little or no advance warning of an impact. we have indicated that there are some general aspects even of contingencies. In many such cases. and hours after the initial indication of an emergency there is little sign that anything happened. There is a movement of organizations. As illustrated in the examples. our observations were that smaller communities tended to incur relatively higher losses for chemical incidents of the same duration. and a similar flow toward it.

3. if it is known. and no centralized sources are available for quick references). The split hinders chemical disaster preparedness and is not helpful in chemical disaster management. there are social as well as technical aspects of preparing and responding to acute chemical emergencies and disasters. A chemical disaster can be occasioned by rather different things. medical personnel are uncertain on measures to take.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 263 what converges. and some behaviors tend to compound the difficulties in the situation and almost ensure lack of coordination. a strong technological bias exists in the planning activities and operational measures undertaken with respect to hazardous chemicals. While in one sense this is undoubtedly true. and frequently requires rather different coping mechanisms. especially in relation to very unfamiliar chemicals. preparedness planning ought to consider three aspects. This also makes for problems in preparedness and response.4 A few implications What are the implications of our study? From a general perspective. There is the strong belief that technical solutions can be found both to prevent and to soften chemical disasters. Also. how to obtain accurate information relevant to the diagnosis and treatment of victims (often the chemical agent is unknown or. From a more specific perspective. can physically have rather different outcomes. Put another way. chemical disasters are more problematical than disasters resulting from most other kinds of agents. problems would still be inherent in the group and human aspects of the situation. chemical disaster agents tend to be relatively more heterogeneous than other kinds of disaster agents. our work suggests that locally based preparedness planning using existing resources can lead to an improvement in integrated community responses. with weak linkages between the two sectors. Finally. There are also special problems in chemical emergencies with respect to exactly how to handle the often overwhelming numbers of mass media representatives. These require the application of a sociological perspective. and how to identify appropriate procedures for neutralization of the chemical threat. . Even if all the technical problems were solved. There is a major public-private sector split. which we have partly tried to illustrate in this paper.

Thus. But these kinds of differences will also manifest themselves in non-chemical types of disasters. including all the recent studies of Bhopal. There has been only limited research in other highly industrialized and urbanized societies in Western Europe and in Japan [20–22]. Our view is that the work done so far.g. outside of the research in the United States and Canada there has been little systematic research on hazardous chemical incidents resulting from transportation accidents. e. To the extent that societies vary from one to another. Additionally. Overall our position is that what has been learned so far about planning for and managing of chemical disasters should be taken as generally applicable unless future research shows the necessity of taking some distinctions into account.5 Cross-societal applications The great majority of social science studies of hazardous chemical incidents has been done in the United States. ACKNOWLEDGEMENTS Part of this research was supported by National Science Foundation (NSF) Grant PFR-7714445 but all views expressed are those of the author and not necessarily those of NSF. .g. there may be variations in social organizational structures (e. while there may be cross-societal differences along some lines with respect to preparedness planning and response managing of chemical incidents. Italy and Japan [26].264 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3. Also. until Bhopal there had been almost no study of acute chemical disasters in developing countries. there is the question of the applicability of what we have reported to all other types of societies. as found in a comparative national level study of natural disaster response in the United States. the degree of centralization of the governmental structure) as well as cultural values [25] which could affect both preparedness and response. there are more similarities than differences [23. suggests that. Even the often drawn distinction between developed and developing societies does not appear very useful for disaster research and policy purposes [27]. Some of the observations in the middle sections of this paper have been reported previously [28]. 24].

DE. Chemical disaster preparedness at the local community level. 3. 4. K. J. (Eds) (1981). Richmond. PHILLIPS.S. Newark. 457–60. J. A Primer for Preparedness for Acute Chemical Emergencies. Larkin.L. 2. In Sociological Research Symposium IX (Ed. VA. J. 343–56. Lewis. (1981). GRAY. Newark. D. QUARANTELLI. GRAY. 239–49. E. TIERNEY. pp.. Woods). Disaster Research Center. Virginia Commonwealth University. D. 4. 8. Louisiana. 12. (1981). L. Blackburg. & QUARANTELLI. Three Case Studies of Organized Responses to Chemical Disasters.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 265 REFERENCES 1. DE. Department of Sociology. 8. J. Williams). B. QUARANTELLI. & TIERNEY. 11. Newark. In Sociological Research Symposium XIII (Eds M. 321–30. University of Delaware. Disaster Research Center. Sociobehavioral aspects of chemical hazards: summary findings on preparations for and responses to acute chemical emergencies at the local community level. (1982). Virginia Polytechnic Institute and State University.L. DE.H.L. Disaster Research Center. 6. L (1984). Newark. HELMS. VA. J. (1984). J. Mutual aid systems in the United States for chemical emergencies. DE. VA. Disaster Research Center. & QUARANTELLI. GABOR. Newark. Sociobehavioral Responses to Chemical Hazards: Preparations for and Responses to Acute Chemical Emergencies at the Local Community Level. Virginia Highway Hazardous Materials Flow. Fawcett & W. J. QUARANTELLI.Hazardous Materials. Social climate and preparations for sudden chemical disasters. D. (1983). Developing a community preparedness capability for sudden emergencies involving hazardous materials. 309–94. University of Delaware.Hazardous Materials. K. (1979). Department of Sociology. University of Delaware. J. E. & PRICE. GRAY.P.L. 115–18. (1983). E. Wiley.Hazardous Materials. QUARANTELLI. QUARANTELLI. T. Honnold & J. DE. J. TIERNEY. Transportation Accidents Involving Hazardous Chemicals versus Those Involving Dangerous Nuclear Material. 4. 10.D. 13. 9. pp. & HUTCHINSON. E.Sculley & J. (1982). (1977). In Safety and Accident Prevention (Eds H. Evacuation Behavior: Case Study of the Taft. Virginia Commonwealth University. E. K. (1980).L. University of Delaware. Social aspects of acute chemical emergencies.Williams). 4 (Special Issue). 7. University of Delaware.Nelson. SCHMIDT. J. Richmond.Hazardous Materials.L.S. (1981). . NY. E. Chemical Tank Explosion Incident. Disaster Research Center. E. 5. New York. Threat perceptions in acute chemical disasters. E.

23. & PADGHAM. (1981). University of California. 26. Disaster Research Center. CA. 20. Newark. McLucKiE. Chicago. Newark. 51–62. SHRIVASTAVA. Perceptions and reactions to emergency warnings of sudden hazards.L. Disaster Research Center. 1978. 15. Bobbs Merrill. Ballinger. & QUARANTELLI. T. National Research Council.L. QUARANTELLI. 25. A.266 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 14. IL. Canada. 24. Bibliography of Publications on Bhopal. Transportation Research Board. Disaster Research Unit. Preliminary Paper No. Warning of disaster and evacuation behavior in a Japanese chemical fire. Temporal Man: The Meaning and Uses of Social Time. (1985). (1982). M. 309. The Peel Regional Police Force and the Mississippi Evacuation. GRAY. Research findings on organizational behavior in disasters and their applicability in developing countries.J. 1. Washington. J. 1986. K. 511–15. Italy. 16. New York. LAUER. Praeger. SHRIVASTAVA. Bhopal: Anatomy of a Crisis. DC. Ekistics. Historical and Comparative Series No. R. (1984). Ottawa. Cambridge. E. National Transportation Safety Board. (1977).Hazardous Materials. 17. Japan and the United States. Risk and Culture. T. K. Analysis of Proceedings of the National Transportation Safety Board into Derailments and Hazardous Materials. Canadian Police College. England. .L. Flixborough—The Human Response. SCANLON. DOUGLAS. P. 22. Industrial Crisis Institute. P. E. Oxford. (1987). (1975). & WILDAVSKY. Effects of Centralization on Disaster Responses 1964–1969. 19. Berkeley. LAGADEC. Hidden Rhythms: Schedules and Calendars in Social Life. In Recent Advances in Hazardous Materials Transportation Research: An International Exchange. (1981). (1982). Washington. ZERUBAVEL. MA. 28. IN. (1980). (1986). P. DC. WESTGATE. 21. University of Bradford. Indianapolis. New York. E. University of Delaware. (1966). E. 97–104. SHIBUTANI. J. First responders and their initial behavior in hazardous chemical transportation accidents. 7. England. Improvised News: A Sociological Study of Rumor. University of Chicago Press. Bradford. 18. M. 107. (1982). University of Delaware. B. IKED A. 27. Major Technological Risk. DE. QUARANTELLI. pp. Pergamon Press. DE.

The high sea traffic has been assigned a special international passage called Route Tango. the narrow sounds and belts. The relatively short distance between the individual coasts means that in a very short time (a few hours) a spill will wash ashore. difficult navigational conditions. Finland and Sweden). NAEP has its own pollution combatting vessels and units stationed at the Copenhagen and Korsør naval bases (the Sound and the Great Belt). her peninsula of Jutland and c. The Danish sounds and belts are among the most crowded in the world. GDR. Copenhagen. Denmark 1 INTRODUCTION With its situation between the North Sea and the Baltic Sea. 7000km. frequently bad weather and a considerable cross-traffic all present a constant risk of collision and grounding. Denmark. The total coastline is c. a large number of tankers. USSR. The national ‘maritime preparedness’ also comprises the Civil Defence Corps whose task it is to combat sea pollution in nearshore areas and if possible stop it from washing ashore. is a typical coastal nation. The forces of the CD Corps (professionals and conscripts) are stationed . Heavy sea traffic. freighters and bulk-carriers pass to and from the Baltic states (FRG. salt meadows and a few rocks). Many special conditions add to the risk of pollution hazards in Danish waters. even with favourable winds and currents.28 Experience Gained from the Pollution Control Operation at Læsø 1985 FLEMMING LIND ARPE Danish Civil Defence and Emergency Planning Agency. marsh. The National Agency of Environmental Protection (NAEP) is responsible for directing and coordinating control operations in cases of oil and chemical spills at sea. 400 bigger and smaller islands. Apart from the traffic to and from Denmark. of a highly varying nature (sandy and stony beaches. Poland.

The total cargo was 3000 tonnes of heavy fuel oil. In cases of comprehensive and complicated shoreline pollution.) NAEP at once dispatched their vessels to the area and early in the morning started air reconnaissance. 2. The local authorities are responsible for shoreline clean-up. marsh and salt meadows). (During the rescue operation later. dunes. The result was a major. In the afternoon a large oil slick was observed drifting in north-easterly direction c. 2. thus very sensitive to any pollution. Læsø also an important nature reserve with sea birds and seals. skimmers. ‘Jan’ of Bremen rammed into the Hals Barre lighthouse near the eastern entrance to the Limfjord.1 The island of Læsø The island of Læsø (112 km2) is a municipality with 2600 inhabitants and a very attractive recreational tourist area due to the beautiful and varied nature (fine sandy beaches.2 The situation On 2 August 1985. does not have sufficient resources to combat a large-scale. In the collision ‘Jan’ received a 20 m long gash in the forepeak tank and the cargo tanks Nos 1 and 2. tools etc. 6·5 . this oil was recovered on the spot by one of the combatting vessels. (local time). local government may apply for practical assistance (know-how and strike teams) from the national authorities (NAEP and CD Corps). complicated pollution of Læsø. a small Danish island in northern Kattegat. in the northern part of Kattegat (position: 56°54′ 05″N 10°30′07″E).m. another spill of more than 100 tonnes occurred. complicated oil pollution.268 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS at 6 regional barracks and equipped with oil booms. however. The first effect of the collision was the immediate spill of c. 2 THE OIL POLLUTION CONTROL OPERATION AT LÆSØ IN AUGUST 1985 One of the biggest oil spills ever in Danish waters was caused by the West German tanker ‘Jan’ of Bremen. plantations. at 02. 200 tonnes of oil. of course. These forces are trained in pollution control operations.40a. This small community.

(Denmark has not yet any remote sensing equipment. The practical work was organized and controlled by the mobile staff and communications unit.m.00 a. Only very light vehicles could be used for . the clean-up had to be done manually.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 269 nautical miles (NM) from the collision site. and try to prevent the oil from drifting ashore. The team arrived at Læsø on 4 August at 02. difficult and dirty job had begun. A vast. with their own vehicles. course 050°) and a large spread of the oil. The tidal water had also pressed the oil into the small inlets and draining canals in the area where it had settled in a layer up to 30cm thick. and communications equipment and outfit for a longer stay on the island under primitive conditions. An oil belt 60–100 m wide polluted 8 km of salt meadow.) The weather in the area had deteriorated considerably. The local government asked for central government assistance. staff. All decisions were made in daily meetings with the local authorities. and the CD Corps over the next 10 days carried out the practical work in cooperation with the local authorities. The oil was washed over the booms (the current sucked the ‘skirt’ towards the sea bottom) and it was impossible to keep them stationary in the high waves. The use of oil booms could not be considered because of extremely difficult weather conditions (strong current and wind. The next day (3 August) oil was observed drifting only 1–2NM from the coast.3 The combatting operation In the course of the first 2 days it became clear that it was not possible to prevent the oil from washing ashore. the strong current (speed 0–8 knots. 100 men from the CD columns at Thisted and Herning (Jutland). Since the oil was now actually on the coast it had become a local task to organize the shoreline clean-up. 24 NM from the collision site) was threatened. Owing to the special nature of the area (salt meadows) and its considerable sensitivity to mechanical impact. high sea and tidal water). 2. The force Was increased to c. the sea clean-up operation was very difficult. materials. Their primary task was to evaluate the situation and the extent of the pollution. Owing to high seas. It soon became clear that the south-west coastline of Læsø (c. NAEP ordered a combatting strike team (20 men) dispatched from the North Jutland CD column at Thisted.

4·5 million DKK (about 0·6 million US$).270 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS transport of the large quantities of waste (oil and polluted vegetation. The CD Corps share of this amount was c. . buckets and special ‘vegetation cutters’ was mostly used for the clean-up operation.). The dirty nature of the work made it necessary to equip the men with protective clothing (waders and gloves). Special cleaning sites were established in the area where the men daily underwent a systematic and efficient cleaning procedure whereby they were completely cleaned of oil. About 1000 birds. Cleaning of materials. equipment and workshop services). This proved essential for their working morale. 3 LESSONS LEARNED FROM THE OPERATION The CD Corps gained some general and specific experience on Læsø. stones etc. materials. (a) We learned the importance of: — having an overall. The cleaning of Læsø was so efficient that no damage was done to nature (fauna and flora). The operation was carried out with efficiency and capability by the staff. mostly terns. the strike team and the logistic service. The men worked every day from early morning to sunset. supply. coordinated preparedness plan to ensure: — speedy and certain judgement of the situation. forks. — speedy communications with/alerting of all responsible authorities. The CD Corps won national and international praise for its efforts. 100 men working continuously for almost 2 weeks. Nor were any negative effects on tourism registered. 11 million DKK (1·6 million US$). With c. sand. became victims of the pollution. it was necessary to have a well functioning logistic service to keep operations going efficiently (feeding. such as shovels. Very simple equipment. equipment and personnel had to be improvised in the open. The oil spill had hit a large number of sea birds in the Læesø area. The waste was temporarily stored in pits dug out in the area and lined with oilproof plastic. The means to maintain these services were brought from the barracks in Jutland. The total costs involved in the oil spill from ‘Jan’ of Bremen were c.

such as environmental organizations and hunts (to kill polluted birds). proposals. coordination and control under difficult circumstances. methods. insurance companies etc. — being able to muster a well trained. — current updating. working plans. — ample and efficient resources. — establishing close cooperation and coordination among all local authorities involved to be able to: — profit from local and special knowledge. — resources being earmarked for current information of the press through briefings and excursions on the spot (the press must always have real and updated information on the situation). capable of turning theoretical knowledge and previous experience (also at international level) to practical purpose. well functioning staff.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 271 — a standing combatting force at a sufficiently high state of preparedness. — having well trained leaders and strike teams with great stamina and good working morale. — ensuring that the formally responsible local authority is not deprived of its competence but is submitted all important operational suggestions from the staff for the final decision (of great psychological importance). — establishing cooperation or contact with other interested parties to be able to inform about the situation or to seek professional assistance or back-up. operations. — having staff who can master planning. and able to improvise. — discuss ideas. — ensure back-up from the local authorities.. — using valuable international contacts who are able to offer concrete advice due to their own research or practical experience. . — a distinct delegation of tasks and responsibilities among authorities. — having an independent and well functioning communications system (internal and external communications on own radio and telephone networks).. etc. resource needs.

— that the time factor is essential. — to put down and mark all relevant facts (resources used. — the press and observers being handled without disturbing the combatting operation. — that the nature of the area and of the pollution made beach cleaners worthless for the purpose.) to be able to maintain control of the operation. — that good working morale is of vital importance to achieve a favourable result of the operation (this is why many . pollution situation etc. — that modern machinery. supply needs. — to put down all improvised procedures and all practical experience gained in order to be able to prepare manuals at a later date for future operations. it was not possible to verify. — that use of sludge sucking equipment in the ‘wet areas’ (draining canals) had to be given up as it sucked up an unacceptable amount of water. large beach cleaners and sludge sucking equipment cannot be used in such a situation (the sensitive salt meadows could not endure being exposed to heavy mechanical traffic). use of military temporary roads to ensure the absolutely necessary traffic had the result that after 2 weeks’ operation in the area the salt meadows were almost intact. — that the existing coastal oil booms and skimmers are of limited use and are completely useless in high seas and with strong currents along the coast. it is essential in order to achieve a good result of the operation that differentiated decision tools are available (use of modern technology will possibly make the decision process easier). which added drastically to transport costs (use of oil separators was impossible owing to the weight of the vehicles). sufficiently early the captain’s statement of the volume of the spill (just a few tonnes). (b) We also learned: — that.272 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS — resources also being earmarked for information of the many national and international experts and observers who wish to gain practical experience during the operation. due to the absence of remote sensing equipment. the existing mechanical recovery equipment is not sufficiently efficacious in rough seas if thin oil slicks are spread over a large area. — to make daily plans for the work and carry them out according to detailed schedules for each working team.

this would make it possible to launch an immediate operation in the areas most sensitive to pollution. symptoms of lack of liquid and salt (headache and nausea) were remedied by giving ample quantities of refreshing drinks with salt added. . as this is also important for morale. thus achieving the best possible results.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 273 resources were used to ensure efficient cleaning of the men every day after work in the extremely dirty environment). a great effort was also made to ensure that the men had sufficient rest and food. (c) We finally learned: — that it would be a good idea to undertake research into the sensitivity of our national coast stretches. — that working in a tight protective suit in the summer heat creates a need to consume liquids.

Holland. In the explosion. . the major consequences of the accident with regard to the policy-making process will be discussed. England. the situation in Limburg and the Western Mining District before the accident will be described. Finally. In its conclusions. in which 14 employees lost their lives. First. 2 THE SITUATION BEFORE THE ACCIDENT One of the most important events that influenced both public and governmental attitude towards DSM was the disaster on 1 June 1974 at Flixborough.29 The Accident at DSM: Learning from a Major Accident in The Netherlands MENNO J. This accident will be looked at from three different points of view. Members of parliament as well as members of local and provincial councils raised questions as to what would happen if a comparable accident should occur at DSM in The Netherlands. 28 people were killed on the premises of Nypro. The regional department of Civil Defence in Limburg made an analysis of the consequences of such an accident in the Western Mining District. Second. with special attention given to the role of the authorities and questions of coordination. the accident itself will be examined. a daughter company of DSM. The Netherlands 1 INTRODUCTION This paper deals mainly with the accident at DSM (the Dutch Mining Company) on 7 November 1975 in the Western Mining District (of the province of Limburg in the south of The Netherlands).VAN DUIN Leiden University. Department of Public Administration. the lack of professional personnel by the local fire squads and other rescue organizations. caused by ignition of a cloud of escaped vapour (due to a broken pipeline).

. within 2 minutes. a committee was formed with respesentatives from different municipalities.45 p.m. Meanwhile. the province and the management of DSM. was organized. Altogether.m. Its main concern was issues of town and factory planning and the environmental problems of DSM... several local authorities of municipalities around DSM developed some preliminary operational plans for specific accidents like those involving faulty gas lines and chlorine leaks. at about 9. In the meantime. 9 persons had been found dead.. Half-way through 1975. public fire squads of different municipalities arrived at the scene. the entire situation changed dramatically.10 a. but the main danger was the threat of a huge explosion of the two nearby gas balls. around 6 p. serving to analyse the potential dangers of DSM. 3 7 NOVEMBER 1975: THE BIG BOOM IN BEEK On that morning. the first attempts were made to find and rescue possible victims. The first body was not retrieved until 1. The explosion caused tremendous fires to break out through the piping system and nearby storage tanks. by 4 p. Also. Attempts to cool them were frustrated by the wind sending the flames and heat towards them. exploded. This was a setback in itself. Suddenly. the predominant feeling was still that a disaster like the one in England was unlikely to happen in Limburg.m. One of the tanks in the tank park cracked open. incidental plans here and there. Although there were some initial preparatory activities. a working group on disaster management.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 275 including those for medical care. a great number of wounded people were found and 45 of them were brought to different hospitals in the area. About 1 hour later.m. Attempts to put out a tank fire continued.48 a.m. At about 10. Due to the difficulty of the task and the bad cooperation between the local and DSM fire squads all efforts were in vain. 7 November 1975. spreading burning petrol all over the place and forcing the firemen away from the immediate area. a leakage in one of the pipelines occurred. The entire tank park flew up in flames. at the beginning of 1975. while members of DSM’s own fire squad attempted to extinguish the fire. . was emphasized as a major problem in coping with a major accident at DSM. This formed a gaseous cloud which.

At 6 p. the last fire in the tank park died out. partly due to operational problems. the coordination between the local fire squads and that of DSM left much to be desired. During the day. If the same were to happen here in the Western Mining District the consequences would be tremendous. The situation remained critical for about 1½ hours. 3. as a result. were very close to the burning tank park and the gas balls. At this time a lot of people. several car crashes occurred. As if that was not enough.. the traffic started moving again after the opening of the Kerensheide intersection.1. hundreds of curious people came to see the big fire.1 Problems of traffic congestion and disaster tourism The accident occurred less than 100m from a busy highway quite near to the major intersection in the southern part of Limburg.2 Some operational problems The public fire departments from the surrounding municipalities functioned far from optimally. By 3 p. In addition.1 Some specific problems on 7 November 1975 Although it is not possible to discuss in detail all the problems encountered. the police had to close the main routes again. Firstly. the municipal firemen returned to the scene. 3. there were technical difficulties such as mismatching connections for water pumps and hoses. Eventually.m. This explosion led to a great disaster. Notably. The necessary blockage of this intersection (Kerensheide) led to traffic chaos on the surrounding secondary roadways. in a similar situation. mostly disaster tourists. an explosion had occurred in Feyzin. on 12 November. they entered the grounds of DSM on their own .1. and the lack of adequate means of communication. At 8 p. 3. traffic was already hindered at another intersection due to road repairs.276 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS In 1966.m.m. a concise picture of the problems concerning operational activities and management of the operations of the responsible authorities is discussed below. France. At that stage the danger of such an explosion was apparent. when the accident radically escalated. DSM never officially requested assistance from the 8 local fire squads.

The gas balls are situated in Beek but. For example. more than 2 hours after the explosion. Everyone had heard the boom and wondered what had happened. 3. operational services spoke of a fire at DSM’s northern location. Women tried to telephone their husbands working at DSM. once again proving how persistent mistakes can be. the police and the mayor of Geleen did not make direct contact with representatives of DSM until only a few minutes before noon. At 1. did the mayors of Beek and Geleen. The fire burned in two cities.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 277 initiative. The explosion took place on the territory of Beek. The situation of the accident itself best illustrates the complexity and the strangeness of this situation. at about 10 a.m. Although this report was quickly corrected to the southern location. manage to contact each other. DSM was situated in no less than 6 different municipalities. The border between the towns of Beek and Geleen cuts right through the tank park. the regional radio station was still reporting an evacuation of schools at Borne (in the north) instead of at Beek. partly due to the . the cities most involved. but during the day. in the case of an explosion. the residents of Geleen. Later the question was posed as to who would have been responsible if one of the public firemen had been killed at DSM. In one of the early reports. Three years earlier in Amsterdam (Marbon explosion) five local firemen died on the premises of a private firm when helping the firemen of the firm. There were six local crisis centres set up and one provincial centre in Maastricht. would be threatened initially.m.1. For some hours the mayors of Beek and Geleen were not sure if the accident was in their city.3 Information and communication problems Directly after the enormous explosion at DSM all the telephone lines were blocked in almost the whole Western Mining District.16 p. and to a lesser degree Stein..30 p. Not until 12. 3. several firemen went to the north.m. This total blockage in Beek and Geleen lasted for hours.1 Lack of coordination The most important problem experienced on 7 November was the rather chaotic and uncoordinated management of the local and provincial authorities involved.

he answered (incorrectly) that only one municipality (Beek) was directly involved because everything had happened in the municipality of Beek. The public received very little or sometimes no information from these centres. So the police warning around 6. to open the windows because of the new threat of a big explosion.2. it appears that lack of experience played a far greater role in his decision not to act than this ‘legal’ reason (which was mistaken in itself). Thousands of people had heard the big bang from the explosion that morning. to close the windows again.278 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS communication jams. They had heard that everything was under control. each handled their own problems. . In the following we will begin the analysis at the central level. provincial and national levels of government.m. 3.1 The central level At the national level. The lack of coordination was the most important issue in the aftermath. hundreds of windows were shattered.. Of course a lot of people missed this message. It is evident that measures taken on the local level are more often directly related to the accident than measures at the central level.30 p. Of the various local authorities. came as a complete surprise for most of the people of Geleen. Of course. and internal/labour safety will be discussed successively. Yet the local citizens heard nothing from the authorities about what had happened and whether there was any further danger. there was hardly any contact between them. alarming reports were sent out over the radio. issued by just a few policemen. that the roads were opened again.25 p. and now it appeared that the danger was still acute. several areas can be distinguished in connection with the accident at Beek/Geleen. Disaster management. the inadequate means of communication influenced this element. When members of the Provincial Council asked their Governor (Commissaris der Koningin) why he did not intervene to coordinate activities during that day.2 After 7 November 1975 The accident at DSM had particular influence on several matters at the local. 3. as well as the message at around 9.m. followed by the provincial and local levels. However.

The weak management. agreed to new legislation. For many years the most important law on this subject was the Safety Law (Veiligheidswet). A more far-reaching law dealing with disasters not only in times of peace but also war. 1981). Recently this law has been replaced by the Labour Circumstances Law (ARBO-wet.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 279 (a) Disaster management. This new law should oblige all the municipalities in The Netherlands to make local disaster preparation plans. The main reason for these reports was the fact that the labour inspectorate was no longer able to check most of these complicated and high-tech organizations by their traditional methods. Several years later this law. would have to wait. These safety reports should help the labour inspectorate to do their job adequately. due to the above-mentioned accidents. 1972). the lack of disaster preparation plans and the administrative chaos did not go unseen by the parliament. speaking of the accident. (b) Internal/labour safety. they have to make an analysis of both the industrial process and the organizational structure. the Safety Law has been changed. Marbon (Amsterdam. Since then. was replaced by a more extensive version. Eventually more than 2000 of these reports will have to be made. At the beginning of 1975. The accident fell into the parliament like a bomb. Although only marginal changes were made. This new regulation gives more responsibility to the management of the firm. All political parties. the Disaster Law (Rampenwet). up to now about 400 have been finished and approved. In the meantime the existing Civil Defence organization was abolished and regional fire departments had been formed. the bigger and more complex industrial and chemical firms have been obliged to make internal safety reports about their organizations. The accident at DSM caused the bill to be discussed before the Christmas recess. The bill was discussed in a committee of the parliament but broader debate had not yet been planned. as announced in the 1975 bill. the Minister of Internal Affairs presented a bill on Assistance in the Event of Disasters and Accidents (Nota hulpverlening bij ongevallen en rampen). DSM and later Seveso (1976) all influenced policy regarding labour protection and labour safety in firms. Within l½ years a concept law was ready. In 1977. Flixborough. it took another 3 years before the law came into force. The interim law (local disaster preparation plans) became the top priority. . the Law on Local Preparation Plans (Wet gemeentelijke rampenplannen).

Beek survived. The mayor of Beek. Urmond and Elsloo should be combined into one big municipality. the Governor reacted as if the entire event had been only a matter of local involvement—the role of the province was and should be only marginal. All industrial grounds belonging to DSM came under the jurisdiction of the municipality of Geleen. a total restructuring of all the municipalities in the middle part of the province (not only the Western Mining District) came into being.2.280 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 3. By purposely by-passing several regular advisory procedures. In order to deal with the powerful DSM. the Minister of Internal Affairs. the provincial government was able to develop this proposal quickly. which went further than mere cooperation between the different municipalities. they described it as an ‘explosion model’ initiated by a panic reaction to the explosion at DSM. That it did not come to anything was to be expected. Support for the proposal was lacking. Stein. and said that the possibility for future explosions would not be reduced by this reorganization. had managed to convince his party member Wiegel. in 1982. the municipalities of Geleen. Three months later the Governor and the Deputies (Gedeputeerde Staten) reacted in quite a different way. This drastic proposal. and protest meetings ended with the formation of 17 new municipalities from the original 56 municipalities. . The result of the proposal would have been a municipality with a very big firm in the centre surrounded by a number of townships.2 Provincial level Following the accident. It was for a long time uncertain if Beek would be one of them. Eventually. and only when more municipalities would be directly involved and coordination is badly needed would the Governor step in. was made in an atmosphere of urgency and a feeling that it was ‘now or never’. Beek. Criticisms against this proposal naturally came from the municipalities affected. Years of lobbying. however. The Municipal Council of Beek reacted by pointing out that this newly developed municipality model completely differed from earlier proposals about the restructuring of the municipalities from 1968 and 1975. during the first meeting of the Provincial Council (Provinciale Staten).

but talk about the positive role of the DSM fire squads and the fine job these men have done. In the long run hardly any changes were made in the different municipalities. A good example of this can be seen in the speech of the mayor of Beek a couple of days after the accident: ‘All those questions asked by MPs and in the Provincial Council can only lead to panic. Why so anxious? It is better to trust the authorities. some preliminary remarks can be made about the learning possibilities and the learning capacity of the different governmental agencies and authorities involved.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 281 3. restructuring of the fire departments and the changes with regard to disaster management. or the authorities. 4 CONCLUSIONS Although the analysis of all eight cases is not finished yet. The mayor of Beek was angry about all . sometimes like to forget. The saying 4We want to learn from the accident’ is more often than not a cliché. The authorities have waited for initiatives by the provincial and central government on restructuring of the municipalities. local and provincial Deputies will ask questions about what has happened and what can be done to prevent a future occurrence. Attention to the issue is one thing.2. especially when death and injury are involved. Accidents are not only occasions from which we can learn.3 Local level After the accident. the events around the DSM disaster were discussed in different Municipal Councils. (1) Accidents. We should not talk about the few mistakes made. occasions for policy innovation. enduring attention and necessary adjustments after the accident are another. Accidents are also situations which we. In these settings the whole atmosphere was defensive: Don’t talk too much about our own mistakes and be nice to DSM (the biggest employer in the whole district).’ This seems to be the reaction of a man who does not understand how the citizens felt during and after the accident of 7 November. are a matter of political interest Members of Parliament.

Something went wrong. Accidents break into the established order and become a top priority. it was 1982 before one municipality got control over the whole DSM area. After an accident one expects firm but. if necessary. was the adequate and brave behaviour of the DSM fire squads. then tomorrow. The only thing that counted. The provincial proposals from 1976 were perhaps useful in terms of learning. Policy making is a time-consuming enterprise. Authorities waited for national help and national guidance in the form of new legislation. This presupposes a certain degree of size and intelligence. After an accident. Unfortunately the real situation is often not so impressive. Investigations and. even more important.282 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS the attention to the things that went wrong. quick response. as well as from normal circumstances. if not today. investigations begin but the importance of the accident decreases with time. even after accidents and disasters. New things come to the fore. Local governments were not able to learn from the accident. but they were a total denial of the vested interests and the power structure in the area. . (3) Policy making is time-consuming People often believe that adjustments. the organization must have a built-in capacity to cope with information. As a result of the over-reaction by the province. Only some marginal adjustments were made on the local level. (2) Appropriate critical mass If an organization wants to learn from an accident. In the case of DSM both the local and the regional level lacked this critical mass. Local authorities were not able to counterbalance the power of DSM. changes and adaptations should closely follow the accident. In the entire area only a handful of professional firemen were available. an accident occurred and everyone agrees that something must be done. regulations and laws are made only after the calf has been drowned. The organization needs to have a policy memory. he said a few days after the accident. The different municipalities lacked the knowledge to be equal partners with the DSM management. but also a state of stability of an organization.

it was ‘ordinary’ administrative factors. which account for this dramatic delay. Technical reports often remain secret. Fortunately some articles have been written about the technical aspects of this accident. for some years the alerting and informing of people has been a non-issue. Organizations and authorities often do their utmost to make the evaluation reports seem as favourable as possible for them. To evaluate is one thing. For example. the role of the authorities. . They can profit from positive reports (more prestige. For instance.) this situation is far worse. Dutch members of parliament and the Minister of Internal Affairs both agreed on the need for a rapid regulation of local preparation plans. the way to inform the public. Unfortunately this second step seems to be far less developed. The knowledge acquired is often hardly disseminated at all. In the beginning everything went well. (4) Learning by informing One of the conditions for learning from accidents is to communicate to others what has happened and what has been done with regard to evaluation etc. to inform others about the evaluation is another. With regard to evaluating and learning from the non-technical aspects (social consequences. more money) but negative reports can lead to less prestige or even dismissal. Rather. etc. (5) Learning and blaming There may exist a conflict between the will to learn from an acident or disaster and the tendency to cover up the mistakes that have been made by the authorities and organizations involved. the present author had difficulty in obtaining the different reports made after the accident.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 283 In the case of DSM this pattern of diminishing interest and decreasing priority can be seen clearly. such as ministerial changes and a gradual decrease of policy priority. but it took more than 5 years for the law to be enacted. Often the goal of an evaluation is not only to learn but also to blame or to punish. Political problems were not the cause of this delay. Only recently have people in the Western Mining District received information on how to act in the case of an accident.

it also scared and affected the people in the neighbourhood. at the time of the explosion. authorities stressed the fact that the accident was a technical event in a private company. After the accident. They almost completely neglected the fact that this accident had.284 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS In the case of DSM (and not only in this case) only technical reports have been made. At least four different organizations have made more or less technical reports (about the cause and about the operation of the fire squads). far-reaching social consequences. The weaknesses of the local and provincial authorities have hardly been evaluated (see the example about the mayor of Beek). The explosion not only struck DSM. .

inspection etc. a Mobil Oil terminal and a pesticide warehouse of Bayer. West of the terminal. In the vicinity of the terminal there are also LPG storage facilities of AGIP and BP. instruction. at a distance of 1 km. Close to the Jet Oil facilities there is a terminal of Greek refineries (total store capacity 500 000 tons) and close to them a liquefied ammonia storage tank of 15 000 tons capacity. Athens.00 on 24 February 1986 a fire occurred at the Jet Oil terminal near Kalochori in West Salonika. In this paper an attempt is made to apply this principle to the Jet Oil terminal plant at Kalochori in February 1986 and to present action taken by the state and the company itself to further the increase of safety in Greek industry. where some 65000 tons of crude oil and 55 000 tons of fuel oil were stored plus 100 tons of naphtha (total capacity 180000 tons).30 Lessons Learnt from Major Fire Accidents in Greece M. 2 THE FIRE At about 12. Greece 1 INTRODUCTION It is good accident prevention practice to ensure that all possible lessons are learnt from an accident. Kalochori is an industrial area located about 7 km from the centre of Salonika (1 million inhabitants). The widely accepted principle that a majority of accidents are a result of human error is confirmed by this first major technological accident in Greece. The lesson learnt from this is the need to change the method of working.VASSILOPOULOS Ministry of the Environment. there is a village with 1000 inhabitants.—in other words to improve risk management. the capital of Northern Greece. training. .

attributed to the enormous spread of the fire. according to the only scientific publication on the event. it was not possible to approach the bunkers of tanks 7 and 8 from the outside of the terminal. together with improper safety maintenance. 1 and. through the drainage system. the fire brigade decided to cool tanks 1–4 and 10. but fortunately no other victims were recorded. As was previously stated. A number of firemen were wounded. It increased rapidly because of leakage at a sluice valve. It is noted that. was the absence of any observation of the safety regulations during welding work on the plants. . units of the Greek army. Nine of the twelve tanks were demolished together with the buildings and electromechanical equipment. This was followed by an explosion of tank 2. At midnight on 17 February. who took control of the situation during the fire. personnel from the forestry service and the neighbouring industries (Greek refineries and EKO) were involved. during welding work on a pipe. after a heavy rain shower just before the event. Hundreds of tons of vegetables were destroyed because of the dispersion of toxic pollutants such as benzo(a) pyrene over a wide area. the army constructed a dam around the terminal. The fire brigade managed to extinguish the fire in this area. At about 14. The main cause.00 an explosion of tank 7 took place and the fire spread to tank 8 where 15 000 tons of crude oil were stored. was transferred to the bunker of tank No. As expected in such cases. the fire started in the area of tank No. The total cost of the site damage was estimated to be about 22 million US dollars. A truck of the fire brigade was also destroyed. fortunately not in the direction of Salonika. Due to the unsuccessful attempts to put out the fire at this early stage. and to a lesser extent to tank 10. Apart from the local fire brigade. presented by a group of chemical engineers. 7 where 4500 tons of fuel oil were stored. To reduce the danger of fire spreading to the whole industrial area. Also. an interministerial council was established under the auspices of the Vice Minister of Defence and the Minister for Northern Greece.286 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS There is evidence that the original fire started in waste oil dispersed on the soil. mostly by heat radiation. the absence of skilled personnel to implement the emergency plan at an early stage. an enormous boilover in tank 8 took place and the fire spread to all the bunkers of the terminal. On Wednesday 26 February a new explosion took place in tank 7 and the fire expanded to tanks 1–4.

Also. a special unit for use in cases of major technological accidents. (3) Automatic fire alarm systems. and contingency plans were reviewed. The major accident in Salonika either accelerated or induced measures to increase risk management in Greece. It is to be noted that. Safe distances between the new tanks have to follow the new regulations. to a safe distance. and on Monday morning at 03. 3 LESSONS LEARNT For the reconstruction of the terminal. to avoid possible dispersion of pesticides in the area. and the major part into a ship. because of a strike.30 the fire was finally extinguished. the Bayer storage facility was transferred using military trucks. the contents were pumped into the storage process tanks of the fertilizer and refinery plants. and risk assessment for dangerous facilities . Two new laws were ratified in 1985 and 1986 covering safety aspects for the workers and the environment respectively. and such personnel are now being recruited.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 287 On Sunday 2 March the fire had gained access to all areas of the terminal through an explosion of tanks 5 and 6. especially for flammable liquids and gases. during the fire. ministerial decrees about new technical regulations. major effort was put into avoiding its spreading to the Greek refineries terminal by cooling the two tanks nearest to the fire. promptly after the event. To prevent possible damage to the ammonia tank. The lack of scientific and technical personnel in the fire brigade was recognized. three major safety aspects were taken into account: (1) Every part of the terminal can be approached easily from inside or outside. Inspections were performed on almost all industrial units where a major accident could occur. An expanded training programme for inspectors and industry engineers is in hand. including appropriate storage and drainage systems. were promulgated. (2) Continuous control using skilled personnel and electronic equipment. Action was concentrated on tank 5 with success. Also. The fire brigade created.

Risk management in Greece has been started. To hope to avoid accidents in the future is unrealistic. as forecast by the Seveso directive.288 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS has to be prepared by the year 1989. . to reduce their impact through continuous effort is possible.

UK 1 INTRODUCTION We can learn from the past only if we are able to recognize similarities between our past experience and our present situation. they also occur with low frequency within any one industrial sector. Although major large-scale failures are high-intensity events. suggesting that the majority of large-scale accidents arise from combinations of individual. University of Exeter.TURNER & BRIAN TOFT Department of Sociology. Much recent research has been moving towards the development of such a framework. In many fields of industrial engineering we have become skilled at making such links. To facilitate such comparisons a framework must be developed which aids recognition of similar types of causal patterns. so the possibility of gaining a greater understanding of these disturbing events is presented to us [1]. To extend such learning it is necessary to start with the assumption that major failures in large-scale systems are not wholly unique. group. Where we are less skilled is in learning fully the lessons offered to us by major failures in largescale complex systems. 20 or even 10 years ago. so they can be analysed to provide information which will reduce the chances of similar events recurring. disregarding the differing contexts in which they may occur. with the result that thousands of routine industrial operations can now be carried out much more safely than they were 50. It is gradually becoming clear that many disasters and large-scale accidents display similar features and characteristics. social and . All forms of learning based upon feedback require that we link patterns from the past with those cues which might alert us to related patterns in the foreseeable future.31 Organizational Learning from Disasters BARRY A. so to learn from them we must make use of a wide range of comparisons from different industrial sectors.

and they attempt to provide information which will ensure that accidents will not arise from similar causes in the future. But the primary focus for action based upon the lessons of the inquiry lies in the recommendations of that inquiry.290 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS organizational factors. and in order to determine in detail the nature of the response to the inquiries we have followed this up by interviewing representatives of the array of organizations involved with five of these incidents. merely upon the basis of reports of an accident occurring. Public inquiries after major accidents already have to serve a number of purposes: they respond to public concern by trying to ascertain exactly how the events in question came about. they provide an authoritative investigatory basis for any subsequent legal action related to liability. some kind of corrective action is likely to be initiated spontaneously by operators of similar plant. A widely publicised component failure. Public accident inquiries have an important role to play in these learning processes. After a major accident. dissemination and use of information relevant to large-scale accident prevention. be efficiently carried out. We chose to look at public inquiries into major accidents which took place in Britain between 1965 . These cycles of events are. rarely considered in a unified fashion. and in this paper we wish to correct this omission and to address some issues which bear upon the problems of the effective generation. however. or upon the basis of reports of an inquiry in the national or the technical press. and if we are concerned to minimize and contain the adverse outcomes associated with major hazards it is important to look at the nature of inquiry recommendations and at the response to them. and they have little opportunity to develop more wide-ranging analyses or to contribute directly to the emerging debate about system patterns which might aid learning. any resulting conclusions must be disseminated effectively and their implications must be translated by individuals and by organizations into appropriate preventive action. To make any progress at all towards this latter goal an inquiry must. for example. In a current study we have been examining the recommendations made in reports from 19 public inquiries into major accidents. and that these combinations display recurring configurations when detached from their specific technical contexts [1–5]. of course. might prompt checks upon similar components elsewhere. or operators in related industries. but they are normally under pressure to give all of their attention to the matter in hand.

by external agencies or both. identifying and . Such an analysis of recommendations displays to us both the areas which were of concern to those conducting these inquiries and the model of diagnosis and prevention which the inquiry body tacitly adopted. the revision or work procedures or by the modification of existing rules or regulations. and all inquiries made recommendations demanding technical improvements and requiring that certain physical changes be made to plant and equipment. demanding the formulation and dissemination of new rules or procedures. for example. to draw attention to personnel issues such as the need for staff training. calling for improved communication about hazards within and outside organizations. doing this by calling for the initiation of programmes of experimental investigations. An accident is always a physical event. The incidents studied more intensively were all accidents which triggered fires or explosions. The model behind these arrays of practical recommendations stresses the importance of: selecting appropriate physical safety precautions. Recommendations were thus concerned to clarify administrative procedures and arrangements. most of these major public inquiries attempted to develop foresight by making recommendations which offered the possibility of forestalling future problems. an administrative and at a communications level. these inquiries having already had their findings subjected to some detailed analysis [1]. we found that it was possible to discern recurrent types of recommendations [6]. and over 80% of their recommendations were accordingly concerned with organizational and procedural matters. for example. When we examined recommendations from these major accident inquiries. But these large-scale inquiries also normally recognized that the accidents were not solely technical events. Organizational recommendations also typically exhibited concern about information flows.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 291 and 1975. They clearly acknowledged them to be socio-technical in nature. and recommending increased supervision. The study is not yet complete and the analysis reported here is an interim one. as well as sometimes proposing actions which ranged more widely where future plans to deal with a particular hazard were concerned. Finally. or by directing calls for action to organizations not immediately implicated in the particular incident under scrutiny. These particular public inquiries sought to control hazards and to prevent the recurrence of major incidents by advocating action at a physical. or to call retrospectively for improved safety-precautions to be installed by. monitoring or inspection of organizational activities by in-house staff.

and it would be a contribution to organizational learning if such blind spots could be reduced or eliminated. it is. rules and procedures up-to-date. The complexity of events associated with large-scale incidents may thus generate ‘blind spots’ in the lessons drawn from them. A technique known as Schematic Report Analysis has been developed in the course of examining public inquiry reports. training staff appropriately. In recent developments of this technique. One of the problems which such inquiries face in attempting to learn from their investigations is that of marshalling the evidence so that appropriate relationships can be observed and the appropriate deductions made. This kind of approach is broadly in keeping with the direction of the recent research referred to above. on occasion. and attending to the supervision and monitoring of processes and individuals within the organizations concerned. but it is presented in specific rather than general terms in each case. This technique has been used not only to summarize a number of public inquiry reports but also to analyse other types of incidents. and then to locate recommendations within this format in a way . prove difficult to extract all the lessons to be learnt from an incident. But we should ask here whether there are also other ways of making maximum use of the considerable volume of investigation which goes into such accident inquiries. in order to ensure that any wider applicability of their findings is brought to notice and that the response to their recommendations is effective. they may unwittingly end up with a limited set of recommendations. in order to explore the combinations of unnoticed events which develop in the ‘incubation period’ prior to a major disaster. the build-up to the Yom Kippur War [7] and the causation of instances of structural failure [8]. for example. keeping working practices. in turn. improving communication about hazardous matters. It may then. difficult for the interrelationships between events to be fully appreciated.292 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS eliminating ambiguous situations. Schematic Report Analysis has been used first to translate the written synthesis of evidence gathered about a particular incident into a graphic form. Organizations in any sector could doubtless learn much about emergency planning for industrial hazards merely by considering their own operations alongside this very general checklist. Since the evidence taken by an inquiry will run to many hundreds of thousands of words. if members of the investigating team are not able to comprehend fully all the implications of the evidence which they have at their disposal.

although four other organizations declined indirectly. Of these. 2 points up the relationship between the inquiry and the recommendations more clearly and more immediately than does the original report. The development and wider use of this and related techniques clearly have a part to play in improving organizational learning after major accidents. It can readily be seen that the schematic diagram subdivides into six separate yet interrelated clusters of events. can be displayed in a single schematic presentation. Wales. only one refused to cooperate with the research. but the recommendations made as a result of the inquiry seek only to intervene in three of those clusters. Following the examination of inquiry recommendations. Very full versions of Schematic Report Diagrams can be stored as nested sets using proprietary programs such as Macintosh Filevision.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 293 which links them to the relevant aspects of the accident investigation. As well as assisting the analysis of the evidence. Thus. by summarizing large amounts of information into a readily comprehensible form. Whether this omission was due to an oversight on the part of the investigators. and diagrams generated from such stores can readily be used as training aids. but the accompanying figures give some indication of the approach which has been used. such diagrams could assist those investigating accidents to clarify their own diagnoses and to identify more clearly the connections being proposed between diagnosis and recommendation. The recommendations of the five inquiries called for action by a total of 23 organizations. In Fig. Figure 1 illustrates how the events of the incubation period of one particular unwanted incident. as seems more likely. they can help to spread the findings of the inquiry in a more readily accessible form. A fuller account of this technique is available elsewhere [9]. in this case a methane explosion at Cambrian Colliery. . they were unable to formulate appropriate corrective courses of action. on the grounds that no one with knowledge of the incident and its aftermath was now available. the diagram set out in Fig. or whether. this project is making a detailed exploratory investigation of the response to recommendations made by five major inquiries concerned with large fires and explosions 10 years or more ago. 2. the various branches of the causal analysis have been isolated and each recommendation of the inquiry has been related to the train of events which it is intended to prevent recurring. staff having died. As indicated above. work currently in hand at Exeter (with the support of the Economic and Social Research Council) is investigating the long-term feedback cycle instituted by accident inquiries.

294 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

FIG. 1. Outlining of individual event SRAD Cambrian Colliery accident.
Source: Ref. 1.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 295

retired or moved on. A further four organizations had gone out of
business since the incident, but interviews were conducted
satisfactorily with representatives of the remaining 14
organizations involved.
Although the analysis of these interviews is still in an early
stage, some preliminary observations arising from them seem to
be of sufficient interest
to set out here. A point which is
immediately striking is the emotional impact which involvement in
a major accident has upon those concerned. Even after an
interval of 10 years or more, these effects appear to be massive
and enduring. One senior manager still did not wish to talk about
the incident which his organization was involved with, because it
still upset him to think about it, even though his organization
bore no responsibility for the accident. As a result of the effects of
the accident, some of those interviewed reported that their
connection with the accident had triggered a major shift in
preoccupations and activities. One architectural practice, for
example, had shifted emphasis of its work so that the bulk of its
work was concerned with safety matters, whilst an individual in
another practice responded to the shock of discovering the fire
potential of furniture used in his building by spending several
years designing safer alternative fittings.
A second observation relates to the clarity of recall about the
incidents in question by individuals, even after an interval of 10
years or more. We have no independent means of confirming the
accuracy of recall, and all studies of memory and recollection
would lead us to expect systematic distortion in such retrospective
accounts, but informants’ discussions of what had taken place
had a very vivid and immediate quality. They had no difficulty in
presenting their clear account of what had taken place, of the
lessons which had been learned, and of how the implementation
process had been carried out. They had all clearly carried away
and retained, in a very accessible form, their own personal lessons
from the incident.
If such an incident recurred, of course, these personal learning
experiences would not be the only issue which it would be
important to ask about. It would be equally relevant to ask
whether the lessons absorbed by these individuals with direct
responsibility for response to the earlier accident had been
satisfactorily transferred to the ‘memory’ of the organization; for
this to occur, they would need not just to have made an impact
upon this particular band of individuals, but to have been
translated into a form where they had become a pervasive and
accepted part of the organization’s mode of operation [10]. Our

296 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

FIG. 2. SARD showing main cluster of events for Chambrian Colliery accident.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 297

298 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

inquiries are centrally concerned with the extent to which this
institutionalization occurred, but as yet we have no conclusions to
offer on this point.
As far as organizational responses are concerned, the accounts
collected did suggest that action was taken to implement relevant
recommendations from the inquiry, and that it was taken very
quickly, delays in implementation occuring only when large
outlays of capital expenditure were needed. In such cases, the
shortfall in safety during the interim period was typically made
up by the devising of new rules and regulations, and by safety
campaigns to make staff particularly aware of the problem. A clear
preference was expressed within these organizations for forms of
safety training which actively involved employees in safety
practices and procedures, rather than merely making them the
passive recipients of additional sets of regulations or directives.
As one would expect following incidents which had excited
considerable public concern, the recommendations for action were
considered at the highest level in all 14 organizations, followed
subsequently by a meeting or a series of meetings with lower
levels of management, and supplemented in some organizations
by information programmes aimed at the general workforce. Such
a pattern of endorsement of action from the top of the organization
clearly contributed to the speed and scope of the reaction to the
recommendations within the organizations.
Whilst the diffusion of information about the response to hazard
within organizations could readily follow the normal hierarchical
pattern used for other types of in-house communications, clear
differences could be discerned when issues of broader diffusion
were discussed. In very large national organizations which
constituted industries in their own right, difficulties of
communication arising from sheer size were compensated for by
the possibility of using standard communication channels to
ensure widespread and rapid dissemination of a particular warning
or instruction to all parts of the industry with a reasonable degree
of certainty about its delivery. By contrast, in a fragmented or
decentralized sphere of operations, the differential response of
small organizations to information about hazard seemed to be
associated with a lack of cross-communication between small
competitors about such matters, few enquiries being made of
others about their level of hazard awareness.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 299

A SYSTEMS MODEL FOR THE REDUCTION OF
SOCIO-TECHNICAL FAILURES
These preliminary observations from our study may serve to raise
some questions about the manner in which recommendations
contribute to feedback and learning, about the assumptions which
underlie them, and about mechanisms for ensuring that they are
more widely known after an investigation. They raise the question
also of whether it would be desirable and feasible to establish
some kind of unitary hazard reporting system which would
overcome
the
problems
of
hazard
communication
within fragmented and decentralized sectors of activity to which
we have just referred. To help to clarify discussion about
organizational learning and adequate feedback after major
incidents, it might be helpful to try to formulate such a system in
model form.
The elements of such a hypothetical model are set out in Fig. 3,
which is based upon earlier work in which one of the authors was
concerned to apply systems thinking to the problem of reducing
the incidence of socio-technical failures [6, 11]. It is sketched out
here not as an immediate policy proposal, but in order to
illuminate the issues which proposals that moved any way
towards such a system would need to confront.
The schematic model in Fig. 3 is best understood by considering
a proposal to initiate a project which would bring about some
change in the environment, a proposal to build a bridge, say, or to
construct a power station. The left-hand side of the diagram sets
out in a schematic form the kinds of events which might then be
expected to follow, as the design is specified and transmuted into
more detailed proposals which can be reviewed for their
acceptability The implementation of the accepted design and the
development of operational instructions for the project enable the
accepted cycle to be completed by the generation of the changes
initially envisaged—the bridge is built and carries traffic, or the
power station is finished and generates electricity.
What this cycle does not include are the activities suggested on
the right-hand side of the diagram, activities concerned with
learning about design. Typically we do have some kinds of
activities which provide us with opportunities for design learning
but these are rarely seen either as complete elements in
themselves or as key contributors to an overall learning system,
particularly when the information from failures is being
considered. Here we are pointing to the need for a set of
procedures which could help to ensure that the lessons which can

FIG. 3. A socio-technical failure reducing system.

300 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

4. LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 301 . Expanded socio-technical failure reducing systems model.FIG.

we hope. and by making arrangements to incorporate this information into design and implementation. 4 to move a little way away from the wholly schematic. the Design Implementation System translates the proposals of the Design System into operational instructions which would include information about cost. as well as from other sources. and into the management of these processes. Here. whilst the third level of the model adds the Design Learning System to the first two levels. the sub-systems shown are intended to be illustrative of the activities likely to be taking place. 3: a desire for change in the environment prompts the specification of a possible project design. collation and analysis of data from known socio-technical failures. The second level contains the first level plus the Design and the Design Specification Systems. 13]. the . which will.302 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS be drawn from socio-technical failures are incorporated into training and working practices in the future. be devised with the benefit of opinions sought from concerned actors in the system. operations have also been separated out into three levels of functioning. The completed specifications will then be translated into firm proposals. by trying to specify some of the sub-systems which a unitary arrangement for reducing sociotechnical systems failure might contain. This could be achieved by making wider provision for the collection. In the expanded model. A Simulated Systems sub-unit is included in the model to emphasize the importance of the possibility of non-destructive testing of proposals in as many failure modes as possible before the design to be implemented is finalized. with problem-solving and the collation of designs from separate sections of the project in order to avoid difficulties of mismatching. rather than being an exhaustive specification. might provide an organizing framework within which discussions of the improvement of the management of industrial hazards might take place. As in the simplified model. We have expanded the model further in Fig. We hope that this model. initially formulated earlier this year [6] by the application of a ‘Systems Approach’ [12. good communications and a two-way dialogue are important in reducing the possibility of failure. Within each system. The core sequence assumed in the model is the same as that already discussed for Fig. as in other parts of the system. At the first level the Design Implementation and the Operational Socio-technical Systems are to be found. through the activities of sub-systems concerned. among other things.

managers and others in training. making them much more aware of the problems and mechanisms of failure. As these are accumulated. setting up the required sociotechnical system in order to do so. and (c) communicate the knowledge gained directly to industry via trade journals and other publications. except perhaps in its clear diagrammatic formulation. As with the earlier diagram. it merely proposes that in designing and operating large-scale socio-technical systems we should make maximum use of available information by setting up and maintaining negative feedback loops to improve our control of such systems. This should generate both the changes initially specified for the project. in the interaction of the processes already outlined with the Design Learning System. together with relevant academic research findings. The Operational System can then be brought into action to start to engineer the proposed project. special notices (as at present in civil aviation). to ensure that current experience influences practice as soon as possible. or should it malfunction to a level which produces a ‘nearmiss’ catastrophe. engineers. As discussed earlier. the system would be expected to: (a) transmit information directly to the Designer Education Systems. and to initiate corrective control action where this is possible. so that information about failures and the responses to them could be made available to designers. the ordering of materials and so on. A token sample of sub-units is included here for illustrative purposes.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 303 provision of labour. which is here sketched out in a little more detail. this part of the system is expected to receive information from a variety of sources about socio-technical failures. and training courses. Should the project fail catastrophically. The three processes specifically identified here set out some of the important general modes of reaction which are involved. the principal focus of interest for present purposes is to be found at the highest level of the model. the procedures indicated here would be brought into play. the logic of the model looks less self-evident. To many there will be little that is novel in this model. Engineering institutions are resistant to suggestions that they might install procedures for . (b) pass information on to sub-systems concerned with the production of Codes of Practice and Safety Legislation in order to influence procedures and work practices. and also ‘condition data’ which may be used to monitor the project. But when we look at practice.

and specialists in the insurance industry who possess information which is vital to the operation of a Design Learning System are reluctant in the extreme to consider making use of this information in the way outlined by the model. of the nuclear industry [14. The difficulties in moving current practitioners a very small step along the way towards the constructive use of failure-derived information indicates clearly that a sociological analysis of current practices in any industry operating large-scale hazardous systems would reveal a pattern much different from the model outlined. or because of the contraints and anxieties imposed upon them by the need for confidentiality and by associated legal limitations upon the dissemination of failure-related information. in practice. CONCLUSIONS At every large-scale accident inquiry the hope is expressed that the investigations will ensure that ‘this shall not happen again’. An all-embracing information-gathering and reporting system would be difficult to devise and to operate. large-scale systems failures. perhaps. and our response to. either because they fail to see any connection between such discussions as the present one and their own commercial concerns. Research into such matters can only be carried out with extreme circumspection if at all. Several factors contribute to this: frequently there is an assumption that . serious discussion of the issues raised by the model is needed if we are successfully to confront the potential for catastrophic losses which results from the continued development in our societies of large-scale systems with high energy concentrations. neglecting many of the social. We should note too that the model itself has its drawbacks. We suggested above that this model might provide a framework for the discussion of some of the issues connected with learning from disasters. and it might generate undesirable side-effects by accelerating current trends towards the over-centralization of information in our society. and making no overt provision for assessing the way in which issues of commercial. political and military power might impinge upon our understanding of. With all of these provisos. 15]. emotional and aesthetic aspects of the processes discussed. It could also be accused of placing an undue faith in a purely cognitive approach to issues of hazard management and failure prevention. But. adequate learning is often constrained.304 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS reporting on and learning from failures. with the partial exception. however.

I.A. (1984). Manchester University Press. 3. J. & TURNER. perhaps. and that they be passed on in such a way that appropriate action indicated by them is encouraged. B. New York. A. B. B. Technology and Society. Basic Books. New York. (1984). PERROW. REFERENCES 1.. and in devising ways of minimizing such failures in the future. second. C. In general. ASCE. BIGNELL. PIDGEON. and no set of techniques for discerning all relevant patterns in the events surrounding the accident. and finally. Often there is also no readily available perspective making it easy to interpret findings at an appropriately general level. (1986). 5. (1978).A. N. Leeds. 6–9 April.A. (Ed. then. learning from disasters requires: first.Nowak). an outlook and techniques which enable appropriate lessons to be drawn from events which are often similar only at some general systemic level.F. Manchester.F. Much remains to be done in identifying patterns of contemporary institutional reactions to failure. that it is pointless to look for regularities in such ‘Acts of God’. Understanding Systems Failures. and we need to look outside our own industrial sector. D.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 305 the particular large-scale incident is unique and unlikely to recur —even. . B. N. an efficient capability to transmit information from these lessons to those most in need of it. Paper presented to the British Sociological Association Conference on Science. The final stage in developing adequate organizational learning after a disaster requires the lessons identified to be passed on effectively to those who need to know about them. TURNER. Man-made Disasters. Normal Accidents: Living with High-Risk Technologies. In Modelling Human Error in Structural Design and Construction. 4. a wide-ranging investigation. & FORTUNE. Learning about such failures is further inhibited if too limited a range of possible comparisons is scanned. PIDGEON. London. The sociological management of safety. & BLOCKLEY. 2. TURNER. Large-scale failures do not recur with great frequency in any single field of activity. as the model discussed in this paper indicates. Wykeham Press. ‘Human error’ and sociotechnical systems failure in structural engineering. examining incidents in other industries and in non-industrial settings if we are to maximize our chances of spotting repeated patterns and of learning from them.

PIDGEON. International CIS Journal. Wiley. (1986). CHECKLAND. 10. Department of Independent Studies. STECH. (1980). J. Open University Press.W. & VEN DER PAS. B. The Organisation and Use of Abnormal Occurrence Data. N.A. 9. Organisations as Systems. Schematic Report Analysis Diagramming: an aid to organisational learning. 12. (1984). & TURNER. (1982). & SPEAR. KALFSBEEK. Systems Thinking. 14. Design practice and snow loading: lessons from a roof collapse. Unpublished research paper. R.. Loss Prevention Manual. (1987). Technical Note No. AMESZ. 67–71.F. Milton Keynes. unpublished undergraduate dissertation. UK. University of Exeter. P. B. 7. 1(2). (1987). 13. LOCKETT. B. 15. A. PER 672/82. B. The Structural Engineer.72. (1987). The Schematic Report Analysis Diagram: a simple aid to learning from large-scale failures. B. Human Factor Failure in Complex Systems. May. Mathtech. FRANCOCCI. Department of Sociology. M. UK. T. F. Maryland. (1981). (1979).J. 8.A. Varese. 11..A. Organisations have no memory. The European Abnormal Occurrences Reporting System. TOFT. PER 1320/87. New York. G. 13. Systems Practice. Bethesda. PRIMAVERA. KLETZ. 1. TOFT. D. R. (1980). 12–23.87.306 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 6. . 64A(3). H. University of Lancaster. Ispra Joint Research Centre. BLOCKLEY. Varese. UK. Political and Military Intention Estimation. Ispra Joint Research Centre .I & TURNER. TOFT..

SESSION VI Information to the Public Prior to and During an Emergency Chairman: A. UK . Belgium Rapporteur: B.SAMAIN Ministry of Health and the Environment.WYNNE University of Lancaster.

It calls for the establishment of local emergency response committees which. However. The Netherlands 1 INTRODUCTION Risk communication to the general public is becoming a central feature of policy in the area of public health and safety as it relates to technological hazards [1. Its Article 8. it will often turn out to be an issue of . in 1986 Congress enacted the Emergency Planning and Community Right to Know Act as Title 3 of the Superfund Amendments and Reauthorization Act. Amsterdam.32 Communicating Industrial Risk in The Netherlands: Principles and Practice PIETER JAN M. the European Commission passed Directive 82/501/ EEC on Major Accident Hazards of Certain Industrial Activities. Title 3 applies to facilities dealing with one or more of some 400 ‘extremely hazardous substances’. among other things.1 states: ‘Member states shall ensure that persons liable to be affected by a major accident originating in a notified industrial activity…are informed in an appropriate manner of the safety measures and of the correct behaviour to adopt in the event of an accident.’ Although with great variety in phasing and quality.STALLEN Institute for Environment and Systems Analysis. 2]. mostly at State level. Communicating about industrial risk may appear to be a simple and direct issue. In response to the Bhopal tragedy. In 1982. This program has been widely adopted and has generally received local support. the US Chemical Manufacturers’ Association unveiled its Community Awareness and Emergency Response program in 1985. shall make provisions for public meetings to discuss emergency plans. most member states are now envisaging procedures to respond to the EEC requirements for public information practices. Initial legal developments in the US have been scattered. The European Council of Chemical Manufacturers’ Federations has generally supported this so called Seveso Directive. However.

as an attempt to mediate between community conflicts. on the other hand. and as a way to educate the public about the difficult trade-offs in regulatory decision-making. there will be a strong preference for quantitative risk analyses to determine the standard setting process. though less visible. on the one hand. as a clinical treatment that has to be monitored carefully. is the political culture which may shape the subject of the risk communication process. it would be worthwhile to identify a number of principles (if not practical rules) to guide the development and evaluation of risk communication activities. or on information regarding risks to the detriment of information about benefits or the trustworthiness of the risk management institution [3]. In view of the value of social and cultural divergences. For example. Such violations not only make effective communication on health and environmental matters unnecessarily difficult. in risk management institutions. it has been suggested [4] that. in public perception and understanding or in other resources may pose serious obstacles to risk communication. The Netherlands. in media coverage. on information regarding the best understood issues. For example. risk communication can be seen as a commodity that has to be marketed. in open and individualized cultures like the US and. economic. . as a warning signal that has to be responded to. These principles might be derived from a variety of sources of applied scientific knowledge since. and of the desirability to provide an equal level of public protection throughout different countries. Thus.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 309 great scientific. to some extent. the argument goes. to a lesser extent. 2 PRINCIPLES Many of the following principles may seem obvious but nonetheless they are often violated [5]. risk communicators may place undue emphasis on information central to their own regulatory policy. social and political complexity. they may also create an adversarial atmosphere which will be counterproductive in future attempts to communicate. the public administration in order to show its accountability will favour explicitness and sciencebased decision rules. Also of importance is the focused effort of more than a decade of risk perception research. Limitations in the science of risk assessment. Perhaps even more important.

social environment and potential for conflicts which must be realized when initiating communication. In some circumstances a secondary purpose will be to solve problems jointly and to mediate risk disputes. when such situations in fact exist. (2) Know your risk communication objective(s) Viewed from the position of the source of communication.310 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS (1) Know your risk problem Risk problems emerge in a variety of policy contexts. These may be: (a) Identification of hazardous substances and installations (b) Siting of facilities and the zoning of neighbouring activities (c) Management of hazardous processes. . its control and enforcement (d) Establishment of emergency plans Each of these has its own characteristic uncertainties. (B) To stimulate behavioural change and encourage the taking of protective action. most clearly so. standards set to acceptable risk) is the typical thing to do when operating under this mode. (C) To inform the audience.g. Other relevant terms under this heading are: to assure that safety arrangement are adequate. Providing factual information about the risk and its context (e. to create an open and responsive atmosphere. risk communication can be undertaken for one or more of the following three major purposes: (A) To convince the public of the trustworthiness of the risk management activities. This purpose is important in so far as one is anticipating emergency situations and. Too often one admits priority to the information purpose (C) whereas the real motive for risk communication (which one fears to communicate publicly) is to achieve credibility for one’s efforts to manage the risks.

health services). inadequate training of manpower. Evidently. (5) Establish the necessary institutional liaisons Risk is a complex institutional phenomenon as well. fire departments. Often resources are insufficient and sometimes they may even be lacking. and have thus become very sensitive to practical constraints to risk communication. insufficient knowledge of communication target. and consequence mitigation. there are often several intermediary groups. nationwide programs rather than replacing the existing practices by new legal and institutional requirements. and agencies at the municipality level) from that for the reactive aspects. Also. enforcement procedures). Risk communication concerns not only the behavioural and technical measures taken at each stage. companies and local authorities who are already experimenting with risk communication in some form. that will see risk communication as of interest to them too (e. Such practices are important not only because typically they have been developed by trial and error. When identifying such limitations the particular situation should be anticipated where the initial message may lead to a public response that will require a second communication effort.g. but also the decision processes involved (e. the media.g. licensing. In most member states the preventive aspects of hazard control are the mandate of a different institution (ministry. Experience with emergency situations shows that it is effective to establish personal links between the relevant sources [6]. such . inspectorate. lacking financial support. carrying out related activities. (4) Benefit from existing practices In most countries one will find. but also because one would like them to be part of the larger. standard setting. The management and regulation of industrial hazards concerns three stages of risk: accident prevention. at scattered places. Limited resources can be: absence of the proper legal mandate to cover the risk issues.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 311 (3) Know your limitations in achieving your objectives Knowing your limitations is just another way of saying ‘know your resources’. exposure reduction.

g. show sensitivity to the different contexts (personal. Thus. it is also a culturally determined phenomenon reflecting political constellations. as it too offers an indirect opportunity for such behavioural control. Many public information campaigns fail because they expose the public to information and values without linking these to the information and concerns that it already shares. opportunities for protection and remedies for harm. provide qualitative information on accident prevention. Risk communicators should also. Without such linkages. both in form and substance. historical) in which the message may be embedded. of the 10−6 type) and the like is only one aspect of providing cognitive control. Such a form of behavioural control will help prevent anxiety and frustration. to make the public understand and control the hazards to which it is exposed. provision of risk information needs to be accompanied by efforts to assist . (8) Place the risk in perspective Risk is not only an objectively verifiable element of a situation. The presentation of quantitative information on exposure levels (e. but he may also make the inference that the risk communicator is not really interested in his problems. It is for much the same reason that the establishment of local emergency committees with public representation is effective. societal values and conflicts. and thus judge this source as untrustworthy. (6) Determine the audience’s needs One of the most difficult tasks facing risk communicators is to develop messages that. local.312 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS commitment is all the more when one has to rely on the resources of others as well. this information should be structured so as to enable the interested citizen to use easily accessible secondary opportunities to gain additional information. and probably more importantly so. (7) Empower the audience with opportunities to control As a general rule it is necessary to use simple and non-technical language. both written and visual. not only may the individual be simply unable to process the message presented appropriately. Preferably.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 313 individuals in understanding the legitimacy of the often implicit risk context. (10) Monitor and evaluate communication performance and effectiveness Even if the state of risk communication were not so meagre. the mere act of communicating may convey the most important information. Moreover. when information and evidence are more complete and control strategies and supporting rationales more fully developed. It is tempting to start communication late in the process. However. risk communication in its various forms should be a repetitive event. sensitivity is developed. interaction between institutions is promoted. false steps and insensitive approaches are nearly certain. Given our limited understanding. Types of contextual information include [7]: (a) Comparisons with other risks of a similar nature (which needs to be done with intelligence and sensitivity) (b) Comparisons with regulatory standards or natural background levels (c) Comparisons with the extent to which the risk can be and/or has been reduced and what that requires in resources Invariably. Such evaluation will provide the central means to assure that accurate information gets through. and the personal perspective which concerns a particular individual at risk (often oneself). experience seems to show that modest but early activity will be appreciated and will not cause undue anxiety. A well designed risk communication programme should therefore incorporate an evaluation plan both of communicators’ performance (conducted in-house) and of its effectiveness (probably at best conducted by outside evaluators). . individuals will need help in understanding the differences between the societal or management perspective with its typical focus on statistical lives. (9) Communicate in good time The probabilistic nature of risk makes it difficult to decide when to initiate risk communication. In a sense. and credibility grows in the face of adversity and the unexpected. failures in practice should be anticipated.

by mass media. or indirectly. Duphar BV). Additional legal provisions are currently discussed to meet with the demand for a more explicit treatment of the risk aspect of the industrial activity. Another source of information on industrial risk is through local governmental bodies. which is the major professional organization.g. 3 PRACTICE IN THE NETHERLANDS In The Netherlands a number of cases of industrial risk communication exist. what information to present first and what to provide at later stages. e. i. At present there is no legal requirement that forces companies to communicate actively and directly with the public. Decisions must be made about whether to communicate directly. e. see principle (2) above. is publishing a series of popular leaflets on properties and risks of a number of chemical substances. In general. Most important in this respect is the Nuisance Act which requires industrial facilities that can cause danger. For example. As guidelines they facilitate the making of a great number of concrete and seemingly insignificant decisions about procedure and content. there are a number of companies that for various reasons have initiated a public information campaign. If there are limited resources available (as is most often the case at the municipality level). In some cases this has been triggered by an environmental event like an accident or a spill (e. The Netherlands). information campaigns by industry seem to focus more on the preventive aspects of risk management.e. risk communication will often necessitate new institutional relationships. Dow Chemical. . Also important is how to structure the information. Since about 1980 the Dutch Association of Chemical Industries has organized national ‘open days’ with some hundred organizations participating (including relevant university departments and government laboratories). one will have to decide whether to spend them on the most accessible audience or on those most at risk. However.g.314 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS All the above principles have to be taken into account when designing a particular strategy for public risk communication. The Royal Dutch Chemical Association. in other cases it has been because of what one might call the management culture with respect to safety (e. damage or nuisance to apply for a license. problems of overlapping responsibility and legal mandates have to be resolved.g. consequently. by mail. The objective is to gain trust or to restore credibility more than to educate the public by improving factual knowledge.g.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 315

In most cases to which the directive applies, the licensing body is
the regional authority or Provincie. The application for a license
must be published in the (local) newspaper, at the city hall, and
via (non-personal) letters to neighbouring residents. During a 30day period the public has access to most of the information
presented by the company to the local authority, and it can
express its view and/or ask for additional information. Related to
this practice of passive communication is the increasing number of
municipalities that are planning actively to provide information to
the public about what to do in case of a disaster of an industrial
origin. No doubt the Chernobyl accident has strongly contributed
to this interest.
The (further)* implementation of Article 8.1 of the Seveso
Directive in the Netherlands will be based upon four provisions:
1. Risk communication as required by Article 8.1 must apply
both to the preventive and to the repressive aspect of risk
management by industry and (local) government
2. It should be active instead of passive communication
3. Plans for risk communication should be developed with close
cooperation between industry and government
4. Article 8.1 should be implemented as much as possible
within existing legal and institutional arrangements
On the basis of these preconditions, and with particular emphasis
on the fourth one, it was decided to adopt the following framework
for the implementation of the Directive in The Netherlands:
— The Provincie activates (if necessary) the relevant parties, i.e.
the company and the local authority.
— The company prepares the information material and aligns it
with information provided by the local authority.
— The Provincie is formally in charge of inspecting the
appropriateness of the information presented.
Having opted for this framework, the running theme for risk
communication is more likely to be ‘the company and the safety of

* The European Commission has raised an argument about whether the
Dutch government was lagging behind in implementing the Directive. The
debate centred around the question of whether the Directive demands an
active approach or also permits passive means of informing the public.

316 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

Table 1 Proposed items of information to be presented to the public (as a
specification of Article 8.1 of the Seveso Directive).

its environment’ than e.g. ‘industrial risk and regulations’ or
‘industrial hazards and emergency behaviour’.
Of course, this general framework still leaves open a number of
substantive and procedural questions. For example, in the context
of the chosen framework, what should be considered ‘appropriate’
information? We propose to consider risk communication as
appropriate when it contains information on a number of specific
items; these are listed in Table 1. Even with this list, risk
communication can range from anywhere between a one- or twopage letter and a full-blown brochure. Also, and thus perhaps
better, one can make a structured approach, giving relatively
modest information to all persons liable to be affected by a major
accident and offering secondary opportunities for information to
those who show a wider interest (see item 7 of Table 1). The choice
of the above global framework and of the running theme suggests
that items 1.2, 4.2, 5, 6.1 and 6.2 may be given more emphasis
than the others, as the latter are more general in nature and less
geared to the specific relationship of the company to its
environment.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 317

To deal with the various issues of content and form, and to gain
practical experience with risk communication along the above
global lines, it was decided to conduct a pilot study first. This
study started in January 1986 under the supervision of
representatives from various government agencies and industry.
The first phase was aimed at investigating communication
practices in Europe, and in The Netherlands in particular. It also
surveyed areas of relevant research. During the second phase,
started in January 1987, a concrete campaign for risk
communication around a real facility was designed and tested.
ACKNOWLEDGEMENTS
Research for this paper was supported by a grant from the Office
of Nuisance Act and Risk Assessment of the Ministry for Housing,
Physical Planning and Environment. The help of Jose van
Eijndhoven is gratefully acknowledged in The Netherlands.
However, opinions expressed in this paper are the responsibility of
the author and do not necessarily reflect those of the granting
agency.
REFERENCES
1.

2.
3.

4.
5.

6.

7.

BARAM, M. (1986). Risk communication: moving from theory to law
to practice. Paper presented at Annual Conference of the Society for
Risk Analysis, Boston, MA.
STALLEN, P.J.M. & COPPOCK, R. (1987). About risk communication
and risky communication. Risk Analysis, (in press).
FISCHOFF, B. (1987). Treating the public with risk communications:
a public health perspective. Science Technology and Human Values,
(in press).
JASANOFF, S. (1986). Risk Management and Political Culture,
Russell Sage Foundation, New York.
COVELLO, V.T., VON WINTERFELDT, D. & SLOVIC, P. (1986).
Communicating
scientific
information
about
health
and
environmental risks: problems and opportunities from a social and
behavioral perspective. In Uncertainties in Risk Assessment and Risk
Management (Ed. V.T.Covello et al.), Plenum Press.
SORENSEN, J., MILETI, D.S. & COPENHAVER, E. (1985). Inter- and
intra-organizational cohesion in emergencies. Int. J.Mass Emergency
and Disaster, 3, 3 .
KASPERSON, R.E. & PALMLUND, I. (1986). Paper presented at the
Annual Meeting of the Society for Risk Analysis, Washington, DC.

33
Hazard Protection Measures in the
Case of the Release of Toxic Gases:
Principles and Description of the
Concepts
W.HALPAAP
Bayer AG, Leverkusen, FRG
1
INTRODUCTION
Fire services are faced by a particularly difficult problem in cases
where toxic gases are released and threaten the environment.
Rapid notification of the incident, as well as fast decisions on
suitable measures and who is to take them, decide the
effectiveness of their actions.
In Bhopal there was obviously no concept for hazard defence in
the vicinity, with the result that the population was directly
exposed to the full effects of the chemical cloud. A large area was
evacuated following the Mississauga incident. The area was
apparently not exposed to the escaping chlorine cloud, since this
would have caused injury to any persons out of doors during the
evacuation.
In contrast to ignitable gas clouds, the concentration of which
must be greater by at least 4 orders of magnitude (ratio=ppm: %)
for them to ignite, toxic gases have a considerably greater range.
Ignitable gas clouds thus usually only occur in the direct vicinity
of a leak, and generally leave no time for the fire service to take
action. Consequently this paper will discuss actions to be taken
following the escape of toxic gases, giving consideration to their
effect on the environment.
When such events occur, there are two fundamentally different
tasks to be performed at almost the same time:
— Action at the operation site to prevent further escape
— Action to prevent effects on the environment
In this paper, special attention will be paid to actions relating to
the effects on the environment, which have attracted growing
attention in public debate. An agreement reached as long ago as

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 319

1979 in Leverkusen between Bayer AG and the City of
Leverkusen, on cooperation in the event of such occurrences, has
since become a model, and it can be expected to be applied on a
supra-regional basis, initially in the Cologne area.
2
PRINCIPLES
Let us start with a fundamental statement: If a toxic gas cioud
escapes, there is no alternative to seeking protection in closed
buildings!
Assuming normal impermeability of the windows (w=0·2), a gas
cloud inside the building reaches a concentration of only 10% of
the outside cloud after 30 minutes. This realization is supported by
the fact that, although the threat to human life is a function of
concentration and time, the concentration has an overproportional effect. A concentration halved by appropriate
measures can subsequently be tolerated for about 4 times as long.
There is no fixed value which can be regarded as ‘hazardous’ or
‘safe’; the transition is a smooth one.
A hazard is almost always preceded by a clear perceptibility or
nuisance. On the other hand, not every annoying smell is
associated with a hazard. This applies in particular for
decomposition gases occurring during major fires. On the other
hand, only very few gases are odourless, such as CO. However,
these are almost insignificant in relation to their long-range effect.
Nevertheless, no matter what the situation, suitable actions can
reduce an otherwise unavoidable effect and reach better values.
As regards the fundamental effect, it is irrelevant whether
harmful concentrations are to be avoided or simply nuisances.
— People should be influenced to react properly as soon as
possible.
— People should be prevented from going or driving into the
affected area, or they should seek protection in closed
buildings, as shown here.
In the event of evacuation being ordered, people would be sent out
onto the streets precisely during the rising concentration phase.
The fact that people have not suffered injury when evacuations
have been ordered and performed in the past must be due to the
concentrations having been so low that such injuries could not
occur for this reason despite incorrect behaviour.

320 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS

This is why the following point should always be kept in mind:
The more hazardous a cloud is, the more important it is to comply
with the procedures discussed here. Evacuation should only be
considered if it can be performed after carefully studying the facts
and taking special precautions.
3
ELEMENTS OF THE AGREEMENT
One essential prerequisite for rapid initiation of action is fast
information as well as specification of the information routes,
contents and subsequent actions. In Leverkusen, this information
is passed from the works fire service to the municipal fire brigade.
The information content determines the actions to be initiated by
the municipal fire brigade in the city district concerned. The works
security department gives the police the corresponding
information at the same time.
If there is no works fire service, or if the technical and
organizational prerequisites are not fulfilled, the municipal fire
service must, as in the case of a transport accident, for example,
obtain this information itself and make its own enquiries.
Realizing how difficult it is to take decisions on far-reaching
actions with the required speed in the first phase of an operation,
and that making enquiries in advance would take too much time,
the following agreements were reached:
(1) Precautionary information is also given on occurrences which
are expected to remain well below the hazard limits, but where a
more extensive risk cannot be excluded. This also includes
occurrences where ‘third parties’ have the impression of being in
danger, i.e. in cases where the danger is only detected subjectively.
(2) The conceivable effects are initially estimated. This concept is
based on the principle that speed is more important than
accuracy, i.e. that measurements are usually not awaited first.
(3) Even the control room staff of the works fire service has the
authority to send appropriate information to the municipal fire
brigade on the basis of messages received without making its own
enquiries.
(4) The information received (be it from employees, the
population or the plant involved) is used to assess the situation,
i.e. whether a small nuisance or a health hazard must be
expected, must not be expected or cannot be excluded. The result
of this assessment is a precautionary notice Dl or an advance
notice D2, D3 or D4.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 321

— D1 means that probably no effects outside the neighbourhood
need be expected, and that consequently no action will probably
be necessary.
— D2 means that the possibility of effects on the neighbourhood
cannot be excluded, but that action will probably not be
necessary, an agreement on such action being reached, if
necessary.
— D3 means that effects on the neighbourhood must be expected
and that the agreed action must therefore be initiated without
delay.
— D4 would additionally mean that a catastrophe warning will
probably also have to be given.
The terms ‘precautionary notice’, ‘advance notice’ or ‘probably’
mean that the assessment of the situation may still change, but
estimation of the situation at the earliest possible time is required.
The resultant actions are therefore of a particular precautionary
nature and take priority over the assessment whether or not the
corresponding hazard actually exists. The action taken in
accordance with this system is therefore not a definitive indicator
of the actual hazard existing.
Experience has shown that numerous notices have been
submitted in accordance with this system on events which,
contrary to the initial assessment, had no effects beyond the
company limits. In such cases, these notices are treated only
internally by the authorities. The willingness to provide such
comprehensive information is thus rewarded by this information
being treated with a certain degree of confidentiality.
4
EXPOSURE AREA/WARNING AREA
We have seen just how important it is to initiate any necessary
action as rapidly as possible. In addition to providing information
on an occurrence as rapidly as possible, it is also important to
agree upon the extent of its effects. In this context, it is important
to relate the estimated effect to a certain area, since the hazard
changes considerably as a function of various factors, including
the distance from the point of origin. The effects at distances of
1000 m, 2000 m or 10 000 m will be completely different. On the
other hand, it is difficult to define such an area when such an
event actually occurs. With the same quantity escaping in each
case, the difference in the areas results exclusively from a slight
change in the wind speed or weather situation.

At the same time. Furthermore. the result would be a large number of different exposure areas. the control centres cannot direct any desired number of vehicles during the initial phase. There were also debates as to whether a larger area should be assumed. These reports then show whether the area thus defined is larger or smaller than the specified area. the quantity escaping. the major parameter. which has been defined as the ‘warning area’ and is intended to fulfil the following requirements: — The cigar-shaped area resulting from the progagation model used is applied in its basic form because it is more likely to allow rational deployment of the task forces than a sector warning. In other words. It is for this reason that the decision was taken at the very start not to prepare or determine any such exposure areas related to concrete occurrences. or only so large. and whether appropriate conclusions must be drawn. an exposure area was selected. . this proposal was rejected. Above all.322 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS If the tolerance limit (perceptibility. The police take up positions at road junctions outside these areas and cordon them off.g. The police and fire service control centres direct their vehicles to the area defined by the wind direction and the templates. e. that the available task forces. hazard) and the quantity escaping were also to be changed. in the expectation that a larger area would mean greater safety. — The area selected is so large. in about 20–30 minutes. irritation. However. They locate the ‘cloud edge’ and report back on the actual location and intensity of the cloud. since a larger area cannot be covered as effectively as a small one. Instead. is unknown and cannot be determined rapidly enough. police vehicles and fire brigade loudspeaker vans can service this area within an acceptable period. The fire service enters the district with its loudspeaker vans and begins to warn the population. It is consequently senseless to expect that an appropriate exposure area could be selected in a concrete case. a greater area does not mean ‘greater safety’ in this phase. they also ascertain whether they are inside the cloud (as assumed) or still outside.

even before a warning is issued. This again shows that the importance of the fundamental idea of interlinking information with the resultant actions has been recognized. the concept foresees the use of existing public warning sirens and appropriate signals telling the public to switch on their radios to hear more detailed information on the warning. It contains the general recommendation that refuge should be taken in closed buildings and windows closed if unusual gases are noticed. — to the general public by the town clerk.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 323 5 INFORMING THE PUBLIC The current concept is rounded off by an information brochure containing general rules of conduct which is distributed: — to the company employees by the company management. i. there are a number of details to be kept in mind when implementing or further developing this concept: (a) The entire city has been divided into numbered warning districts with specified routes for issuing warnings by loudspeaker vans. The texts have been kept as brief as possible and the vehicle speeds selected accordingly. if necessary — The principles for establishing rules of conduct giving particular consideration to toxic gases — The translation of these principles into rules of conduct for the population In conclusion.e. (b) The information system has also proved of interest for the exchange of information between public agencies. It is. 6 ADDITIONAL COMMENTS The major elements of this concept are as follows: — The checklist as a basis for the company informing the municipal authorities as rapidly as possible — The agreed exposure area as the basis for the action to be initiated in the first phase. conceivable that a warning area assigned to advance notice D3 is allocated a further area where the fire brigade and . At the same time. The district numbering system is based on a neutral point—the Leverkusen motorway junction—not on the company grounds. for example.

Since the Bayer works in Leverkusen have never experienced an event resulting in such effects on the neighbourhood. This is why reports on effects (nuisance. etc. Such a concept naturally also gives rise to the question of the occurrence and effects of possible accidents. . A general order specifying complex or selfcontained breathing apparatus would make large-area actions virtually impossible. irrespective of the specific properties of any particular substance. coughing. In addition.) take priority over measurements. protection through simple breathing equipment. the substance escaping can often not be identified immediately. The fact that it has been possible openly to discuss and agree on this concept in Leverkusen is a reason for satisfaction.) (d) Filter masks will still suffice in the case of large-area warnings. (Details of the additional difficulties in making such measurements or of how difficult it is to obtain useful measurements suitable for analysis will not be mentioned here. On the contrary. The only exception to this rule should be in the direct vicinity of the point of origin. but without developing a feeling of fear.324 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS police control rooms are merely to be informed—in the sense of Dl or D2. we had to make it clear why such actions were nevertheless being prepared and rules of conduct issued—because possibly something could happen. it is my opinion that the existence of a clear concept has even created an additional feeling of safety. i. People accepted that such an incident was regarded as ‘not impossible’ by way of precaution. Road blocks should always be located beyond the cloud so that the use of filter masks is not necessary. (c) It cannot be emphasized often enough that the actions described for toxic substances are largely correct. in view of events occurring elsewhere.e.

to counteract the nuisances and risks which it presents. the media (and particularly television) have a considerable effect on public opinion —the fourth power. While in recent decades the public has become more and more aware of problems of the environment. The major accidents which have occurred in recent years. France 1 INTRODUCTION When confronted with the development of techniques. to acquire credibility in a situation of calm — To prepare for a possible crisis situation and manage information during the crisis . the industries which manufacture basic materials for the other economic sectors have made little progress in their communication with the general public. have led to reconsideration and a change in attitude. On the other hand. Most often they have faced nonspecialist journalists when accidents occur. Demystification is more difficult when technologies make use of knowledge which is no longer at the level of the man in the street. and a certain number of events the amplitude of which has been considerably exaggerated by the media.GROLLIER BARON Institut Français du Pétrole. The present ideas regarding information to the public as far as industrial risk is concerned may be summarised as follows: — To convince the public of the social value of the activity of the establishment.34 Industrial Risk and Information to the Public R. without preparation and in circumstances when it is difficult to achieve credibility. the general public is divided between admiration and the old spectre of the sorcerer’s apprentice. Vernaison.

at the extremes. 15– 20% of people fiercely for or against the industry. On this subject. which gives a positive image. which gives a negative picture.326 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS The first phase. and those whom one can call the general public. etc. that of television is greater than that of the press. Among other things. 2 CREDIBILITY: SOCIAL VALUE There are two sorts of public for an industrial establishment: those who live around the establishment and who are exposed to its nuisances. which is very important. the mayor inspires most confidence (he defends the interests of his electors). — With regard to industrial risks. it was tested on ecologists. even though the credibility of the media is quite small. — They are not worried about pollution by the neighbouring factory but are worried about pollution created by factories further away. and at the centre 60–70% of undecided people whose opinion can be summarised as follows: — They are satisfied with the presence of the factories which bring employment and thanks to which the areas benefit financially thus enabling the building of facilities for the public (swimming pools. is an indispensable preliminary if one wishes to control efficiently the second.).1 Facts The enquiry carried out by Adicra in 1980 among populations living near chemical factories to the south of Lyon showed that the inhibitants were divided into three segments. of the sources of information. sports complexes. they lack credible reference. 2. The first needs reassurance while the second expects more general information. when an audio film on pollution was filmed. the test showed that the ecologists were first of all aggressive but then posed questions when they found that they were dealing with people who were competent to . but wish to know whether all reasonable precautions to reduce the risk of accident have been taken. It should be noted that. they admit that there is no zero risk.

thanks to Rhone-Loire scholarships. From this study and from experience of dealing with the media. one must show that the factory has the permanent aim of reducing nuisances. and of limiting .2 What information should one provide? Following these reports. etc. etc. one must also discuss the negative counterpart of these positive aspects: — Pollution and nuisances — Risk of accidents On these two themes. mentioning its renown in such and such a field. comfort. accessible to everyone. value added tax. In the case of direct information.). car parts. avoiding technical or learned terms as far as possible. show samples of products used or known by the general public which use substances produced by the factory (drugs. On this occasion one can detail the advantages (price. medicines. weight. efficiency. Adicra on the one hand tried to make the chemical industry of Rhone-Alpes known through the media and. — Indicate its contribution to the national economy: volume of business. etc. — Explain what the products are used for if they are not directly commercialised (which is the case most often in chemistry). plastic objects.). — Mention the regional prestige of the establishment by e. sub-contracting and work created. the originality of its products.g. To acquire credibility. Germany and The Netherlands. the following rules governing the way of informing the public have been drawn up: — Give information in simple language.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 327 give them explanations for problems which they had only encountered through the media. revenue for the community and amenities which it has been able to finance. — Show the establishment’s contribution to the local economy: direct employment. 2. exports. performed a study on the actions undertaken in this field in Switzerland. its position at the national and international level. of avoiding accidents which are always possible although improbable.

To do this.). one must not deny the dangerous character of certain products when they are dangerous. their training. although it is a good idea to convince elected officials. one should highlight the professional competence of the factory staff. illness. cars.). these follow the opinion of their electors whatever their personal convictions — Use of staff who are particularly credible vis-à-vis their frends and relations. A triumphant attitude does not favour credibility. etc. they must be encouraged to defend their establishment — Newspapers and business brochures — Local press: — reception of journalists. drowning. etc. etc. These risks should also be given a relative value by comparing them to risks encountered in everyday life (lightning. One may also mention the means available for safety and the performance obtained.328 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS consequences. electricity. — insertion in newspapers of articles on factory life (new installations. commercial successes. etc. It should be noted that. but explain how one can mitigate these dangers. the money spent on these activities. . neighbours. etc. general public. sports equipment. important visits.3 How should the information be given? There are various means which must be adapted to local situations and to the people whom one wishes to convince. professionals in the health service. pupils. In this type of communication it is important to refer to subjects which are known by the man in the street and to his references in matters of probability.). It is very important to be objective on these two themes to acquire credibility. by comparing the activity of the factory with familiar activities whose risks are balanced by a certain value (use of town gas. As an example. 2.). One may cite: — Open days which can be oriented towards certain sections of the public (teachers. the regular performance of safety exercises. One may also help the public to draw up a balance between advantages and inconveniences.

g. This will be the case e. section Special Intervention Plan. it is better to .LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 329 — articles in free newspapers — Local radio and regional television. …The mayors of parishes at risk must receive privileged information from the Directorate of the Establishment and the Commissaire of the Republic Appendix II of the Orsec plan Technological Risks deals with population information cards: ‘…It is obvious that the nature and magnitude of risks for the population varies considerably depending on accident dynamics. such as confinement in the house or evacuation.g. The preparation of safety advice in the form of a card will thus be considered as imperative in all the cases where risk analysis shows the possibility of rapid dynamic accidents. The interministerial communication of 12 July 1985. for associations or organised meetings. dealt with the problem of informing populations: ‘…In most cases the systematic application of these measures will be suitably prepared by information of the population. e.’ In France on 12 July 1985 the Minister of the Interior and Decentralisation issued circular No. 85/170 to the Commissaires of the Republic on new safety planning for technological risks. corrosive or asphyxiating gas ‘If one has an industrial complex which includes several neighbouring installations of different types. — Video films shown in schools. implying immediate countermeasures. together with the circular on the Orsec plan Technological Risks. when there is danger of explosion or of rapid propagation of a toxic. 3 PRELIMINARY INFORMATION FOR A CRISIS SITUATION The CEC Seveso directive of 24 June 1982 concerning the risk of major accidents in certain industrial activities (82/501/CEC) states in Article 8: The member states will ensure that people who may be affected by a major accident caused by a notified industrial activity in the sense of Article 5 are informed in a suitable manner of safety measures and procedures to be followed in the case of accidents. by the distribution of information cards giving safety advice.

4 INFORMATION IN A CRISIS SITUATION There is no lack of examples where bad management of information at the time of an accident or incident increased its gravity in public opinion well beyond the reasonable (Seveso. It is a good idea to prepare for the distribution of this card by preliminary information in order to avoid the creation of unjustified fears. taking account of the following considerations: — The card will be drawn up on one page with the company or companies on the heading and the Directorate of Civil Safety or the municipality of the factory or factories involved.330 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS prepare a single card. Obviously the information card must be adapted to particular cases.). One can note that it would be a good idea to have a standardised European alert code. rather than several documents Industries thus have an obligation to keep the population informed of the risks involved in their activity and of the procedures to follow. One should thus keep in mind some characteristics of information for the media if one wishes to understand how a process of disinformation can be started: . etc. specifying that it is just a complement to the efforts already made on safety.) may pose problems. Montlouis. the Sandoz accident. They must prepare an information card in liaison with the authorities concerned. Joint cards are recommended. The population should be advised to test the radio warning at normal times (no danger situation). — A radio warning on a fixed frequency (local radio) is more efficient than the telephone for obtaining instructions or information. It is recommended that the telephone system should not be saturated in crisis periods. Chernobyl. stadiums. etc. — People responsible for public places must receive special information to enable them to complete the safety advice for their establishment. including all the useful information and advice. the Rheims transformer. Open air amenities (swimming pools. Preferably stations should be chosen which broadcast continuously and which have a certain permanence.

now the information media sell ‘news’ or events. This supposes that this function is envisaged in the organisation of emergency plans and attributed in the same way as other intervention functions. — When journalists have information they try to verify it by comparing the points of view to the extent that they have the time. Answering the press quickly supposes that one does not wait for the drawing up of a communique by the authorities.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 331 — The media must sell their services. saying what one does and . this is the case in most chemical accidents. remain available and have received some training to enable them to inform the public correctly. — Information has more value if its source is credible. a man who bites a dig is’. cold. — Journalists who deal with different facts (including accidents) are not specialists and do not have any particular competence. unexpected. this also has the advantage of establishing personal relationships with the journalists and of obtaining a certain credibility. nature. etc. by anticipating the media. some journalists even say that it is better to publish unreliable information quickly rather than wait for confirmation. — Information very quickly loses its value. who must be of a good level. It is very important to take the initiative in information to avoid the propagation of rumours which could cause panic movements in the population. unusual. The information must be objective and limited to describing the facts without making any assumptions. — One often says that for journalists ‘a dog which bites a man is not news. anticipating them by taking the initiative of announcing the event as soon as possible. One must be able to answer the press very quickly. Good training consists in receiving the press occasionally. The media news event is characterised by its rare. — The journalist who covers an event finds his information where he can from the most accessible people who are not necessarily those who are most competent or most objective.e. such a communique will very probably be too late and will not answer the needs of local journalists. it is a perishable commodity. it has more value when it is fresh. fulfil public demand. — If he wishes to ‘sell’ his article it must draw the reader’s attention with eye-catching headlines. One can deduce the best way of managing information in a crisis situation from these characteristics. This implies that the designated people. i.

He should be designated a priori when the PPI is being prepared. there should be just one spokesman for industry and the administration. once the operations are terminated. For major accidents. provision of food and lodging. Local radio and television may play a very important and positive role in the giving of orders. the media should be informed of the outcome. the procedures to be followed to facilitate their task according to the importance of the accident (making available a room with means of communication: telephones. documentation. the information for the public is carefully prepared.332 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS does not know. it involves informing the people involved in the accident. the organisation of a press centre thus helps in accident management. In the United States. no statement should be made about the gravity and cost of the accident. If the accident is large and will last for some time. there is another problem of information of the public which has been mentioned above. reproduction apparatus.). record of entries/exits. evacuation of the population. it includes a detailed list of journalists with no omission with their details. etc. which involve the PPI. Initial information and declarations must only concern indisputable facts: — Type. telex. Do not try to believe that one can give a smattering of technical information to journalists concerned with various matters. This was shown in particular in Mexico where the television gave indications on the organisation of transport of wounded by the underground to the various hospitals. they are what they are and they must be dealt with as such. For very large accidents which may require e. cards. facilities for photographs. Communiques transmitted to journalists by telephone or telex could with advantage be confirmed by special courrier. One should also say that the press will be kept informed of the first results of the enquiry.g. a radio frequency is reserved for the transmission of these messages. audio-visual material. place and time of the accident — Installation name — Product involved — Control measures put into practice As long as no precise figures are available. identification badges. . In US emergency plans.

. Industrialists have understood this situation very well.Giraud said recently. one must be ‘media-oriented’ because disinformation is a considerable risk for our societies and can lead to aberrations. As A. In the Grenoble region the head of an establishment where a risky investment is underway informed the public during the administrative enquiry with a remarkable card. they have organised training sessions for communication with the media. according to the old proverb. It is better to spend a little more on public relations and a little less on superfluous investments to increase the illusion of safety.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 333 5 CONCLUSION For a long time industrialists have believed that. as can be seen in certain fields. one should let sleeping dogs lie.

GUTMANN Battelle Institut. Frankfurt. the proper solution depending on the specification of the semantic and sociological demands. FRG & G. Their communication qualities decide their value. however. FRG 1 INTRODUCTION Hazard alarm systems are a special case of communication systems. how may the public be prepared for potential industrial hazards? — The technical aspect. . The analysis in terms of communication systems has to cover three essential aspects: — The semantic aspect: what is the information to be transmitted and to whom? — The sociological aspect regarding the receivers of alarm messages: how may the public be informed without producing unwanted reactions. if they happen they should be handled in such a manner as to minimize harmful consequences. We shall not discuss here the balance between increasing efforts on measures to avoid hazards and increasing efforts on better alarm system. Bonn.35 Requirements for the Planning of Industrial Hazard Alarm Systems with a view to the Application of Modern Communication Systems WOLFGANG ULRICI Ecomanagement Consultant. We start from the notion that incidents can happen despite every effort to avoid them.

. Hazard communications system characteristics.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 335 FIG. 1.

This is particularly important for industrial hazards when fast and well founded reactions are needed in order to protect the plant from technical as well as from secondary damage.g. and basically. ‘passive’ means requiring people to take immediate action in order to protect themselves and leave the hazard to itself for the moment. hazard alarm systems may tell people what they are expected to do next. These differences show up in the rates and amounts of information that are to be transferred. hazard alarm systems are simply to indicate that there is danger imminent. are characterized by elevated amounts of information to be transmitted. the other links needing only medium-rate channels. hazard management. Active reaction systems. In this context. ‘active’ means counteracting the hazard and turning to knowing more about it. Third. The essential differences are found in the purpose of triggering passive or active reactions of those concerned. The second task of a hazard alarm system is to inform on the character of the hazard. and hazardcombating services. In addition to this. with a view to industrial hazards. on the other hand. in a later phase. hazard alarm systems will provide for the means of communication among those concerned in an incident. the number of communication partners and the direction of the channels (one-way. in a later phase. two-way) will have to be considered. . This will advise people to protect themselves. indicating e. 1. The advice may be very simple. but only medium to low amounts of information to be transmitted. why the orders given are necessary. We may distinguish two major kinds of alarm systems for different stages of the reaction on to hazard. These characteristics are tentatively drawn up in Fig. such as ‘seek shelter immediately’. in other cases it may be more elaborate.336 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 2 BASIC GOALS AND MODELS OF HAZARD ALARM SYSTEMS First. high-rate channels are required only between hazard observers/analysts. Fourth. Passive reaction systems are characterized by high information transmission rates.

For this. The other channels with particular requirements are those from hazard management and authorities towards the public during the first. being short for reaction time reasons. time scale of evolution. in the first. So it is of crucial importance to shape the passive alarm system so as to convey . The essential requirement is that alarm messages. In the first phase of industrial hazard alarm. In the active reaction phase. process involved. emissions.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 337 3 SEMANTIC ASPECTS Semantic aspects are of particular importance in all those cases where information must be transmitted at a high rate. Even if people have been trained in general.e. and during the active reaction phase among hazard management. hazard observers/analysts and hazard-combating services. messengers may not be sufficiently fast. must be shaped so as to be unambiguously decoded by those concerned. as now the information must suffice to make quick and optimal decisions for combating the hazard. Telephone links will no longer satisfy the needs. any simple alarm signal must be completed by concise (for time reasons). The transmission of images will be necessarry. the requirements for the link between the hazard observers and hazard management will even grow. it is never known for sure if all those concerned in an incident are indeed properly trained and if they have memorized their parts well. specifie information from technical sensors and human observers/analysts that is sufficient for a well founded decision. as their channel width is insufficient. passivereaction phase. particularly as communication both ways may not be feasible because of the number of communication partners. The channel width of a telephone link should be sufficient for this purpose. i. as the number of decision-makers will be small. as hazard management and authorities need sufficient information in order to be able to decide on triggering the internal and public alert systems or not. This imposes restrictions on the amount and the complexity of the information to be conveyed by the public alarm system. the bottleneck being the alarm centres with hazard management and the authorities. Typical data to be transmitted are: affected plant. the alarm system must be able to yield information on the character and the severity of the hazard. requirements are strongest for the links between the site of the incident and the hazard management and further to the authorities. passive reaction phase. potential consequences.

this will be possible only if the messages are transmitted to all those concerned. it may. The alarm system is to guide the actions of those concerned. even if they have been trained for it. Some particular considerations will be necessary for this problem. However. if possible self-explaining. however. and if they are decoded by them as intended. our societies are largely emancipated—some call them ‘ego’ societies—and people are not used to obeying orders if they think they know better. the design of the technical realization of the hazard alarm system. with the public. messages leading to standard reactions. needs are for continuously widening public channels. the intangible consequences of neglecting the information dimension of hazards may be serious. how are you sure to convince them that an official order to leave their cars will be to their advantage? After all. the problem of serving a large number of people with a continuously growing amount of information remains. there is no guarantee that people will behave the way they are expected to. but their transmission rate may slow down. an incident is not a disaster by itself. How will they be reached? How should they behave— turn to flight or seek shelter? If the latter. 4 SOCIOLOGICAL ASPECTS A sector that is often treated with little care in the planning of hazard alarm systems is the communication of the hazard management with the public. . develop into one by affecting people and by wrong reactions. The amount of information that must reach the public is in many cases considered as minor and restrictable to the needs of the passive reaction system. In our opinion this view is inadequate. This part of the system is seemingly unimportant for active reaction purposes. particularly in an emancipated and critical society.338 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS easily understood. In a later stage. The problem is best illustrated with people who are driving in their cars at the time of alarm. A disaster is formed by the interaction of an incident and the people concerned. to a large extent. Even if the potential consequences of the hazard may be technical at the moment. knowledge of the characteristic reactions of people and of their social environment will determine. Confidence in the technical and political decision structures may be essential for the survival of an industrial plant From the sociologist’s point of view. Thus.

A good example of this is the French alarm system Grands Barrages in which the essential messages are precisely formulated and standardized in advance. the order ‘seek shelter and close your windows’ may be accompanied by the reasoning: ‘fire in chemical plant at site xyz. hazard management and authorities should provide reasoning together with orders. one ‘on’ and one ‘off’ sign. (2) As often as possible. (3) The alarm message should not only contain the reasoning for the orders but also a positive message such as: ‘You can avoid harm if you act like this…. Arguments may (and should) be short. they are alarmed.g. whoever needs more information should obtain it. This is a rather delicate point with industrial hazards. Therefore. However. as the population and the public media tend easily to . Training local people to more elaborate reactions to alert signals is a good idea in principle. hazard managers should be trained for their use. and they will be inclined to panic. special training appears to be necessary only if the alert system and the standard reaction should be totally inadequate. sociological reasoning leads to additional requirements (which translate into technical design requirements) such as: (1) It will be generally helpful to keep the public alert signals as simple as possible. people potentially wanting to participate in the hazard planning and to get internal details. emitting noxious gases. in order to improve the efficiency of the alarm system.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 339 So. this again implies that triggering an alarm should always include interrupting the radio programmes in favour of alarm messages. This is not an argument against training in general but only against special training. e. Without this. they are stressed. This will be clear from the consideration of the state of people at the moment of receiving the alarm. even if it can be done financially. this is a delicate point for industrial hazards.g. further messages will follow soon on radio and TV’. (4) In an emancipated society. this will leave you secure for the moment. It is useful to include the shaping of messages and their reasoning as early as in the hazard alarm planning phase. This implies that people must be kept informed as long as the alarm is effective. e. turn off air conditioners’. people might be inclined to turn their air conditioners to full power. So they will be eager for precise directives as to what to do for the best and to know that their needs are cared for. and to train people to just one standard reaction that will be viable in the majority of cases: ‘seek shelter and turn the radio/TV on’.

A solution that may largely satisfy the needs of the public as a whole may consist in a quasi-intercommunicative system: telephone lines to the centres of administration and of hazard management. . this may raise public confidence in the hazard management and help to calm down public stress. particularly if helpers from other countries are involved. the appearance of the hazard managers or their PR officers and of the authorities on the radio/TV would have the effect of demonstrating that the public is not excluded from first-rate information but rather included in the discussion of hazard management and of the consequences of the incident. there is the problem of time dependence of information needs. which may be the case for incidents near the borders of a country. a link to the next radio station. In the first moments after an incident. and suggestions on when and where to get more detailed information. — Messages must be unambiguous. but it may help to manage the first problems of general interest. of a couple of telephone lines to a specially trained PR officer. this does not rule out keeping certain information confidential. namely the installation. there may be some confidential information involved that should not enter the public scene. whereas afterwards more precise information will be necessary to suppress unwanted reactions such as panic or troublesome curiosity.340 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS lose confidence in official statements if there remains only the slightest suspicion that something essential and important for their well-being is hidden from them. Still. This is not a solution for private problems. — Hazard-combating services should be able to communicate with the hazard management both ways. and the broadcasting of information that is of common interest. as somehow the bad truth will come out anyway. The simplest technical solution. Harmonizing these terms is felt to be urgent. On the other hand. This includes even such simple items as the naming of intervention means. (5) The rules for interaction between hazard management and the public should be followed for communication among the hazard-combating services as well. In addition to the amount of information needed. the needs of the public concerned may be fulfilled by just knowing how to protect themselves. will not work because they would tend to be quickly blocked. Moreover. This means that: — Orders to the hazard-combating services should be followed by the reasoning behind them. it is important for the management to keep the initiative.

it was assumed that the authorities. partly public. — The alarm system must reach all people concerned with certainty.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 341 (6) Up to now. such as providing cars with two brake systems that work on different principles. there are certain basic principles to be followed: — The alarm system must be reliable. — The alarm system must reach those concerned in time. active elements being involved only to keep the system from firing. To provide proper functioning in case of hazard is probably conisdered by most planners to be more important. even if this should tend to blunt the instrument to a certain degree. An example of this is the railway security system that causes the engine to stop if the driver does not react . It is our opinion that this is in industry’s own interest and will not need further discussion. being partly hazard management. The first of these requirements is meant to include a guarantee against missing alarm as well as against false alarm. The requirement of sure firing in the case of hazard is taken into account by providing the system with proper redundant elements. Another guiding principle adding to reliability is the passive design of those parts of the alarm system that have to be most reliable and timecritical. and most technical systems give priority to hazard warning instead of suppression of false alert. will be informed as comprehensively and as quickly as possible.g. e.1 Basic principles For any technical solution to the alarm system. 5 TECHNICAL DESIGN OF HAZARD ALARM SYSTEMS 5. the alert should be triggered by passive rather than active elements. as the easier and more reliably the alarm system is triggered the more it is prone to false action because of spurious commands. These two demands are to some degree contradictory. the special role of the authorities in industrial hazard alarm systems has not been specifically considered. It is good technical practice to do this by establishing different systems in parallel. There has to be some compromise.

the lines. In some cases it may be useful to provide the alarm system with a monitoring system indicating the orderly status of the alarm system.342 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS in time to a certain periodic signal.2 Technical options Considering the semantic and sociological requirements restricts the choice of communication means to rather few typical options: — High-rate.e. The technical alarm system must be planned as flexibly as possible. high-amount transmission that addresses only a few partners is provided by TV and special data transmission links. hazards usually do not behave according to planning. i. a portable warning system that is permanently worn may be helpful. However. The second requirement means that the alarm system keeps open the lines to those concerned or provides for reliably opening them in case of hazard. Yet another principle that produces reliability is to provide sufficiently wide and redundant information channels. these channels must be properly secured. The next requirement rules out certain information media such as newspapers for high-rate information as they will not add to the efficiency of combating the hazard. this system is not feasible. These transmission means will be useful . even if a hazard system is planned with much care. 5. The future public ISDN (Integrated Services Digital Network) system applies digital transmission to existing telephone links and will provide data transfer rates of 64kBaud in the first stage. It should always be kept in mind that. such as the public. Apart from technically securing. If only a few persons are immediately concerned. without fixing too many details. Being weak elements of transmission. So the solution of providing different alarm systems for different phases of a disaster and for different degrees of concernedness must be adapted to the mobility and number problem. protecting. So only the fastest information transmitters may be allowed at the hazard site. Finally. there is a problem caused by the mobility of people. the technology should serve well for the purpose of future hazard alarm systems. but for large numbers of persons. as this adds to the complexity of the system and thus provides another potential source of errors. the alarm system should be flexible enough to make the management of unexpected situations possible. such a monitoring system should be checked very carefully for usefulness.

LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 343 for hazard observers and analysts to reports to hazard management during the active reaction phase. They will be used for basic alert signals. poses the greatest problem. and with the authorities in the passive reaction phase. with the exception of the link between hazard management and the authorities which may be served now by radio intercom.e. 2). — Medium-rate. low-amount communication addressing a large number of people appears to be provided best by sound signals of low information content. whistles. bells (the use of the latter is. The question may be raised of whether one should combine the hazard alarm systems for the passive and the active reaction stage into a maximal hazard communication system (cf Fig. sirens. the information of general interest to be broadcast from there. Still. newspapers appear to be too slow a way for mutual communication. not in the technical sense proper but rather in the sense of organizing it. in the future with preference for ISDN . has been mentioned above. better. however. They should be used by the hazard management to communicate with hazard observers/analysts. high-amount communication among few partners may be done by couriers or. — Medium-rate. They are to be used for communication between hazard management and authorities in the active reaction phase. — High-rate. not recommended. medium-amount communication between hazard management and authorities. — High-rate. i. with the hazard-combating services in the active reaction phase. A possible solution for a quasicommunicational system. and from there to the next radio station. The answer is ‘no’ because. with a large number of partners. including telephone links leading to the hazard management centre and to the authorities. as they are reserved for religious purposes and not all people are used to paying heightened attention to them). medium-amount communication among a few partners will be possible by radio or telephone links. there is a strong time dependence of needs that are best served by different systems. future ISDN systems. for most of the links.

i. suggest the proper immediate action to protect themselves.3 Passive reaction alarm systems 5. . if necessary in combination with the appropriate training of local people. one that conveys just the information ‘danger’. as this will. — some kind of short ‘all clear’. Maximal hazard communications system characteristics.g.3. three long pulses. alert signals should be coded as simply as possible.1 Links from hazard management or authorities to the public For reasons of reliable decoding. FIG. The requirement for a warning system that reaches more or less all people concerned in time with unambiguous messages leads to recommending the simple siren whose range of signals may be restricted to: — permanent continuous or oscillating sound as long as the cause for danger is in effect. e. 2.344 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS 5.e. A single-information alert system. may be appropriate for warning the public in a first step. as feedback will not be feasible for most of the people concerned.

this will raise the task of more elaborate coding (and decoding). three long pulses) means ‘all clear’. If one wants not only to convey the ‘danger’ information but also to provide more. the second element indicates the distance to the hazard (short/long). short. The successive elements may be given mnemotechnically easy meanings. Even if a whistle for the transfer of complicated signals is applied. It is a very important principle to make sure that the alert signal proper should never be fired just for test purposes. long. the rather frequent firing of the whole alarm sequence for the sake of technical testing bears the risk of blunting the instrument. the decoding help information that must be present in memory should be as simple as possible.g. even without announcement. These whistles have proved to be very reliable. it is useful anyway to provide for siren signals as the basic alarm. in foggy weather when short message transmission and reaction time is important. The siren does not appear appropriate here because of its limited potential to transfer more than the most simple signals in the short time required. one extra sign (e.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 345 For testing purposes. short. Thus. singleelement signs shall not be allowed.g. another instrument such as a sonore whistle will be better suited. long. short. up to three elements per sign. except for the SOS code: short. The problem then is the training needed to decode the information. this is high for radio amateurs. This yields a unified alarm system for general use . An example of a code of proven efficiency is the Morse code. In some countries. So what appears to be possible is to code the information in short sequences of. say. there is the problem of repeating the message on short terms in order to reach even those who did not get it the first time and to permit them to log in and get the whole message. It contains four elements (the two significant ones being short pulse and long pulse. people tend to ignore the alarm signals. the first element indicates the time for reaction (short/long). e. long pause. but there are probably not many who still remember the Morse code they learnt with the boy scouts. long pause.g. and the short and long pauses used for separation of the elements and signs) that are easily transferred with few errors. for this case. In addition. this kind of instrument is used by vehicles such as railway engines and ships. in order to be decoded fast and reliably. Incidentally. short. long. short. the ‘all clear’ signal may be fired every now and then. e. In order to decode unambiguously short pause and long pause. each sign denoting precisely one message. This means that.

the hazard management and the authorities will need high-rate. just by cutting the sirens. information channels in order to decide upon the alarm message to the public. without the danger of misunderstandings. The widest high-speed information channels required . and to create general confidence in the efficiency of the alarm system. this system will drastically reduce the probability of false alarm. where the signals should be interpreted first by human observers/analysts. but only medium-amount. 5. particularly with industrial hazards. A slight problem exists with the reliability of this more complicated alarm system. it should be obligatory to cut any radio/TV programme in the alarmed area in favour of the transmission of information relating to the alarm and the incident causing it.3. What adds to it is the notion that the whistle may be tested apart from the general alarm system.4 Active alarm reaction systems 5. The number problem will not play the same rule as with the informing of the public. A couple of telephone lines leading to the hazard management centre as well as leading to the authorities will probably be sufficient for the purpose of passive reaction. 5.2 Links from hazard observers to hazard management and authorities In principle.1 Links between hazard observers/analysts and hazard management The rate and amount of information necessary in order to decide upon the most appropriate action may be enormous. The links from sensors may lead decentrally not farther than to the local control centres on site. As most people are equipped with radio and TV. Blocking radio/TV transmissions in favour of alarm messages is recommended in order to inform people comprehensively. An important aspect should be observed. there may still be the problem of simultaneously processing information from several sources. to relieve the telephone system.4. because of the sometimes rapid evolution of incidents. however. On the other hand.346 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS and a superposed special system for special use.

The link back may be provided just by conventional radio to a receiver with the observing personnel on site. A special telephone line. it has the drawback of being a rather rigid.e. alarming and guiding system within the firm.3 Links between hazard management and authorities The authorities may want elevated amounts of information. 5. Short-wave intercoms or mobile telephones should be best. How can this information be reliably transmitted to the hazard management centre (and. it appears that this will argue against such a system. cameras and loudspeakers. should be sufficient for most purposes. high-information communication system. since fixed telephone and video lack the portability required for the operating (and not just observing and analysing) personnel. So it may be a good idea to equip the observers/analysts with light video cameras to transmit images from the hazard site to the hazard management centre.4. using microphones. It is felt that such a system would provide for optimal flexibility. Another conception might consist in using the electric power lines for transmission. combined with guarding the confidentiality of industrial information. a future ISDN system is expected to be . More than that. need portable means of intercommunication. i. maybe in combination with a future ISDN. in most cases the fire services. Besides. to nobody else)? One idea is to install a firm-based monitoring. but will not need it immediately.LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS 347 for hazard management are TV channels. and they are available on virtually any spot of the plant. additional system that requires intensive care.4. from where transmission goes via the lines of the electric power system to the hazard management centre which is equipped with a corresponding receiver/decoder.2 Links from hazard management to hazard-combating services The hazard-combating services. This will probably raise suspicions that the system might be used for other watching purposes as well. The configuration of such a system might then look like this: a portable video camera with a microphone will transmit with weak power to a local transponder/digitizer that is plugged into the next socket. the next best solution being radio or mobile telephone. for reasons of confidentiality. 5. they will provide the power necessary to drive the high-rate.

4 Links between hazard management or authorities and the public An idea for a quasi-commutational system has been described above. It is assumed that the industrial hazard management should have access to the authorities’ public alarm system in order to fire the alert signal if need be. making other communication links more or less obsolete. and that the technical systems are designed according to the needs of those. it is suggested that sociological aspects be increasingly considered when planning industrial hazard alarm systems. may be considered if the hazard management is located far from the site of the incident. . high-amount data transmission such as for the communication between hazard management and the outside. the management will usually be near the site and thus not need remote links. It will be the favourite future link between them. in the case of industrial hazard.348 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS particularly suited just to the needs of medium-rate. for the purpose of triggering the alarm and of setting the passive reaction system into action. whistles and radio/TV appear to be sufficient if properly organized. Regarding new communication technologies. possibly via satellite. Remote radio/TV links. It is important that this should include the close cooperation of industrial systems and of the authorities as well as the information policy towards the public.4. hazard-combating services. 6 CONCLUSIONS In this paper. the coming appearance of wide-channel telephone (ISDN) technology will offer a good opportunity to reshape the communication systems among the hazard management. the existing means of sirens. The probability of misuse appears to be minimal. 5. and the public authorities towards better efficiency. However.

F. France H.Gow.B.BOISSEAU. CEC J. UK Panel members: P.KAY Formerly Health and Safety Executive. FRG B.SCHNADT.HEFFERNAN.W.QUARANTELLI. USA H.WYNNE. UK .L. Ireland E.CONCLUDING SESSION Panel Discussion and Conclusions Chairman: R.

In France. problems had been identified in some instances in producing site-specific emergency plans.and off-site. Three countries had already implemented laws which dealt specifically with the Seveso Directive but Italy as yet did not have such a law and was dealing with the matter through a number of administrative measures in which a network of ministries and organisations was involved. . In Germany. In such situations. particularly in rural areas. and the organisation and means necessary had to be specified. a framework had already existed for the tackling of emergencies. Italy. papers were given on the organisational aspects of emergency planning for chemical accidents in Germany. UK and France. the first step was the identification of the hazard. Within most countries. In the UK. and the introduction of the Seveso Directive had drawn further attention to the industrial sites which presented major accident hazards.36 Summary of the Concluding Session ORGANISATIONS IMPLEMENTING EMERGENCY PLANNING In the first session. followed by an assessment of the consequences of possible accidents and then the drawing up of emergency plans which could mitigate these consequences. there was a considerable history of emergency planning for industrial establishments. and this had led to proposals for amendments to the regulations and for the preparation of further guidance. This led to proposals for on-site plans (POI) which were the responsibility of the manufacturer. Within these countries there were significant differences in the way responsibilities had been allocated between various authorities. the responsibility for off-site planning had been given to the local authority at county or equivalent level. both on. and off-site plans (PPI) which were drawn up under the authority of the Prefect.

Implementation of the requirements of the Seveso Directive was easier in cases where there was already a suitable framework in existence. it was emphasised that an emergency plan that is placed on a shelf to gather dust is worthless and that . an essential feature was the definition of the people who had to carry out various functions. although these were more of a practical than a fundamental nature. The first stage was the preparation of an accident scenario which determined the type of warning system. Finally. to the outside public and. emergency planning had been organised to produce the same positive effects. Particular stress was laid on the need for effective communications to factory personnel. ON-SITE AND OFF-SITE EMERGENCY PLANNING DESIGN In the discussions of the design of emergency plans. It was also necessary to define the type of cooperation that was required. and the coordinator who ensures liaison with outside bodies. When a major emergency occurred. and plans should therefore be as simple as possible. Two particularly important roles were identified: the operational controller who is responsible for the direct response to the incident. the roles of the participants and the overall experience could be evaluated. in particular between industry and the authorities. Some differences did emerge in the approach of different countries.CONCLUDING SESSION 351 Although there were different legal backgrounds and organisational structures in these countries. and this was an area in which further work was in progress. the first need was for a rapid response. It was stressed that plans must be flexible and well defined. to the media. the level of resources required and the amount of coordination that was necessary. In the design of emergency plans. EXERCISES AND AUDITING OF EMERGENCY PLANNING Throughout the conference. and the thorough training of those people in their roles. also very important. between authorities. there was a general convergence of points of view. In this way. There were many difficulties in predicting the consequences of accidents. there was the need for realistic exercises to check the design of the plan and to ensure that it was kept up-to-date.

TECHNIQUES FOR EMERGENCY PLANS The great interest in the application of techniques for emergency plans was reflected in the range of discussions in this section. Modelling could also be a useful tool in designing practical exercises. it was most important to evaluate the experience that had been gained and to assess the performance of the participants. As a result of exercises. The other question related to evacuation. the safest procedure for the public was to remain indoors with doors and windows shut. In the course of this session. One consequence of the need for immediate response. Often. and the lack of immediate availability of some personnel. It was an essential part of emergency exercises that they should be drawn up with clear objectives in mind. and to simulate the results of an exercise. Table-top exercises provided a useful way to test procedures. There was also the danger of overloading communication systems. which involved moving men and equipment. When and how to evacuate would always prove a difficult problem. and the point was made that evacuation should only be considered as a last resort. in some cases it had taken up to an hour to get management staff members to their posts. During the initial stages. alterations in telephone numbers. and a lack of understanding of the roles of others. During the discussion of these particular aspects. Examples were given of theoretical models that could be used to evaluate the effectiveness of existing plans. the sequential telephoning of key personnel took time. should also be carried out. faults had been revealed. An overview was also given of the possibilities and limitations of expert systems and other artifical intelligence methods. Once the exercise had been completed. two issues with general implications were also discussed. defective equipment. There was considerable interest in the use of computers as tools for planning and protection activities. but practical exercises. particularly those in key positions. both theoretical and practical considerations were addressed. technical data were provided from which the hazards presented by industrial installations could .352 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS there must be effective exercising and review of these plans. There was the need to provide the emergency services with information so that they could instantly identify the dangerous substances involved and their possible effects. Finally. meant that units would initially operate with only 50% of their manpower. these included the non-availability of key personnel.

the need and extent of emergency planning could be determined. Indeed. There was rapid development in the field of information technology as systems achieved greater and more rapid response and more flexibility. A more direct application lay in on-line systems for computerised alerting systems. assessing the actual extent of an incident using data from atmospheric sampling/meteorological measurements. In order that a correct interpretation of these events could be made. There would also be further opportunities to develop the use of modern information technology and computerised tools in evaluating problems and assisting in their solution. A study of major incidents in America had emphasised the importance of considering emergency planning as a process.CONCLUDING SESSION 353 be assessed. Computers and modern information techniques provided a wide range of applications. LESSONS LEARNT FROM EMERGENCY MANAGEMENT OF MAJOR INCIDENTS The lessons learnt from major incidents were of two kinds: those related to a specific emergency and those where the lessons could have a much wider application. and assessing the possible consequences of emergency measures. promising lines of research were being pursued. and also for developing the background for accident scenarios as a basis for planning for emergencies. storage and retrieval of information. it appeared that their use for support of the decisionmaking process in emergencies is premature. one solution might be to collect data by means of permanent monitoring systems. There was also a need to ensure that information from the lessons would reach the people who were most in need of it. With respect to the application of expert systems. With fixed installations. questions were raised as to the availability of computers and measuring devices in the early stages. Because of the rapid nature of the development of major emergencies. This led to a deeper understanding of the phenomena arising in disaster situations and would enable emergency managers to cope with unforeseen configurations. and examples were given of off-line programmes that had been developed for risk analysis. from this information. Unexpected events often occurred . However. it was pointed out that the study of incidents in fields other than industrial hazards would often establish repeat patterns. it was necessary to establish a suitable framework for comparison.

attention should also be drawn to the efforts that had been made to ensure that the plant operated safely and. while no attempt should be made to conceal the hazard. Article 8. indeed. During an actual emergency. A review was also given of the requirements for alarm systems. There were a number of areas in which information was necessary: the need to alert people that they lived near an industrial activity with the potential to cause a major accident. Relationships with the media were also very important. INFORMATION TO THE PUBLIC PRIOR TO AND DURING AN EMERGENCY In the Seveso Directive. the nature of the hazard that could be presented. In some cases. In extreme cases. There were different ways in which information could be supplied but they all should be aimed at clarity and simplicity. because of a poor and confused response to an emergency. There was considerable concern from industry initially that the provision of this information could cause unnecessary public alarm. However. particularly to prevent misinformation during an actual incident. Examples were given of the way in which post-mortems on specific incidents had led not only to proposals for technical improvements but also to the consideration of wider issues. it was accepted that radio provided the best means of communication but there was some difference of opinion about the effectiveness of loudspeaker vans. One result of this process was that industry had found that it paid to be as open as possible. changes had been made to local authority organisations. It had been found that there was a much greater impact if personal contacts were made. . and the type of action that should be taken in the event of an emergency. the severity of a major incident could be such that action was decided on a national level and could lead to changes in the law dealing with industrial installations. experience has shown that this was not the case and that the public was more realistic about the appreciation of risks than had generally been believed. and it was suggested that. a robust overall plan was needed which could be applied to these situations.1 requires that persons liable to be affected by a major accident be informed in the appropriate manner of the safety measures and the correct behaviour to adopt in the event of an accident.354 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS in the crisis phase of disasters. such as the organisation of the emergency services. to stress the economic benefits provided by the installation.

for example in small factories. SUMMARY AND CONCLUSIONS The conference has shown that there is worldwide interest in emergency planning. there are some issues that need to be further investigated before formalising guidelines. The content of the communication and the means of its transfer to the media should be further investigated. However. There is basic agreement on the approach to emergency planning but some practical differences are inevitable owing to different national structures. The involvement of the public can also present difficulties. in particular the relation between the content of a safety report and emergency planning. This leads to some general conclusions: (1) The problem of communication with the public deserves further debate. Such a device could prove particularly useful in situations where sophisticated resources are not always available. if an evacuation was required. emergency services and local authorities. This was reflected in the number of papers from different countries and in the information that was presented and discussed. when and . For example. the very nature of an emergency plan can require quite complex relationships between manufacturers. (3) The establishment of general guidelines would be beneficial. It might be food for thought to consider the organisation of a special ‘workshop’ to cover these points. The conference provides a good example. This work could draw on the experience of countries with advanced emergency planning systems and lead to a consistency of approach. The framework so created could also provide useful information for installations that fall outside statutory requirements. it would be necessary to explain to the public the necessity for this step and the facilities that were available. especially those linked with the development of probable or worst case scenarios and consequently preparedness.CONCLUDING SESSION 355 Finally. (2) There is a need to continue to exchange information and experience on emergency planning. The presentations showed that some common problems exist. there are difficulties in predicting the consequences of an accident and hence the area likely to be affected in an emergency. and the Commission (DG-XI) should consider how to build on the work in progress. One of the common themes of the conference was the need to continue working on the improvement of emergency planning systems. Also. in some types of storage and in rural areas.

356 EMERGENCY PLANNING FOR INDUSTRIAL HAZARDS how to evacuate. available monitoring and alerting systems. . safety distances and allowable doses. (4) The progress occurring in the field of advanced informatic tools should be carefully followed and encouraged. existing expert and data base resources. in the near future. integration of plans from other plant in the vicinity. etc. user-friendly informatics for use in emergency management will reach maturity and hence play an important role in this and other related fields. It is to be hoped that.

V. CETESB—Environment Ag.L.Giles College 16 Northumberland Avenue London WC2 5AP UK . Av. Danish Civil Defence and Emergency Planning Agency 18 Vordingborggade 2100 Copenhagen Denmark Bellamy. Danish Police Ridderstraede 1 2100 Copenhagen Denmark Arpe. H. Frederico Hermman Jr. ENICHEM Anic Piazza Boldrini 1 20097 S. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Bardolet Casellas. Direccio General de Provencio Via Layetana 69 08003 Barna Spain Andersen. TECHNICA Ltd. SNAM Progetti S.Donato Milanese Milano Italy Arioli. A. Via Luigi Settembrini 52 Milano Italy São Paulo Brazil Avouris. 345 Bellamy.. N. A. A. L. D. S. G. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Ancarani. F. St. Lynton House 7–12. P.LIST OF PARTICIPANTS Amendola. CEC—DG XVII 200 rue de la Loi Bruxelles Belgium Ancillotti. J. Prof. Tavistock Square London WC1H 9LT UK Aventurato.Donate Malanese Milano Italy Baun. RISO National Laboratory 4000 Roskilde Denmark Barone.

M. K. ROHM and HAAS Italia SpA Via V. Norwegian Petroleum Directorate PO Box 600 4001 Stavanger Norway Blok. S. Ass. Ministero Sanità Via Listz 34 Roma Italy Boato. Carrasco Arias. P.358 LIST OF PARTICIPANTS Bernard. Ministry of Environnement/ DRIR 84 rue de Faretra Toulouse France Boissieras. V. J. L. BEFA GMBH Postfach 8901 5030 Huerth FRG Binetti. J. P. RHONE POULENS TcL 129 rue Servient 69398 Lyon France Bork Kristoffersen. K. Centre Italiano CTL Corso Venezia 37 Milano Italy Boisseau.Pisani 26 Milano Italy Bo. AGIP Raffinazione ENI Piazza Della Vittoria 15 16121 Genova Italy Cassidy. HSE Room 405—Magdalen House HSE Stanley Precinct .GRENFIL Via Emillia Ponente Imola Italy Birden. J. Ministero Sanità Via Liszt 34 Roma Italy Civil Defence Corps 18 Vordingborggade 2100 Copenhagen Denmark Bressan. L. P. Europeenne Gas Liquefile Paris France Caroselli. Gobierno Civil Protec. Civil Avda Marques de la Argentera S/N Barcelona Spain Casarino. G. Province of South-Holland Konigskade 1 2596 AA The Hague The Netherlands Caputo. R. Inspection du Travail 26 rue Zithe Luxembourg Brynjulusen. E.

Gorizia Italy de Palma. MONTEDIPE SpA Via Rosellini 15–17 20124 Milano Italy Corti. ICI Safety & Environm. G. F. T. Gruppo LEPETIT Via G.A. D. G.S. Vigili del Fuoco Via Valleggio 15 Como Italy de Witt. R. H. Regione Emilia Romagna Via dei Mille 21 Bologna Italy Essery. G. Sociology Via Malta 2 3417. Department ICI PO Box 8 Billingham Cleveland UK Fary. . 33 del Sempione 20017 Rho Milano Italy Cipolla. F. of Int. CEFIC—BP Chemicals Limited Boness Rd Grangemouth Stirlingshire Scotland Egidi. Inst.Murat 23 20159 Milano Italy de Marchi. SAGERI 2 rue Ancelle 95521 Neuilly-sur-Seine France Cooney. Ministero Protezione Civile Via Ulpiano 11 Roma Italy Clausen. CEC—DG XI 200 rue de la Loi Bruxelles Belgium Dickie. B.O. BRENK SYSTEMPLANUNG Heinrichsallee 38 5100 Aachen FRG del Bino. Cleveland County Fire Brigade Endeavour House Stockton Road Hartlepool Cleveland TS25 5TB UK Corigliano. L. L. W. AGIP Raffinazione SpA S.LIST OF PARTICIPANTS 359 Bootle Merseyside UK Chiesa. G. B. Topsikring A/S 4 Borupvang 2750 Ballerup Denmark Consolini.

ENEL CEN ‘E.le VV. Via Gregorio XVI 3 32100 Belluno Italy Ginnity. G.V. M. . Generalitat de Catalunya C/Via Layetana 69 Barcelona Spain Genesco. Electrowatt Engineering Services Stanford House Garrett Field Science Park South Birchwood Warrington Cheshire WA3 7BH UK Folino.M. D. Altos Hornos de Vizcaya C/Carmen 2 Baracaldo Vizcaya Barcelona Spain Garcia. G. Frederico Hermman Jr. B. Prof.360 LIST OF PARTICIPANTS Energiesystem Nord GMBH Walkerdamm 17 2300 Kiel FRG Av. GILBY Associates Nethertoun Glebelands Road Knutsford Cheshire WA 16 9DZ UK Ginex. R. FIAT Engineering-Torino c/o Ministero Protezione Civile Via Ulpiano 11 00193 Roma Italy Galvan.di Galeria Roma Italy Garay Unibaso. E. Bruel & Kjaer A/S 2850 Naerum Denmark Giocoli. DATAMAT SpA Via Simone Martini 126 Roma Italy Fox. M.M. CETESB Gilby. F. Comando Prov. L. CRE Casaccia SP Auguillarese 301 00060 S. 345 São Paulo Brazil Feliu. P. R.E.A..Fermi’ Saluggia Italy Gilby. M. AGIP Petroli SpA Via Launentina 449 00142 Roma Italy Gomes da Silva. Ministry of the Interior 1 Place Boauvau Paris France Filippelli. M.F.

Fernando Sousa 11 Lisboa Portugal Gow. H. M. J. Ind. M. Chim. Frederico Hermman Jr. E. R. North East Thames Regional Health Authority UK National Health Service Addison House 32–43 Chart Street London NI 6GP UK Gremmen. F. Industria Italiana Petroli S.lla Varese Italy Greenhill..S. H. V. TUEV Rheinland Postfach 101750 5000 Koeln 1 FRG Holt. CETESB Environment Ag. 4 Ireland Heffernan. Aert Van Nesstraat 45 The Hague The Netherlands Grollier Baron.Jonica 74100 Taranto Italy Haddad. ASAHI Synthetic Fibres Killala . ASAHI Synthetic Fibres Killala Co. Cons. J. Inst. Av. Dow Chemical Europe Sa. J. E. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Honan. NIER Via Stefano Bologna Italy Gullo. FRANCIS SpA Via Origgio Caronno P. CEC—JRC Ispra JRC Ispra Establishment 21020 Ispra Varese Italy Graziano. Prof. Department of Labour Mespil Road Dublin.LIST OF PARTICIPANTS 361 Direccae Geral da Industria Av. Français du Pétrole BP3 69390 Vernaison France Guerrini. D. Mayo Ireland Hesel. Norsk Hydro AS PO Box 646 5001 Bergen Norway Holtbecker. 345 São Paulo Brazil Heffernan. A.

Vigili del Fuoco Via Gnocchi 22 Como Italy . L. ANSALDO SpA Via D’Annunzio 113 Geneva Italy Mancini.M. Province of South-Holland Konigskade 1 2596 AA The Hague The Netherlands Kay. Johns Close Penn Bucks HP10 8HX UK Lepore. Statens Brnadinspektion Kongevejen 207 Copenhagen 2830 Denmark Kahl. Università di Pisa Via Damiano Chiesa 5 56100 Pisa Italy Lucchini. BRUEL & KJAER A/S Naerum Hovedgade 18 Copenhagen 2850 Denmark Lambardi. M. USSL 68 Rho (MI) Via Galdino da Varese 27 Varese Italy Madsen. London) Cotswold St. G. H. Seitnerstrasse 70 8023 Hoellriegelskreuth FRG Kal. ICARO Srl Via Sonnino 9 Pisa Italy Mangialavori. G. F. R. Protezione Civile Ufficio Procivil. ENEA—DISP Via V. Mayo Ireland Johansen. (formerly Health and Safety Executive. L. L. E. N. P. ETS Via Cisanello 32 56100 Pisa Italy Mammone. Ind. Linde ag Industrial Gas Div. FAI Via Cardano 8 Milano Italy Macchi. G.362 LIST OF PARTICIPANTS Co. Via Ulpiano 11 Roma Italy Lanzino. Dipt. ISPESL Via Urbana 167 00184 Roma Italy Loprieno. I. L.Brancati 48 Roma Italy Leonardini.

Emergency Planning Advisory Committee 590 Jarvis Street Toronto Ontario Canada Melis. Studi) V. EDRA Srl Via Gradisca 8 20151 Milano Italy Marchionne. 116/2 001444 Roma Italy Max-Lino. R. AGIP Petroli SpA Via Laurentina 449 00142 Roma Italy Mocke. M. Portuguesa de Isoc. G. B.Tupini. T. Ministero degli Interni Via De Pietro Roma Italy Mariano. 24D03 Cédex 45 92078 Paris La Defense France Marshall. UKAEA Srd Wigshaw Lane Culcheth Warrington Cheshire UK Milone. I.M. G. ESCOM PO Box 1091 Johannesburg South Africa Montini-Trotti.LIST OF PARTICIPANTS 363 Marangoni. ENEA (Dir. 5 Ivy Road Shipley West Yorks. Apartado 30 3861 Estarreja Codex Portugal Marlier. F. DNS SpA Via Nazionale 59 Merano Italy TECHNICA Limited Lynton House 7–12 Tavistock Square London WC1H 9LT UK Marchant. J. BRUEL & KAER Viale U. E.D. M. V. P. Mellin. G. M.Regina Margherita 125 00198 Roma Italy Mostarda. ISOPOR—Ca.A. BPCI 80 New Forest Drive Neath UK Michell. UK Mattia.M. FLEXA SpA Via Custodi 25 Italy . ELF France 2 Place de la Coupole B.

Belliardstraat 18 1040 Bruxelles Belgium Olivier. Z. Ministry of Housing & Environment Athens Greece Odou. Ch. SpA Via Stazione 90 Mornago Varese Italy Picciolo. DATAMAT SpA Via S. G.Cap. Rheinisch Westfaelischer Technischer Ueberwachungs-Verein Steubenstrasse 53 4300 Essen 1 FRG Neuhoff. GALSTAFF Ind. Civil Protection Nat. Van De Vlaamse Gemeensch.M. Kernforsch. Society TNO Aeldoorn The Netherlands Poli. del Servei d’Ordenacio i Seguretat Vial Ajuntament de Barcelona Av. Min. S. M. Portal de l’Angel 8 08002 Barcelona Spain Pesarini. C. A.Donate Milanese Milano Italy Pietersen. C. VINCOTTE Avenue du Roi 157 1060 Bruxelles Belgium Olivieri. M. Service Rua da Bela Vista à Lapa 57 Lisboa Portugal Pares. M. Servigo Nacional de Proteccao Civil Rua de Bela Vista à Lapa 57 Lisboa Portugal Nivolianitou. ISPESL Via Urbana 167 00184 Roma . ENICHEM S. X. P. J.364 LIST OF PARTICIPANTS Mueller. J. G. Cologne Fire Brigade 5000 Koeln 60 Scheibenstr. London Fire and Civil defence Authority 20 Albert Embankment London SE1 7SS UK Paesler-Sauer. Zentrum Karlsruhe 7500 Karlsruhe FRG Paranhos Teixeira. U. 13 FRG Nicolau.Martini 126 00143 Roma Italy O’Reilly.

Dow Chemical GMBH PO Box 1120 2160 Stade FRG Pucciarelli.Brancati 48 Roma Italy Rossi. L. Cap. & Envir. Norwegian Petroleum Director PO Box 600 4001 Stavanger Norway Samain. P. S. Ind. R.M. ENEA—DISP—ARA—SCA Via C. SNAM Progetti S. Comandante Vigili del Fuoco Via Legnani Varese Italy Saltroe. Ass. SHELL Int. Varese Piazza Monte Grappa Varese Italy Pruess. F. Bruxelles Belgium Schlanbusch.F. F. Min. M. A. Petr.E. Santé Publ. A. DSM—CVMD . W. Circumval. E. National Fire Service Rua Julio De Andrade 7 Lisboa Portugal Ricchiuto. PO Box 162 2501 AN The Hague The Netherlands Ribeiro de Almeida.Donato Milanese Milano Italy Rueda. TUEV Rheinland Postfach 101750 5 Koeln 1 FRG Schouteten. L. Norsk Hydro SA PO Box 646 5000 Bergen Norway Schnadt. Disaster Research Center University of Delaware Newark Delaware 19716 USA Raadsen. A. F.lació 1 08003 Barcelona Spain Sacchetti. Unitat Operativa de Gestio Ambiental Ajuntament de Barcelona Pg.LIST OF PARTICIPANTS 365 Italy Pozzi.Brancati 48 00144 Roma EUR Italy Quarantelli.lli Lamberti SpA Via Piave 18 Albizzate Varese Italy Rubino. ENEA—DISP Via V. My. H. D.

University of Exeter Department of Sociology . R. S. Tema-Terr. CEC—DG XI 200 rue de la Loi Bruxelles Belgium Toft. Diagonal 647 08028 Barcelona Spain Singleton. Ministerium des Innerns Schillerplatz 3–5 6500 Mainz FRG Sigales. B.De Gasperi 16 S. Cons. Via Bordoni 4 Milano Italy Serafini. P. F. Univ. A. Dow Chemical Company King’s Lynn Norfolk PE30 2JD UK Smyrniotis. SNAM Progetti V. B. TNO PO Box 541 Appeldoorn The Netherlands Steininger. CEC—DG XI 200 rue de la Loi Bruxelles Belgium Stallen. Dornier System GMBH PO Box 1360 7990 Friedrichschafen FRG Steur. Comando Prov. B. F. Bayer AG Fackbereich Brandschutz D—5090 LeverkusenBayerwerk FRG Tasias.le Vigili del Fuoco Via Gregorio XVI 3 32100 Belluno Italy Sesenna. H.A. A. S.Donate Milanese Malano Italy Serafini.Juan De La Salle 6 08022 Barcelona Spain Testori Coggi. ENICHEM Agricoltura VIA Medici del Vascello 26 Milano Italy Siegmund. W. Politecnico Catalunya Av. T. P. Engineers & Planners A/ S Houedgaden 2 PO Box 51 Birkeroed Denmark Semprini.y Medio Ambiente S.J.366 LIST OF PARTICIPANTS PO Box 603 6160 MH Geleen The Netherlands Selig.

D. Ind. Calor Teo. 1 Ireland Turner. Department of Environment Custom House Dublin. Ministry of Social Affairs and Employment DG of Labour Balen van Andelplein 2 NL—2273 KH Voorburg The Netherlands Van Duin. M. B. 147 Patission Str. 12 Ireland Tuohy. M. M. (LP Gas) Longmile Road Dublin.O. Can. M. ECO Consult Bluecherstr.J. AKZO NV PO Box 186 6800 LS Arnhem The Netherlands Van der KOOI. E.LIST OF PARTICIPANTS 367 Amory Building Exeter UK Via Desenzano 17 20146 Milano Italy Tognoli. SNAM Progetti S. H.H. Pelegri Barcelona Spain Tominez. H. R. M. Umweltbundesamt Bismarkplatz 1 1000 Berlin 33 FRG Valerio. W. M. Province of South-Holland Konigskade 1 2596 AA The Hague The Netherlands Tuite. Athens . Ministry of Envir. ORSA Via Colombo 60 Gorla Minorc Italy Vallmes Rodoreda. M. B. INKE SA Po.Donate Milanese Italy Van den Brand. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Vassilopoulos. 13 5300 Bonn 1 FRG Uth. Research University Leiden Rapenburg 59/2311 GY Leiden The Netherlands Van Leidekerke. University of Exeter Department of Sociology Amory Building Exeter UK Ulrici. USSL 68 Rho (MI) Van der Hooft.

G. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Versteeg. Civil C/Caballeros. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Wynne. Stone-Webster V.Ferrara 36 Roma Italy Zappellini. G.C.L. F. Rensslaer Polytechnic Inst. B. University of Lancaster Lancaster LA1 4YN UK Zanarelli. Jefe Servicio Prot. NIER Bologna Via S.368 LIST OF PARTICIPANTS Greece Versino. 110 Eighth Street Troy NJ USA Wendler. M. G. G. J. Syreco Srl Via Roma 1 Besozzo Varese Italy Zanuzzi. E. G. ENEL Via F. CEDA Srl Via Cagnola 29 Gazzada Varese Italy . B. Ministry Vrom Drvd Stamstraat 2 Leidschendam The Netherlands Villanueva Munoz. 9–2c Valencia Spain Volpe.le Regina Margherita 11 Monza Milano Italy Zanelli.Stefano 16 Bologna Italy Zaro. G. W. S. F. CEC—JRC Ispra JRC Ispra Establishment Ispra Varese Italy Wallace. TOP Srl Via San Lorenzo 12–9 16123 Geneva Italy Volta. Università di Pisa Via Diotisalvi 2 56100 Pisa Italy Zani.

352–4 procedures. 328 Bhopal. 155–7. 197 Appropriate information. 360–1 Back-distance. 119. 3. 96–7 semantic aspects. 108. 205. 354–6 basic goals and models. 345–56 sociological aspects. 343. 343. 137 CHEMIC. 206 Bayer AG. 33. 155–63. 118. 332. 263– 6 Chemical emergencies—contd. 252 Chemical emergencies. 151. 152. 3. 327 Bromine. 354 passive reaction. 298. 160. 355 links from hazard observers to hazard management and authorities. 22 Chemical agents. 356 links between hazard observers/ analysts and hazard management. 317. 255–72 protection measures. 205 Basle. implications of research study. 191. 355 links between hazard management or authorities and the public. 349–51 links between hazard management and authorities. 159–60 369 . 171–2 Association of Civil Defence and Emergency Planning Officers. 325 April Storm. 301 CECOP. 345 basic principles. 270–1 initial response. 255. 299–301 Accidents. 122 Bantry Bay. 257–62 managing responses to. 129–30 Artificial intelligence. 91 Ammonia. 251. 346–9 technical design. 343–56 active reaction. 243 Chemdata. 118 Advisory Committee on Major Hazards. 167–9 251–73 cross-societal applications. 23–4 Act on Calamities.Index Accident inquiries. 354–5 links from hazard management or authorities to the public. 93 Adiabatic Expansion. 36. 243 Cambrian Colliery. 351–2 Alert schedules. 271– 2 impact contingencies. 40 Auditing. 206 Alarms. 168 CFK. 352– 4 links from hazard management to hazard-combating services. 349–56 technical options. 69 ARIES Emergency Centre. emergency action for. 194. 115–31. 35.

336 Chemical installations. 206. 28–30. 218 CIMAH. 136 Chemical industry. 336–8 Crisis situation. 85–97 Chemical Security Group. 226–7. 205 Chemical Industries Association. 208. 40 Credibility. 205 Chemsafe. 137 Chemistry. 70. 81. 83. 212–13 Civil defence. 235 potential risk domain. 227 problem domain. 222–4 cognitive control domain. 317 Community disaster preparedness. 253–4 CONOCO. 338–2 Dangerous Substances (Conveyance by Road in Road Tankers and Tank Containers) Regulations 1981. 178. 286–7 County Emergency Planning Officers’ Society. 226–35 management information. 31 Civil Protection in Peacetime Act 1986. 123 Dielheim. 328 Communication(s). 168 Chemical substances. 133–15 Cologne. 173–7 framework for analysis and design of. 220–2 use of problem representation. 108. 323 Chlorine. 157–8. 275–6 Civil protection. 36. 69 Control of Major Industrial Hazards Regulations 1984. 252 Decision making. 247 . 134. 155–60. 161–3 oil pollution control. 225 emergency response. 32 Civil Defence Corps. 226 mitigation resource domain. 118–22. 243–4. 118 DENZ code. 234 social science approach. 220 mental strategies and heuristics. 254–5 Community social organization patterns. 17. 63. 34. 225. 34 Classification Packaging and Labelling of Dangerous Substances Regulations 1984 (CPL). 161–3 Data banks. 158–9 Decision support systems. 136 Chernobyl. 43. 243–6. for. 217. 37–8. 124–7. 74. 108. 299–300. 7. 343–56 problems of. See CIMAH Co-operation. information for. 230–5 Denmark oil and natural gas transmission. and. 137. 226 decision sequence. 96–7. 135 Danish National Fire Inspectorate (DNFI). 219–40 artificial intelligence approaches. 162 Dansk Olie & Naturgas A/S (DONG). 26–8. 219–24 system science approach. 143. 225–6 implementation of. 31–2 Civil Defence Act 1948. 227 nuclear power installations. 286 Community Awareness and Emergency Response program. 81–4 Coordination problems. 266– 70 Chemical engineering. 155. 151–2. 175–81 Dense Cloud Dispersion. 220 state of the art. 135 Cleveland County Fire Brigade. 64. 219–24 management science approach. 24–6 Dayton.370 INDEX situational contingencies.

50–1 recommendations. 47 design of. 26 Emergency Control Centre. 82 ENEA/DISP. 65–9. 155–7 Research Centre (DRC). 330–1 Falkirk District Council. 36. 155. 283–92 Dutch Association of Chemical Industries. 41 Evacuation. 126 Emergency planning. 84 Federal Anti-Pollution Act. 83. 252–5 Prevention Management. 139–40 table-top. 157–8. 22 Fire accidents. 248–9. 155–60 major accidents. 118. 243–9 protection in vicinity of hazardous plants. 97. 97. 205. 95 Fire services. 36–40 techniques for. 47–57 example. 327 Fireball effects. 137–40 physical-type. 360–1 aim and execution of. 293–5 Fire-fighting. 167–9. 49–50 technique. 138 Expert systems. 359 Emergency Planning and Community Right to Know Act. 322 EMERCOM. 38– 40. 245–6. 212. 85–92 exercises and auditing. 216 Fixed-site situation. emergency and intervention plans. 99–113 Garmisch-Partenkirchen. 211 Emergency plans definition. 115–31. 65–70. 52–6 stages for realisation of. 160 Exercises. 43–4. 6. 81. 285 FIAT-SDT. 185–90 Emergency Services. 118. for. 207. 248–9 . 74 Prevention Service. 3–15 safety concept. 88 Federal Republic of Germany chemical plants. 217. 33. 288 France. 107–13 European Commission. 115. 317 Emergency Planning Authority. 317 European Disaster Medicine Centre. 56–7 external. 287–8 preparedness planning. 283–5. 107–13 on/off site emergencies. 251 Response Act. 38. 222–4 Explosions. 283. 203 establishing. 156 Danish oil and natural gas transmission. 161–3 off-site. 201–3 Estarreja. 143–4. 120. 173–7 to release of toxic substances. 9–12 Feyzin. 215 Flixborough. 83–4. 285 DSM. 211 Emergency Planning Officer. 23–4 organisations implementing. 128–9 EPDES (Emergency Planning DESign) code. 21–4 field of application. 288 management. 211 Emergency Coordinator. 88 tourism. 48–9 manual. 155–60. 208–12 accidents.INDEX 371 Disaster Law. 207. 210. 301 Explosive materials. 91–2. 244 Prevention Plan. 121. 49. 171–83. 255–7 Flammable materials. 210 Exposure area. 361–2 Emergency response decision support for.

74. 65 Controller. 141–2. 135. 135 Hazards. 230 potential risk domain. 224 Information technologies. 255–7 Italy. 89. 197–8 Information brochure. 43 Law on Local Preparation Plans. 213 Hospitals. 230 Information systems for emergency management. 94 Industrial Head Coordination. 68. 209 level. 6–9 Hazardous Substances (Labelling of Road Tankers) Regulations 1978. 179 Industrial Emergency Plan. 63. 66–7 In-plant training. 207. 288 Labour protection. 191–8 GC/MS devices. 142 elimination. 208 Indian Point Nuclear Power Plant. 335–42. 3–9. 206. 206 categories of. 63 procedures. 118 transportation of. 275–81 Large inventory top tier sites (LITTS). 288 Learning from disasters. 96 Industrial risk. 212–13. 331 Information problems. See Onsite In-plant incidents. 328. 329 Leverkusen Model. 293–5 Labour Circumstances Law. 179 Grangemouth. 63 identification. 293–5 Kalochori. 331–2. 55. 152– 3. 17–30 Jet Oil terminal fire. forms of. 208 identification of. 122. 335–42 Industrial safety service. 69–70 Internal Plan of Operation (IPO). 95 Information and Calculation System. 90 Health and Safety Executive. 75–9 reduction. 217 Incident control. 363 Information sources emergency management resources domain. 34 Hazardous materials. monitoring of industrial activities. 119. 121. 141–2 effect. 211 mitigation. 182 In-house operating plan. 286 Information requirements. 78 assessment. 207 Law for Protection against Catastrophes. 93–7 basic elements of. 78. 81 Greece. 140–5 Control Point. 89 Hazardous activities. 186 GEN-X expert system. 63 sources. 125 protection measures. 209 containment. 75 . 288 Labour safety. 144 Hydrocarbons. 288 Læsø. 63 definition. 293–5 Hazard analysis. 205–18 Leverkusen. 47 In-transit incidents. 78.372 INDEX Gas escape. 327–33 Protection Plan. 137 Incompatible land uses. 4 notification and survey of. 73–9 Law of Classified Installations. 192–3 Hazardous installations. 94 Industrial health service. 297–313 Legislation. 36 areas. 134 Hazardous substances. 120. risk analysis of.

73. 11–12. 191. 43. 215. 301 Methyl isocyanate. 253 Mobilization procedures. 341 Oil industry. 73–9. 63–7. 195–7. 149–50 Notification of Installations Handling Hazardous Substances (NIHHS) Regulations. 205 Marbon. 339 ORSECHYDROCARBURES. 362–3 Major industrial risks. 323 Nypro. 252 Local Authorities. 100–1. 205 ORSEC. 205–18 Manfredonia. 96–7 Monte Carlo techniques. 100. 161–3 On-site emergency planning (POI). 47. 359–60 aim and general principles. 202 MONTEDIPE. 81 pollution control. 317 Major Hazard Regulations. 76–9 Major hazards. 73 North-Rhine-Westphalia. 59–62 Nantes accident. 64. 342 Mexico City. 39 LPG. 43 Nuclear industries. 100–1. 275 National Chemical Emergency Centre. 10–11. 284. 82 experience gained. 191–8. 65. 64–6 Operational problems. 171 Nuclear power installations. 102 Nuclear activity. 34. 208–13 Mobilization of resources. 37–8 London Fire and Civil Defence Authority (LFCDA). 318 Methane. 247 Liquefied petroleum explosion. 252 Nuisance Act.INDEX 373 Limburg. 38 Major Accident Hazards of Certain Industrial Activities. 317–26 industrial emergency planning. 283 Off-site emergency planning (PPI). 144. 205–18. 96 Natural Gas Coordinating Committee. 155–60 Notification methods and procedures. 33 Logistics. 283–92 communicating industrial risk. 144 Local Government Act 1972. 54. 103. 194 Methyl isocyanide. 44. 43 . 35 Mexico. 275–81 transmission. 81–4 Major incidents. 190. 205 Major Incident Control Committee (MICC). 44. 288 Markov model. 243–9 lessons learnt from. 102–3 National Agency of Environmental Protection (NAEP). 93–7 NKA/INF project. 128. 285 Oppau. 178–82 Nuclear power generation facility. The accident at DSM. 235 Nuclear wastes. 86. 3. 285 Liquefied gas explosion. 48. 202 Mathematical models. 99. 161–3 Netherlands. 235–9 Nuclear plant emergency response. 205–18. 235–9 North Rhine-Westphalian Law for Protection against Catastrophes. 193. 252 Mitigation. 137 National Toxicity Information Centre. 167 Media coverage. 161 Natural gas transmission. 206 NPK fertiliser. 67–9 definition. 244–5 Macintosh Filevision. 248–9 Liquefiee gas tanker. 301 Maintenance. 191.

63. 34. 360. 109 Security service. 185–90 SMPC. 216. 319 sources. 68 Small inventory top tier sites (SITTS). 318 maps. 288. 128 Provincial Council. 205 Schematic Report Analysis. 323. 321 Risk analysis. 20 San Carlos. 205. 185–6 Police Force. 93. 363 Seveso disaster. 322 Risk contours. 48 Phosgene. 65. 318 appropriate. 9–10. 115 PEE/PLASEQTA. 254 Social value. 321 management. 245–6 Risk. 3. 68. 275 Portuguese National Civil Protection Service. 324. 167–9 Special Intervention Plan (SIP). 127. 252 Plume models. 43 Petroleum refining industry. 100 Particular Contingency Plan. 205 San Juna Ixhautepec. 130 Site Main Controller. 325 limitations in achieving objectives. 107. 36. 19 problem. 289 Public inquiries. 81 Petroleum products.) Regulations 1986. 66. 148–9 Rheinische Olefinwerke (ROW). 295. 135 Royal Dutch Chemical Association. 300 Schematic Report Diagrams. 43. 167–9 Pesticides. 144 Pollution hazards. 319 performance and effectiveness.374 INDEX ORSECRAD. probabilistic nature of. 298 Public Relations. 66. 43. 338. 108 Phosphorus trichloride. 129 Risk— contd. 124. 109 SNPC. 115. evaluation. 301 SDPC. 196 Probabilities of lethality (LTL). 210 hazardous activities. 100. 147–53 Responsibility assignment. 43 ORSECTOX. 78 Safety Law. 99. 48. 7 Petroleum Regulations. 288 Safety reports. 68. 295. 299 Regulation on major industrial accidents. 118 Propane. 136–7 Seveso Directive. 36. 167–9 PEQHU. 199–204 Risk assessment. etc. 319 objective(s). 252 Spain. 144 Radio. 342 Radiological emergency. 209. 12–14 Response capability assessment. 18. 3. 317. 107. 20 Road Traffic (Carriage of Dangerous Substances in Packages. 247 Provincial Contingency Plan of Civil Protection. 109 Social climate. 47 . 217. 192–3 models of. 107. 207 SMART system. 306–12 Somerville. 107–13 Potchefstroom plant. 96 Public transport. 178–82 Recommendations. 295 Petrochemicals. 288 SIGEM system. 95 Self-help schemes. 94 information. 336–8 Socio-technical failures. Massachusetts. 322 Safety equipment. 318 Risk communication. 317–26.

147–53 Warning area. 48. 22–3. 307 Systems models. 47. 38. 76 Yom Kippur War. 119. 330–1 WHAZAN code. 122 SRPC. 90 Spreading angle. 107 Störfallverordnung. 210. 209 Vulnerability assessment. 190 UKHIS (United Kingdom Hazard Information Warning System). 76 Technological Risks. 118. 217 Toxic substances. 200–3. 167–9 TMI-2. 285 Training. 123. 327–33 Toxic materials. 205–18 United States. 205 Thermal radiation effects. 341 Transport of dangerous substances. 126 Wilhelmshaven. 295. 69–70. 194 United Kingdom. 78 Technical Operations Management (TEL). 339 Telephones. 257 Validation. 212. 215 TIGRE computer code. chemical emergencies. 216 Vinyl monochloride. 108 Vulnerability analysis. 83–4. 35. 43. 244–5 Working Environment Act. 157–8 Technical Task Force (TEL). 251–73 US Coast Guard. See Off-site Special protection. 68 Television. 35 Toxic gases. 97. 306–12 Systems overviews. 38 Vapour cloud explosion (VCE). 300 Systems Approach. 207. 54. 185–90 Traffic congestion. 3 Structural failure. 22–3 Trial and error strategy. emergency planning. 94 Works Task Force (WEL). 31–41. 178 Tolerance limit. 342 TELRAD. 300 .INDEX 375 Special (off-site intervention) plan (PPI). 330 Toxic environment. 68 Texas City. 134 Union Carbide.