Professional Documents
Culture Documents
RESOURCE ARTICLE
INTRODUCTION
Q1 By Mary Beth Mulcahy,
Chris Boylan, challenge requires that risk manage-
A series of highly publicized incidents Q2
Samuella Sigmann, ment knowledge must be readily
have highlighted the physical hazards
Ralph Stuart shared between laboratory workers,
of energetic, pyrophoric, and flamma-
their colleagues, and administrators.5,6
ble materials used in academic re-
To be effective, risk assessment and
search laboratories.1–3 These
communication tools must also be
incidents have brought attention to
adaptable to research processes as they
Mary Beth Mulcahy is affiliated with the challenge of managing safely the
change in response to evolving scien-
the U.S. Chemical Safety Board, 1750 multiple, rapidly-evolving chemical
tific needs.
Pennsylvania Avenue, NW, Suite 910, processes that are characteristic of
Washington, DC 20006, USA. the laboratory environment. To keep
pace with this evolution, the Associa- SAFETY MANAGEMENT IN THE
Chris Boylan is affiliated with DNV tion of Public and Land Grant Univer- LABORATORY SETTING
GL, 1400 Ravello Drive, Katy, TX sities (APLU) has suggested that a new
77449, USA. approach to laboratory safety is re- Academic laboratories host many dif-
quired. The APLU suggests that the ferent operations serving a variety of
Samuella Sigmann is affiliated with new approach should address both purposes. Teaching, research and ser-
A.R. Smith Department of Chemistry, technical and cultural aspects of this vice laboratory functions are routinely
Appalachian State University, 525 issue.4 Several hazard identification, intermixed with personnel moving be-
Rivers Street, Boone, NC 28608, USA. evaluation, and management tools tween them as needs arise. This diver-
have been described in the National sity in laboratory operations is a
Ralph Stuart is affiliated with Environ- Academy’s Prudent Practices in the significant challenge when addressing
mental Health and Safety Keene State Laboratory and the American Chemi- safety management in this setting.
College, 229 Main Street, Keene, NH cal Society’s publication Identifying Academic laboratories are quite di-
03435-2502, USA. and Evaluating Hazards in Research verse in scope and activities. For the
Tel.: +1 802 316 9571 Laboratories to address the technical purposes of this paper, ‘‘laboratories’’
(E-mail: rstuartcih@me.com). side, but the cultural aspect of the refer to workplaces that conform to the
1871-5532 ß 2016 Published by Elsevier Inc. on behalf of Division of Chemical Health and Safety of the American 1
http://dx.doi.org/10.1016/j.jchas.2016.10.003 Chemical Society.
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
Occupational Safety and Health Organizations and individual labo- The bowtie approach described in this
Administration’s (OSHA’s) definition ratories must support safety educa- paper presents an opportunity to
of laboratory ‘‘scale and use’’ defined tion and training by establishing improve institutional understanding
in 29 CRF 1910.1450.7 In these labora- clear safety expectations through of laboratory risks by facilitating com-
tories, diverse chemicals are used in policies and procedures and then munication and management of these
multiple processes in quantities that maintaining oversight services that issues among the wide variety of sta-
one person can safely manipulate using identify opportunities for improving keholders in academic laboratory re-
traditional laboratory safety practices. training; search. Stakeholders include
These practices include generic engi- Appropriate emergency response university presidents, senior adminis-
neering controls such as laboratory ven- and waste management services trators, laboratory faculty and staff,
tilation and chemical storage devices, must be provided for the laboratory environmental health and safety staff,
emergency response equipment, work- work being conducted, and and students.4 Such a diverse group of
er training and oversight, and personal Continuous monitoring and re- stakeholders means that communica-
protective equipment. It is important to sponse is required at all levels of tion tools may have to be redesigned
remember that many research activities an organization to ensure adequate with different levels of detail to address
in higher education go beyond this def- and effective barriers to prevent or various questions from different
inition by involving significant non- mitigate laboratory incidents. groups. Fortunately, software for
chemical hazards, using amounts of building bowties is available to support
chemicals beyond OSHA’s definition Compounding the challenge of effec- the reuse of information developed for
of laboratory scale, or by using materials tive safety management is the increas- one set of stakeholders, thus commu-
for which insufficient hazard informa- ingly interdisciplinary nature of nication tools may easily be adapted
tion is available. laboratory science. As biologists, physi- for the needs of others.12,13
As noted by both the US Chemical cists and engineers collaborate with To address this daunting list of
Safety Board (CSB) and the National chemists, the complexity of the hazards needs, ongoing communication among
Research Council (NRC), the tradi- presented increases correspondingly. the various stakeholders at an academ-
tional laboratory safety model is being Control strategies for chemical hazards, ic institution is necessary. A graphical
challenged in the modern academic biological agents and physical dangers representation of an institution’s safety
laboratory.1,8 Emerging factors, such can differ significantly and sometimes barriers and controls is likely to be
as the increasing turnover and diversi- compete for attention and resources. significantly more valuable to support
ty of laboratory workers, and the crea- Cross-disciplinary collaboration can re- communication than a collection of
tion of new materials with unknown sult in workers being involved in labo- text-based policies, procedures, and
hazards, have rendered the traditional ratory processes that they may not be checklists – especially when commu-
laboratory safety practices listed above well educated in. For this reason, it is nicating with audiences who do not
inadequate, taken alone, for addres- important to develop a safety commu- have a background in the chemical
sing research hazards. Preventing nication system that clearly highlights sciences or risk assessment.
and mitigating laboratory incidents significant hazards while being easy to
requires effective management of a sys- use, share, and modify as work changes.
tem of barriers that includes physical, It is also important to remember that THE ORIGIN OF BOWTIE DIAGRAMS
operational, and organizational ele- the management of laboratory safety
ments. For example: has the potential to affect people other Using graphical imagery to describe
than those physically conducting the safety systems has been shown to be
The institution must determine what work. The disruption of neighboring effective in supporting hazard commu-
facilities and management resources activities on many campuses and the nication needs. In 2000 James Reason
are required to support specific re- legal impacts of the 2008 fatal fire at presented a popular model depicting
search proposals (including the abil- the University of California, Los the progression of an incident through
ity to house new chemicals and Angeles (UCLA) made it clear that lab- a series of leaky barriers.14 This model,
assess additional hazards they create oratory incidents can have conse- aptly referred to as the ‘‘Swiss cheese
in the laboratory) and determine quences impacting not just the model’’, is presented schematically in
whether such facilities are available principle investigators overseeing the Figure 1. In this model the slices of
when such proposals are funded; work, but neighboring researchers cheese represent ‘‘barriers’’ such as
Laboratory equipment with appro- and institutions as a whole in terms of those outlined above, while the holes
priate controls must be provided for both productivity and reputation.9,10 represent barrier deficiencies in specif-
the experiments being conducted; Even incidents that result in no injuries ic elements in the system. These defi-
Safety equipment must be identified, can have serious financial conse- ciencies can allow a threat to penetrate
provided, maintained, and inspected quences, such as recovery costs of over the system and result in an incident. An
to ensure that it will function as $1,000,000.11 increased number of barriers with
designed if it is needed (i.e. emergen- These considerations mean that a smaller and fewer holes will lead to a
cy shut-offs, safety showers, eye larger network of stakeholders should more robust barrier system for prevent-
washes, etc.); be included in developing safety tools. ing incidents.
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
Figure 2. Generic bowtie diagram depicting multiple threats that can escalate to a loss of control of a hazard, and in turn,
progress to a variety of negative consequences.
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
escalation of a threat to the point unintentional or intentional scale-ups bowtie for the group. The participants
where it becomes involved in a top may be different, once safety critical initially listed the following as poten-
event. If a top event occurs, then the amounts are exceeded, the barriers to tial threats that could lead to exceeding
mitigation barriers on the right side prevent or mitigate consequences of a safety critical limits (top event):
of the bow tie are intended to either detonation are the same. Therefore,
stop or minimize the severity of un- choosing ‘‘exceed safety critical limit’’ No written procedures
wanted consequences. allows the bowtie to cover multiple No personal protective equipment
threat scenarios for a single hazard. (PPE) policy
Secondly, choosing ‘‘exceed safety Untrained laboratory workers
A LABORATORY EXAMPLE critical limit’’ as opposed to ‘‘explo- Inadequate supervision
sion/detonation’’ puts the risk man- Lack of communication
In March 2016, the Division of Chem- agement focus on a point in time
ical Health & Safety (DCHAS) pre- when the laboratory workers can still Listing failed or degraded barriers
sented an interactive symposium at respond and avoid a detonation alto- (such as those listed in the bullets
the American Chemical Society gether. Figure 3 presents the initial above) as threats is a common mistake
(ACS) national meeting. The sympo- bowtie diagram with the group-decid- in bowtie development. In response,
sium was held to assess the value of the ed hazard and top event. the workshop leaders guided the group
bowtie methodology in a laboratory Several potential consequences toward threats describing an initiating
setting.16 The example used in the ac- were suggested: event that if left unchecked could es-
tivity was based on the 2011 CSB case calate to a laboratory worker exceed-
study of a laboratory incident at Texas Injury/fatality due to detonation ing a safety critical amount of energetic
Tech University (TTU).1 Property damage due to detonation compound. After this instruction, the
To construct the bowtie, participants Reputation damage due to detona- group identified three key threats:
were asked to identify its parts in ac- tion
cordance with widely accepted best Loss of business/productivity/grant Intentional synthesis scale-up of en-
practice in the oil and gas industry in funding due to detonation ergetic material
the following order17,18: Regulatory/external review Inadvertent synthesis scale-up
Unauthorized (criminal) activities
1. Hazard/Top Event Potential consequences extend be-
2. All Consequences yond the individuals and laboratory Figure 5 shows the developing bow-
3. All Threats involved and can effect multiple tie diagram which now includes the
4. Preventive Barriers departments in the university, as ap- group-identified threats that could
5. Mitigation Barriers parent from the above list. Although lead to the top event (exceeding the
property damage and regulatory/exter- safety critical limit). Finally, the group
Initially a variety of possible top nal review would be valid conse- discussed and listed the various pre-
events were identified, but the hazard quences, due to the time constraints ventive and mitigative barriers. The
was generally agreed upon by all parti- the workshop leaders focused on three resulting bowtie diagram based on this
cipants as ‘‘energetic material,’’ Other consequences listed in the evolving workshop activity is shown in Figure 6.
possibilities for the top event discussed bowtie diagram shown in Figure 4. The bowtie diagrams in this manu-
were: Identifying potential threats was the script were prepared using BowTieXP
most challenging part of constructing a software.13
Intentional scale-up
Explosion/detonation
Exceed safety critical limit (In this
case, a safety critical limit is the min-
imum amount of material that could
cause permanent bodily injury if det-
onated)
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
Figure 6. Bowtie diagram for the detonation of energetic material in a laboratory setting.
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
barrier can degrade and fail. In the lab- can be the starting point of an acci- 3. Kemsley, J. Spark from pressure gauge
oratory setting, operational barriers are dent investigation to determine how caused University of Hawaii explosion,
necessary and heavily relied upon. This and why the barriers had failed, were fire department says. Chem. Eng. News,
inadequate, or were bypassed. 2016 Web Date: April 19, http://
is illustrated on the bowtie in Figure 6
cen.acs.org/articles/94/web/2016/
where the only barrier listed to prevent
04/Spark-pressure-gauge-caused-
‘‘intentional scale-up’’ is a policy not to CONCLUSION University.html [accessed 28.04.2016].
synthesize amounts greater than the 4. The Association of Public and Land-
safety critical limit. Bowtie methodology could prove use- grant Universities, A Guide to Imple-
ful in aiding universities to further menting a Safety Culture in Our
improve how they manage their labo- Universities. http://www.aplu.org/library/
USING BOWTIES TO FACILITATE ratory risks. It provides a structured safety-culture/file [accessed 31.08.2016].
RISK CONVERSATIONS approach to identifying key safety bar- 5. National Research Council. Prudent
riers and controls so their strengths Practices in the Laboratory: Handling
Bowties are part art and part science. and Management of Chemical Ha-
and conditions can be monitored more
There is no single ‘‘right’’ way to create zards, Updated Version;; National Aca-
effectively to prevent barrier degrada-
demies Press: Washington, DC, 2011,
a bowtie. As long as a group developing tion. These are the types of improve- doi:10. 17226/12654.
the bowtie leverages in-house exper- ments necessary to go beyond current 6. American Chemical Society. Identify-
tise and identifies their most significant laboratory safety practices and achieve ing and Evaluating Hazards in Research
hazards and the needed controls, then a proactive safety culture. Laboratories. 2015. http://www. acs.
the bowtie can be a useful tool to Bowties facilitate effective risk con- org/hazardassessment [accessed 06.09.
facilitate risk conversations amongst versations by providing a visual aid to 2016].
numerous stakeholders. For example: prompt conversation rather than rely- 7. Occupational Exposure to Hazardous
ing on sometimes cumbersome text- Chemicals in Laboratories, Definitions,
As the bowtie in Figure 6 indicates, 29CFR1910.1450(b). https://www.osha.
based policies, procedures and check-
gov/pls/oshaweb/owadisp.show_document?
mitigating potential consequences lists. Presenting information in multi-
p_table=STANDARDS&p_id=10106 [acces-
may require resources outside of ple formats supports learning by a sed 06.09.2016].
the chemistry department such a diverse group of workers whose learn- 8. National Research Council. Safe
public relations department to han- ing styles, languages skills and techni- Science: Promoting a Culture of Safety
dle media attention. cal expertise can be expected to vary in Academic Chemical Research; Na-
Bowties help depict whether a risk significantly. Universities can also ap- tional Academies Press: Washington,
management approach is focused on ply this technique as part of their audit- DC, 2014, doi:10.17226/18706.
preventing dangerous situations or ing process to help identify where 9. Torrice, M.; Kemsley, J. Patrick Harran
mitigating them after they have oc- barriers may need to be strengthened and L.A. District Attorney Reach Deal
curred. For example a fire hazard In Sheri Sangji Case. Chem. Eng. News,
or added, and this can drive the crea-
2014 Web Date: June 25, http://cen.
will not be controlled by PPE and tion of effective actions toward further
acs.org/articles/92/web/2014/06/
so a well-developed bowtie would risk reduction efforts. Ultimately, bow- Patrick-Harran-L-District-Attorney0.
highlight that PPE is only helpful ties provide a good roadmap for asses- html [accessed 31.08.2016].
in mitigating potential burn injuries, sing the quality and number of risk 10. Kemsley, J. AAAS rescinds election of
not in preventing the hazardous controls to prevent serious laboratory Patrick Harran as a fellow. The Safety
conditions that can lead to a fire incidents and facilitate the identifica- Zone by Chem. Eng. News, 2015
from occurring in the first place. tion of clear actions to improve con- Posted on December 23, http://
A graduate student or other labora- trols and advance the ‘‘cultural’’ side of cenblog.org/the-safety-zone/2015/12/
tory worker initiating a procedure safety management. aaas-rescinds-election-of-patrick-
can review an existing bowtie as a harran-as-a-fellow/ [accessed 06.09.
2016].
reminder of the critical barriers to
11. Bierman, P. The University of Ver-
ensure they are in place and func- REFERENCES mont, Cosmogenic Nuclide Laboratory
tioning before initiating work. 1. U.S. Chemical Safety and Hazard In- and Geomorphology Research Group,
During a laboratory safety audit, an vestigation Board (CSB), 2011. Texas Memorial Day 2007, was not such a
inspector without expert knowledge Tech University Laboratory Explosion: good day http://www.uvm.edu/cosmolab/
on a process and its hazards could Lubbock, TX, 2010, January 7, Report ?Page=fire.html&SM=about_sub_menu.
review a bowtie and ask about the No. 2010-05-I-TX. http://www.csb. html. [accessed 06.09.2016].
gov/assets/1/19/CSB_Study_TTU_.pdf 12. BowTie Pro, Home Page, http://
safety management systems con-
[accessed 06.09.2016]. bowtiepro.com [accessed 04.09.2016].
nected to the barriers indicated on
2. Kemsley, J. Researcher Dies After Lab 13. Software Products, CGE Risk Manage-
the diagram. This can include talking Fire: UCLA research assistant burned ment Solutions, http://www.cgerisk.com/
with the individuals responsible for in incident with tert-butyl lithium. software [accessed 04.09.2016].
monitoring and maintaining differ- Chem. Eng. News, 2009 Web Date: 14. Reason, J. Human error: models and
ent barriers. January 22, http://cen.acs.org/articles/ management. Brit. Med. J. 2000,
Incidents and serious consequences 87/web/2009/01/Researcher-Dies- 320(7237), 768–770.
occur when barriers fail. Bowties Lab-Fire.html [accessed 06.09.2016].
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003
JCHAS 921 1–7
15. The history of Bowtie, CGE Risk Man- 17. A Barrier Focused Approach: How to http://www.aiche.org/ccps/community/
agement Solutions, http://www.cgerisk. Get Started with Process Safety, Vol. 2 projects/project-237-guidelines-barrier-
com/knowledge-base/risk-assessment/ (Draft), ENFORM. The Safety Associa- risk-management-bow-tie-analysis [acces-
the-bowtie-methodology [accessed 04.09. tion for Canada’s Upstream Oil and sed 04.09.2016].
2016]. Gas Industry, 2016 (http://www. 19. ANSI/AIHA/ASSE Z10 Committee.
16. Mulcahy, M.; Boylan, C. Introduction enform.ca/files/pdf/process_safety/ Occupational Health and Safety Man-
to Bowtie Methodology for a Labora- Barrier_Focused_Approach_How_to_ ag. Syst. 2012 viii.
tory Setting. Presented at the 251st get_started_draft.pdf [accessed 10.06.
American Chemical Society National 2016].
Meeting & Exposition. Boston, MA, 18. Project 237: Guidelines for Barrier
March 15th, 2016. Risk Management (Bow Tie Analysis),
Please cite this article in press as: Mulcahy, M. B. et al31., Using bowtie methodology to support laboratory hazard identification, risk
management, and incident analysis. J. Chem. Health Safety (2016), http://dx.doi.org/10.1016/j.jchas.2016.10.003