ANSI/ ANS-2.12-1978 ANSV/ ANS-2.12-1978 guidelines for combining natural and external man-made hazards at power reactor sites No longer being maint American National StANSV ANS-2.12-1978 American National Standard Guidelines for Combining Natural and External Man-Made Hazards At Power Reactor Sites Secretariat American Nuclear Society Prepared by the American Nuclear Society Standards Committee Working Group ANS-2.12 Published by the American Nuclear Society 555 North Kensington Avenue La Grange Park, Illinois 60525 USA Approved July 25, 1978 by the American National Standards Institute, Inc.reek dn Ans Steel eel» mn onder Amarloan oni eng tatu i tri Naan Bente sea ee ee ee Standard szisunoe of an American Nationa! Standard does not in any sespack preclude rn Someta toes sear ee ee teeta So ee Soci crenata eetacene ae sedate era cs Ee casas pony nant Se WOM sara as casuer telat ssa oe crete aot neta recat we we a rma aie teens tel pees ool So eae tle vais onto soy Published by American Nuclear Society 555 North Kensington Avenue, La Grange Park, Illinois 60525 USA Price: $35.00 Copyright * 1978 by American Nuclear Society Any part of this Standard mey be quoted. Credit lines shoul read “Extracted from American National Standard ANSIANS-212-1978, with permission of the publisher, the American Naclear Society.” Reproduction prohibited under copyright convention unless written per mission is granted hy the American Nuclear Society, Printed in the United States of America oForeword ris force isnt pert of American National Stand © In May 1974, the ANS.2 Subcommittee, Site Evaluation, was given the respons for developing a standard to determine the combination of natural and external man- made hazards to be considered in plant design, A working group was formed and designated ANS-2.12, “Guidelines for Combit ‘Natural and External Man-Made Hazards for Power Reactor Sites.” A scope for this effort was developed by the working group as given in this standard. ‘The working group met in September and November 1974. In January 1975, the working group was selected to participate in an experiment on accelerating the establishment of nuclear standards which was sponsored by the American National Standards Institute (ANSI) using funds provided by the U.S. Nuclear Regulatory Com- mission. Battelle-Northwest conducted the experiment under contract with ANSI. The key ingredient of the experiment was a one-week working session at the Battelle Seat tle Research Center with Battelle providing an executive secretary and clerical assistance. The first draft of the standard was a product of these three meetings. ‘The working group felt that this standard should attempt to: (1) include a thorough candidate list of the natural and external man-made hazards to be considered. (2) establish a probability approach for combining natural and external man-made hazards, (3) develop a set of exclusion criteria which would allow candidate hazards or hazard combinations to be judged as not requiring consideration in nuclear power plant design. (4) provide a methodology fer combining hazards which designers could use. (5) give a list of standard hazard combinations. € {6) shere possible, efer to other existing ANSI standards for characterization of the individual natural hazards, (7) provide @ reference document for external man-made hazards, and (8) attempt to provide a probability level for hazard combinations where possible. ‘The scope and these guidelines have been followed in this standard. In preparing this standard, it was found that data on the probability of occurrence and ~ the duration of natural hazards are very scarce, particularly for the extreme “probable maximum events.” The ANS-2.12 working group partially circumvented this status by _— using the concept of recurrence intervals for the natural hazards which occur more frequently than the “probable maximum events.” It is left to the designer to determine the magnitude of an event with a given recurrence interval for his site. More work needed in defining the probabilities of occurrence of severe or extreme phenomena. Another area which needs more evalu: jon is whether, and to what extent, natural harards cause or increase the probability of external man-made hazards, as for exam- ple, the effect of ice or fog on airplane accidents. Similarly, research is required on the degree of dependence or effects of one natural hazard causing another, as for example, the relationship between a tornado and rain. ‘The ANS-2.12 Working Group wishes to clearly state that it is difficult to precisely establish the probability of occurrence of natural and external man-made hazards. The phenomena are complex and the probability of each is a function of parameters such as geographical location, time of year and nature of the hazard. There are different Probabilities for the various intensities of a hazard. When one takes the next step, to~ develop a methodology of combining these hazards and attempts to use a probability acceptance limit for the combination, as was done, the task is even more difficult In this standard, the working group took what they consider to be an engineering ap. © proach. It is believed some guidance on hazard combination is needed relativelyquickly for nuclear power plant designers. The working group saw that in some areas designers are working on the basis of conservative assumptions in states of un- certainty, where the data are incomplete and not perfectly known. The working group therefore used their judgment in setting down the bounding values for the frequency of ‘occurrence of certain hazards and their durations. In some instances estimates were made based on the group’s collective experience. Summaries of data on hazards and published methodologies for estimating their probabilities of occurrence were reviewed. Where possible, values were used which are already generally accepted in the nuclear industry or which are believed to be acceptable, based on past licensing ac- tions. When the standard hazard combinations in this document were derived, they were not further examined for their consequences to see if the damage which they would cause to a nuclear plant is acceptable. This will involve further refinement and study by the designer. Only hazards external to the plant are considered in this standard. In-plant hazards such as internal missiles and pipe ruptures are not included. Sabotage is not included asa man-made hazard. ‘This standard does not address the effects of a combination of hazards on a nuclear plant, nor does it consider repair times ar the plant process condition which is ap- propriate to the hazard combination. Further guidance on these matters is expected to be addressed by the reactor criteria documents under development by ANS-50, Power Reactor Systems Committee.” In defining durations for hazards, the repair time of the plant is not included because the plant operating state subsequent to the event will be significantly altered by the original event. Hence only the duration of the event itself as specified in the standard is used. For example, if an SSE occurs the plant will be shutdown for inspection and evaluation. ‘This standard docs not include the development of specific loading combinations, load factors or behavior requirements to be used in design because these criteria are to be determined in accordance with the applicable codes or standards governing the design of the specific systems, components or structures being considered. In addition, it should be understood that the rational development of design requirements, not other- wise covered by applicable codes and standards, includes an evaluation of the con- sequences as well as the probability of oceurrence of all hazards, both man-made and natural, and external and internal to the reactor system. ‘The working group believes the approach used in this standard will be useful to nuclear plant designers in developing severe and extreme load design criteria, Working Group ANS-2.12 of the Standards Committee of the American Nuclear Society had the following membershi urgent & Lundy Engineers S.J. Milioti Chairman, American Electric Power MK. Ravindra, § Schulte, Stone de Webster Engineering Cor- ‘Service Corporation RO. Barnett, Tennessee Valley Authority P. Gotthch, Damen & Moore eneral Electric Company LL Beratan. US. Nuclear Regulatory Commission ‘Offshore Power Systems 6 Kost, Bnaineering Decision Anatvais Compans, J. H. Wiggins, oH. Wiggins Company The Plants” and draft standards under develop Design of Stationary Gas Cooled Reactor Plants” an ANS-4.1( an EMEBR Nuclear Power Plant” “Contributions were also made by W.R. Sugnet of Westinghouse Electric Corporation and CE. Johnson of the US. Nuclear Regulatory Commission who served as members of ANS 2.12 for « majo portion of this ef fort. NG. Wittenbrock of Battelle, Pacific Northwex: Laboratories served as ecutive secretary during the ANSE sponsared experiment in writing nuclear standards‘The membership of ANS-2, Site Evaluation, at the time of approval of this standard was; RV. Bettinger, Chairman, Pacifie Gas & Electric Company E. Becalante, Los Angeles Department of Water “€ Power 4A Fischer, Dames 4 Moore MLL Goldman, NUS Corporation W.W. Hays, US. Geological Service GE. Heim, Sargent & Lundy Engineers GF Hoveke, Sargent & Lundy Engineers G.S Hunt, Southern California Edison Company D.H. Johns, Southern California Eiison Company U. Kappus, Dames Moore RB Keever, Nuclear Technology, Ine Ed. Keith, EDS Nuclear Inc CR McClure, Bechtel Power Corporation ‘The American National Standards Committee N18, Nuclear Design Criteri following membership at the time it reviewed and approved this Standard: Organization Represented ‘American Chemical Society American Concrete Institute ‘American Nuclear Society ‘American Society of Civil Engincers American Society of Mechanical Engineers ‘American Public Health Association, Ine. ‘American Welding Society ‘Atomic Industrial. Foro Blectric Light & Power Group Pederal Power Commission Health Physics Society Institute of Elec cal & Rlectronice Engineers Institute of Nuclear Materials Management [Nuclear Energy Liability & Property Insurance Association US. Baviconmental Protection Agency US. Energy Research & Development Administration US. Nuclear Regulatory Commission Individual Members 8. J. Miliot, American Blectrie Power Sereice Cor poration GW. Nicholas, Dames & Moore Noble, Dames Moore Pickel, Oak Ridge National Laboratory RozPserescas had the Name of Representative CE. Stevenson PL. Roggenkamp (alt) PE Mast, GL. Wessman M. 1. Goldman ©. Gogotck (ale) 3.8 Bitel 1H. Helyoak (ale) 4. R. Coleman JR McGutfes G. Sherwood 3 Howard 3. E Sohngen tale) ‘AP. Donnell BLP. Chew (all) JM. Smith, Je RL. Clark (alt) RG, Benham Bitharz, Jr (alt) “SAtendt LOA Strom (alt) Catal 1. Magiani (ale) fe Salivan G_ Burley (alt) FX. Gavigan GA. Arlotto 1M. Kehnemasa (alt) SEN. Cramer RJ. Creagan JF. Gibbons TD. dones Td. Pashos D.R Patterson, AH. Redding G.C. Robinson RP. Schmit JF West2. Definitions ........2...0005 3. Introduction . 4. Natural and External Man-Made Hazards ........... 4.1 Exclusion Criteria for Individual Hazards 4.2 Hazards Usually Excluded 43 Consideration of Other Haz 8. Characterization of Hazards 5.1 Dependency of Hazards. 6. Methodology for Combining Hazards........ ? 6.1 Exclusion Criteria for Combination .- 6.2 Computation of Independent Event Combination Probabilities. 6.3 Computation of Probabilities for Event Combinations Involving Dependent Events ... 7. Standard Hazard Combinations. . 7. Criteria for Using List of Standard Combinations 7.2 List of Standard Hazard Combinations.............- 8. Design Approach for Combining Hazards. 8.1 Design Approach Sequence ....... 9. References. ‘Appendix A. ‘Mggeailtx Busteret tcc. Ser Figure 1 Dependency Reletioahipe of Natural and Man-Made Figure 2 Flow Chart for Design Approach Sequence .... Figure C-1.. Figure C-2.. ay LSGuidelines for Combining Natural and External Man- Made Hazards at Power Reactor Sites 1. Scope This standard presents guidelines which will allow power reactor designers to select (1) com- binations of natural hazards, (2) combinations of external man-made hazards, and (3) com- binations of natural and external man-made hazards to be used in the design of power reac- tor structures, systems, and components. ‘The hazards to be considered in these categories are identified. Probability level acceptance criteria are defined which will enable designers to select the combinations of hazards which are design bases for a particular site. Methods for caleulating probabilities of combinations are discussed. definitions will be useful in understanding this standard. aircraft impact. Accidental impact of an air- craft into a safety-related structure, system or component such that the resulting missile, fire, or smoke could affect the ability of the struc ture, system or component to perform its in- tended safety function, combined event. An event consisting of the simultaneous occurrence of two or more natural or external man-made hazards. dependent events. Event combinations for which the occurrence of one event gives in- formation about (increases or decreases) the possibility of the occurrence of the other event. For example, the occurrence of a hazard such as, an earthquake may increase the chance of an ex- plosion at a nearby industrial facility. extreme environmental load. Load which is credible but highly improbable. flood-coastal. Abnormally high water on open and semi-enclosed bodies of water resulting from storm surge and tsumami, precipitation, tide, wind-wave activity, and possible flood at nearby stream. food-lake. Abnormally high water on enclosed bodies of water resulting from high lake level, storm surge and seiche, precipitation, wind-wave activity, and possible flood of nearby stream. flood-river. Abnormally high water on an inland stream resulting from precipitation and snowmelt runoff, possible ice blockage, wind- wave activity, and possible dam failure or stream diversion. independent events. Event combinations for which the occurrence of one event does not give information about (increasing or decreasing) the probability of the occurrence of the other. If two events A and B are independent, the conditional probability for the occurrence of A given the oc- currence of B is simply the probability for the occurrence of A alone (unaffected by the oc- currence of B). industrial or military facility accident. Ex- plosion, missile, fire, toxic gas release or other potential hazard from a fixed facility. man-made hazard. An accident involving vehicles, equipment, or structures created by man which occurs external to a nuclear power generating station and has the potential for causing damage to safety related structures, systems, or components or a nuclear power generating station. missile. A mass which has kinetic energy and is, unrestrained. mutually exelusive. Two or more events which cannot physically occur simultaneously. natural hazard. A natural phenomenon which has the potential for causing damage to safety. related structures, systems, or components of a nuclear power generating station, operating basis earthquake (OBE). That‘American National Standard ANSUANS2.12-1978 ‘earthquake which, considering the regional and local geology and seismology and specific characteristics of local subsurface material, ‘could reasonably be expected to affect the plant site during the operating life of the plant; it is that earthquake which produces the vibratory ground motion for which those features of the nuclear power plant necessary for continued operation without undue risk to the health and safety of the public are designed to remain fune- tional. (1]" pipeline accident. Rupture of a pipeline carrying a gas or liquid under pressure which can explode or ignite or create a toxic gas cloud or environment which ineapacitates personnel or degrades equipment operation. probability of occurrence (frequency of oc- currence). The mean annual rate of occurrence of a hazard parameter within a range of values. ‘The limits of this range are indicated by the definition of the event. These limits can be specified to consistently cover the entire spec- trum of parameter values as illustrated in Ap- pendix A. For example, a 100-year wind means the probability of exceeding this wind speed in a year is 1/100. For purposes of this standard, all hazards have their parameter values? divided into segments which are characterized by discrete design events (such as Safe Shutdown Earthquake (SSE). retaining structure failure. Failure of a dam, levee, breakwater, seawall, or similar confining structure, due to causes other than flood, earth- quake or tsunami (since the retaining structure failure consequences must be included in the consideration of these natural hazards). safe shutdown earthquake (SSE). That ‘earthquake which is based upon an evaluation of the maximum earthquake potential con- sidering the regional and local geology and seismology and specific characteristics of local subsurface materials. It is that earthquake which produces the maximum vibratory ground motion for which certain structures, systems, and components are designed to remain func- "Numbers in brackets refer to corresponding numbers in Se: tion 9, References. Parameter value characteriagy an event a to its intensity eg. wind yelcity for a tornado, oF elevation of a flood. tional. These structures, systems, and com- ponents are those necessary to assure: (1) the integrity of the reactor coolant pressure boundary, (2) the capability to shut down the reactor and maintain it in a safe shutdown condition, or (3) the capability to prevent or mitigate the consequences of accidents which could result in potential offsite exposures comparable to the guideline exposures of 10 CFR 100. [1] standard project flood (SPF). The hhydrograph representing runoff from the stan- dard project storm (or snow melt, or both,) as specified by the U.S. Army Corps of Engineers. seiche. A standing wave oscillation of an en- closed water body that continues, pendulum fashion, after the cessation of the originating force, which may have been either seismic or at mospheric. severe environmental load. Load that could infrequently be encountered during the ‘operating life of the nuclear power plant. storm surge. A rise above normal water level on the lake, open sea or coast, due to the action of wind and atmospheric pressure. surface vehicle explosion. Accidental ex- plosion of land or water vehicles, including barge, truck, and railroad car. surface vehicle impact. Accidental impact of a surface vehicle with or near a safety-related plant structure, system, or component caused by an out-of-control vehicle due to operator error, vehicle failure, or due to natural hazards such as a flood. tornado. A violently rotating column of air pen- dent from a convective type cloud and nearly always observable as a funnel cloud or tube. Tornadoes have large rotational wind speeds, pressure gradients along their radii and trans- lational movement. A tornado can create struc- tural loadings and, in the case of a floating nuclear plant, can create plant motions. A tor- nado has the potential for creating missiles, the characteristics of which depend on the intensity of the tornado, The design criteria are depen- dent on the geographical location of the site forwind and missile velocities and differential pressure drops, and on the physical make-up of the site for the missile spectr tsunami. A series of sea or lake waves produced from the displacement of water by either a local or distant submarine earthquake, volcanic erup- tion, submarine or coastal landslide. A tsunami may cause flooding, loss of cooling water (on flowback or drawdown), impact loads from ‘waves or floating debris, or both, and erosion of earth foundations from structures. ultimate heat sink. The water source for the cooling water system of a nuclear power ‘generating station which performs two principal safety functions: (1) dissipation of residual heat after reactor shutdown, and (2) dissipation of residual heat after an accident. It can include a ‘complex of water sources, including necessary retaining structures (e.g, a pond with its dam, or a river with its dam), and the canals or con- duits connecting the sources with the cooling water system intake structures for a nuclear power generating rtation, Cooling towers or por- tions thereof which are required to accomplish the ultimate heat sink safety functions are in- cluded in this definition. waves. Surface motion in an ocean or large body of water caused by winds, barometric pressure anomalies (e.g., squall lines, etc.) and seismic disturbances. For floating nuclear plants, waves can induce plant motion which much be considered in plant design. Waves in- crease the effective water depth to be considered in the design of a plant sited on or near a large body of water. The dynamic effects of waves are considered in the definitions of floods in American National Standard “Standards for Determining Design Basis Flooding at Power Reactor Sites,” N170-1976 (ANS-2.8). [2] 3. Introduction It is current practice in the design of nuclear power generating stations in the United States to protect against the effects of natural and ex- ternal man-made hazards which can occur at the plant site. Certain natural hazards and how their individual effects are evaluated are discussed in ANSI/ANS-2.8-1977(N170), [2] and ‘American National Standard ANSI/ANS.2.12-1978 in draft standards under development? A similar series of standards have not been written, for external man-made hazards. However, a discussion of the missile effects of surface vehicle and aircraft accidents is presented in a proposed draft standard* It is the purpose of this standard to establish a methodology for identifying combinations of natural hazards and external man-made hazards, for consideration in plant design. As existing standards do not cover all of the individual ex- ternal man-made hazards, this standard can also serve as a reference which a designer can use in examining a specific site for protection against individual external man-made hazards. ‘The concept of an acceptably low probability of ‘occurrence is used in this standard as the basic criterion to establish a division between com- binations of natural hazards and external man- made hazards which must be considered in the plant design and those which need not be con- sidered. sidelines for Power ‘and Other Extreme Wind Parameters at Power Sites.” ANS.23, Assigned Correspondent: George Faulting at Nuclear Power Reactor Sites,” ANS.27. Assigned Correspondent: Cole Fe McClure, Bechtel Power Corpuraton, Papel American Standard "Standards for Evaluating Water Supply tnd Waterborne Ragncige Trpor for Power eset Sites: Ground Water," ANS-29. Assigned Correspondent DIL Siefken, Sargent & Lundy Engineers, Proposed American National Standard “Guidelines for Determining Ground Water Supply/Ground Water Transport of Radionuclides for Power Reactor Sites,” ANS-2.17- Assigned Correspondent: PJ. Fiyan, Bechtel Corporation: Propemet ‘American National Standard "Guidelines for Determining ‘Taunami Criteria for Power Reactor Sites” ANS-24 Assigned Correspondent: Norman R. Wallace, Bechtel Power Corporation. Proposed American National Standard “Stan- dards for Eyalusting Surface Water Supply for Nuclear Power Sites” ANS-2.13. Assigned Correspondent: Ulich ‘Kappus, Dames & Moore; Propared American National Stan- dard “Guidelines for Evaluating Surface Water Sup plySurface Water Transport of Teadionuclides for Power Reactor Sites,” ANS218. Assigned Correspondent: Ulrich ‘Kappus, Dames and Moore ‘Proposed American National Standard “Plant Design‘American National Standard ANSI/ANS2.12-1978 ‘The overall safety risk to the public due to a hazard is concerend not only with the probability of occurrence of the hazardous event but also with the probability that an essen safety function is significantly impaired by the event such that there is release of radioactivity. Specific evaluation of such consequences. is beyond the scope of this standard, but should be considered in the final selection of combined events to be used for plant design. Furthermore, the functional impairment implied by the con- sequence (of the initiating event considered here) could arise in other ways. Any attempt to estimate consequences from external hazards ‘must not produce only partial estimates which could be misleading. For example, if an external event causes loss of all off-site power, the failure rate of diesel generators. may enter into Equation 3-1 below. ‘The probability of the consequences is expressed by considering three distinct probabilities: (1) P(o), the probability of the occurrence of the particular hazard event, (2) P(ffo), the con- ditional probability of significant impairment of safety function, given the occurrence of the hazard event, and (3) P(c/f, the conditional probability of significant consequence to the Public, given the impairment of function. The probability of significant impairment of function P() = Piflo) P (0) (Eq. 3-1) and the probability of consequence to the public Pie) = Plelf) P(A) Plolf) P(f/o)P(o). (Eq. 3-2) ‘This standard is primarily concerned with examining hazard combinations based on ac- ceptably low values of P(o) alone. Other industry working groups are assigned to consider how the combinations of hazards in this standard are accounted for in design. The methodology described in this standard screens ‘out those hazard combinations which have oc- currence probabilities of less than 10°* per year. ‘Though extreme events such as the SSE and PMP are determined deterministically, ap- proximate probabilities are given for these ex- treme events for purposes of screening those jons of natural and man-made events which should be considered for design purposes. For those combinations of events for which the occurrence probability is equal to or greater than 10°® per year, a best estimate of the con- sequences of the combination of hazards shall be made. If this estimate shows that the con- sequences are unacceptable (for example, if radiological consequences to the public exceed Title 10, Code of Federal Regulations, Part 100 [1], guideline doses), then the plant shall be designed to make the consequence acceptable. If the evaluation shows the consequences are ac- ceptable, then the hazard need not be con- sidered in the design basis. 4. Natural and External Man- Made Hazards ‘The list below gives hazards which shall be con- sidered in the development of a design basis for combinations of natural hazards, combinations of external man-made hazards, and com- binations of natural hazards with external man- made hazards for power reactor sites. Natural Hazards Earthquake Flood-Coastal* Flood-Lake Flood-River Rainstorm* ‘Tornado ‘Tsunami Wind* “Includes hurricane effects External Man-Made Hazards Aircraft Impact Industrial or military facility accident Pipeline accident Retaining structure failure Surface vehicle explosion Surface vehicle impact ‘These hazards occur external to the plant and could act on safety-related structures, safety- related equipment located outside the plant structures (eg, piping and tanks) or the ultimate heat sink either by direct contact, or as a resultant effect of a hazard striking nearby which could prevent safety-related systems or components from performing their intended safety functions‘This standard assumes that the design bases for safety-related structures, systems, and com- ponents account for applicable individual natural or man-made hazards. ‘Natural events more frequent than the 100-year mean recurrence interval, which may be com- bined with other extreme events, eg. the SSE and a 25-year flood, are not addressed in this standard because such high recurrence interval events are normally accounted for in the design as ambient or initial conditions. 4.1 Exelusion Criteria for Individual Hazards. Other hazards have been considered and have been eliminated from the design com- binations in this standard. Each of the following criteria provides an ac- ceptable basis for excluding hazards from the evaluation to determine design combinations of hazards: 1. The phenomenon is of equal or lesser damage potential to the plant than other phenomena against which the plant has been designed. 2, The phenomenon by itself has a probability of occurrence less than 10-® per year. 3, ‘The phenomenon is shown to be unable to occur close enough to the plant to affect it. 4. The phenomenon is included in the definition of another phenomenon. For example, storm-surge and seiche are included in lake flooding; toxic gas is included in pipeline ac- cident or industrial or military facility accident. 42 Hazards Usually Excluded. The list of usually excluded hazards is given in Table 1. ‘The hazards in Table 1 shall be reviewed for the particular site and plant design, using the ex- clusion criteria of 4.1, to confirm to the extent required that these hazards can be eliminated 4.3 Consideration of Other Hazards. It is believed that the hazards identified in Section 4, “Natural and External Man-made Hazards,” and Table 1 form a complete and sufficient list of natural and external man-made hazards which are important to the safety of a nuclear power plant. However, should some other hazard be identified as an important hazard, the designer shall consider it using the methods of this standard. ‘American National Standard ANSI/ANS.212-1978 5. Characterization of Hazards ‘Table A-1 in Appendix A gives the parameters for the natural hazards and Table A-2 in Ap- pendix A presents the external man-made hazard parameters. Appendix A discusses the basis for the values in these tables, gives the methodology for com- puting the probability of occurrence for each in- dividual hazard, and cites the references used. ‘The discussions in Appendix A are for in- formation only; the appendix is not a part of the text of this standard. There are several methods of calculating the characteristics of a natural or man-made hazard and the characterization of hazards is site dependent. Although the values in Tables A-1 and A-2 eon- tain a large measure of conservatism and for most sites are considered upper bound values, the designer shall not consider these values as realistic parameters for the individual hazard for his site. Realistic values for probability and duration of individual hazards are beyond the scope of this standard, and where such values are needed, they shall be determined by a site specific analysis. 5.1 Dependency of Hazards. Figure 1 shows a matrix of the dependency relationships of hazards. ‘This figure was used for committee calculations as described in Appendix D. Natural hazards do not cause or occur con- currently with other natural hazards in Figure 1, except for wind and rain which occur together during a hurricane. Some other natural hazards are mutually exclusive, e.g, the various flood hazards. Still other natural hazards show no dependency since they have different effects, eg, the 100-year rain is not the same rain which is used in calculating severe floods (e.g, standard project flood, (SPF). The dependent aspects of some of the actual hazards listed in Figure 1 are already included in the broad definitions of this standard. Therefore, the hazards are shown as independent in Figure 1. For example, the flooding effects of a long duration rain on a river or river basin are already included in the definition of river floodAmerican National Standard ANSI/ANS2.12-1978 ‘Table 1 Individual Natural and External Man-Made Hazards Usuall Excluded from Combinations of Hazards Exclusion Criterion No. As Defined Hazard in Seetion 41 Remarks High summer I temperature . Ultimate heat sink is designed for at least 30 days of operation, including consideration of evaporation, drift, seepage, and other water losses. (See Note 2.) Waterspout 1 Tornado governs. Loading due to water in spout not governing. Sandstorm 1 Tornado and windstorms cover this. Blockage of air intakes with particulate matter should be considered. (See Note 2.) 3 For most sites in the US. Fog H Could, however, increase probabi of man-made hazard involving surface vehicles or aircraft. Forest fire 1 Site eared for such fire. Con room habitability required for ‘smoke and protection of diesel air intakes. at Drought . Assumes multiple source of ultimate hheat sink or ultimate heat sink not affected by drought; e., cooling tower with adequately sized basin Lightning 1 Lightning protection is considered in the plant design. Does not have additive stress effects on structures ‘when combined with other hazards {See Note 2) | Frost 1 Snow and ice govern | Meteorite a Less than 10-6 per year itself. (3] “Thew natural hazards provide some of the design bases for the ultimate heat sink. iRiver diversion Soil shrink-swell ‘consolidation Low lake or river water level Toxic gas Avalanche Landslide Low winter temperature? High tide, high lake level, ‘or high river stage Waves “Those natural hazards provide some of the design bases for the ultimate heat American National Standard ANSI/ANS2 12-1978 ‘Table 1 — continued Exclusion Criterion No. ax Defined Section 4.1 Remarks Other missiles govern (See Note 1) Included in effects of coastal flood. ‘This event is considered in the ‘evaluation of the ultimate heat ‘sink. Should diversion become 1 hazard, adequate storage is provided, (See Note 1.) Ultimate heat sink is designed for at least 30 days of operation, including consideration of evaporation, drift, seepage and other water loses. Control room habitability required for toxic gas accident. This, assumes no operator action outside the control room is required to render the consequences of the event acceptable. For most sites in the U.S. See Ref. [4,5] for discussion of avalanches. For most sites in the U.S. ‘Thermal stresses and embrittlement rot significant or they are covered by other codes or standards. (See Note 2.) Included under flood. Included under flood.American National Standard ANSUANS2.12-1978 ‘Table 1 — continued Criterion No. as Defined Hazard in Section 4.1 Remarks Seiche 4 Included under flood. | Precipitation Regional 4 Included under food. Storm surge 4 Included under flood. Tee cover 1 See remark under snow for roof loading. Tee effects on intake structures on northern lakes ‘may require design consideration, 4 Tce blockage of rivers causing flooding is included under river flooding, Tce blockage causing loss of cooling water flow must be considered in design of intake and discharge structures, Snow si When snow (or ice) load in excess of design live loads is considered in the design of power plant structures, the resulting load combination is treated as an extreme environmental condition with unit load factors. In general, the normal design load (ex, U = L4D + 17L) controls in comparison to the total load of U = D + S, where D = dead load, L = live load, and S = snow (or ico) load. See Ref. (6,7] 4 Snow melt which causes river flooding is included under river flooding. Notes to Table 1 (1) Siterelated characteristics, such as subsidence due to subsurface pumping, mining, cavities, o alteration of ground water regimes; liquifaction potential, or sensitive soils and rocks which have expansive, heave shrink-swell characteristics are natural Phenomena which shall be considered as potential hazards These potential hazards shall be considered and eveluated during the ‘Ste suitability evaluation procesa. Characteristics either reslt in (a) the site being unsuitable or (b) proper design consideration fand construction techniques are employed such that they are no longer considered hazards. Under the latter ease, slope stability ‘and differential settlement and their resulting lond-causing phenomena shall be given proper design consideration similar to any ther structure or system. As such these siterelated characteristes are not considered natural hazards and ned not be cone Sidered in the development of combinations of natural and external man-made hazards. (2) Environmental hazards such as this which could affect plant equipment and instrumentation are normally considered in the plant design separately as environmental protection© ‘The causal relationship of natural hazards to man-made hazards is difficult to establish. For example, it seems reasonable to assume that wind or rain could have a causal relationship to a surface vehicle accident (e-., surface vehicle explosion or impact). The data available have not been analyzed sufficiently to establish that wind or rain caused the surface vehicle hazard. Very few data analyses are available concerning ‘American National Standard ANSY/ANS-2.12-1978 the influence of natural hazards in causing man- made hazards, However, the data base for man- made hazards, limited as it is, does include the influence of natural hazards. In many instances where particular natural hazards exist, additional precautions are taken to reduce the probability for a man-made hazard to occur. For example, additional navigational Figure 1 Dependency Relationships of Natural and Man-Made Hazards £ i 5 ale a - z il: 3 Flale i g/2|2|2 g/4a/2 4 . sis {slele|s]ele t7ilalels Flood Coastal wIMTr}a}a{a]a T}1] >} sol sao s | 6 | 6 | 5 | 6 elsirie Flood-bake MiTi}i}a}a]o Tt] >| as biol 5 | 6 | 5 | 5 | 6 cls|7leé Flood River Thr} 1 ]rys Tt] P| sielslsle eisi7ie Rain t/t] > 1 |r) |] as 6 | 6 | 16 9 slalele Tarmado TPM] a T|1) >| a é| s/o éisi7ieé Trane r]a T|1) >| as | éisi7ie Wind D T{1) >| 8 cisi7leé ‘Aisa Tapact m[rle ts bal 6s] 6 | 6 Milaary ov Teatal T{iyiyo Pacilty Accident slelele Pipeline Accident T]r]o slele Retaining Structure 1) Faire oie Surface Vehicle 13 Impact 5 D = Dependent 1 = Independent M = Mutually Exclusive (Numbers refer to notes on following page)American National Standard ANSYANS2.12.1978 Notes for Figure 1 General: Notes 1-4 pertain directly to dependency/independency relationships. Notes 59 gives baves for entries D1, or M in Figure 1 (1) Dependency for hurricane only. Otherwise independent. If dependent, note 7 applies instead of note 6 (2) Collapse of river dams includes potential barge impact on retaining structures (3) If impacting vehicle can also cause significant explosion effets, relationship is "D” instead of "I", and note 7 applies in- stead of notes 5 oF 6. (4) Effects of rain and earthquake included under various flood definitions (6) Hazard precluded by. or included in, definition of other hazard (6) Hazard recognized as having no general dependency relationships, (7) Hazards gonerally recognized as having a reasonably probable dependency relationship. (8) Effects of hazard combination are normally not concurrent or are non-additive on plant structures, systems, or com- ponents. (2) Although the natural hazard can increase the probability of the man-made hazard occurring. insufficient quantitative data are available to determine any dependency relationships. In addition, carrective measures are normally taken in instances ‘where natural hazards influence man-made hazards, eg. increased lighting, better navigational ads, etc. Furthermore, during severe weather conditions, aircraft and surface vehicle movements are normally reduced near the hazard by grourving or re- routing (10) For sites on estuaries oF on lakes near stream mouths, see Sections 92 systems are installed at airport facilities and shipping lanes where poor visibility (due to rain, snow, fog, etc.) frequently exists. For con- servatism, however, several man-made hazards have been treated as being dependent on rain or wind. For other natural phenomena, such as earth- quake, flood, or tsunami, a dependency relation- ship was also assumed as shown in Figure 1 for any combination where at least one example at 1 hypothetical site could be provided. For in- stance, floods could cause surface vehicles (barges) to impact an intake structure at the ultimate heat sink. As an upper bound for those combinations, complete dependency was used in the sample calculations of Appendix D, which led to the standard hazard combinations in Sec- tion 7, “Standard Hazard Combinations.” The assumption of complete dependency for those combinations is extremely conservative, however. Because the majority of nuclear power generating stations are not located near poten- tial man-made hazards, it is not expected that these dependent events will require con- sideration for most designs. In the case where such potential man-made hazards do exist for a site, it is expected that the designer will specifically evaluate the strength of such a dependency to determine whether the com- bination can be neglected for design purposes. 6 Methodology for Combining Hazards is For the purpose of this standard, event com- 10 ind 9.2.3 of Ref. [2] binations are considered to occur over a time in- terval not exceeding the shorter of the durations of either event. 61 Exclusion Criteria for Combination. Combinations of two or more natural or ex- ternal man-made hazards need not be con- sidered in plant design, if one or more of the following criteria are satisfied: 6.1.1 Acceptably Small Probability of Oc- currence. If the probability of simultaneous oc- currence of the combination of hazards is equal to or less than 1 x 10° per year at the plant site, the combination need not be considered in plant design. The basis for selection of 1 x 10° per year as an acceptably small probability of oc- ‘currence is discussed in Appendix B. j 6.12 Hazard Effects Are Nonadditive. Where the effects due to the individual hazards in a postulated combination do not produce loads on the same part of the plant, the need to consider this combination in the plant design is eliminated by giving due consideration to each of the individual hazards comprising the postulated combination. 6.1.3 Effects Bounded by Another Design Combination. Where the effects on the plant due to one combination are determined to be less severe than the effects of another com- bination, which are part of the plant design basis, then the documentation of this deter- mination constitutes the necessary con- sideration of the less severe combination in the plant design. 6.14 Mutually Exclusive Hazards. Com- binations of hazards which cannot occur simultaneously due to the physical laws ofnature shall be excluded from consideration in plant design. 6.2 Computation of Independent Event Combination Probabilities. The probability for the simultaneous occurrence of two in- dependent events is: PpPp P( 2 events combination] = where pi and po are the probabilities of occurrence, per year, of the two events; ty and tz are the event durations (in minutes); Y is the number of minutes in one year. This formula is valid when it is assumed that the durations of these events (tx and tg) are much less than one year and pi and p2 << 1. The derivation is given in Ap- pendix C. ‘The probability fot the simultaneous occurrence of three independent events is given in Appendix , but is not presented here as it was not used in any of the calculations. 6.3 Computation of Probabilities for Event Combinations Involving Dependent Events. ‘The probability for the simultaneous occurrence of events 1 and 2 is P (2 events combinations) POMP, (Eq. 6-2) where P (2/1) is the conditional probability for the oc- currence of event 2, during the occurrence of event 1, given the occurrence of event 1. Quantifications of the degree of dependency be- tween natural hazard causes and man-made hazard effects are generally not well established. In addition, because these conditional probabilities may be site specifi, it is up to the plant designer to estimate the conditional probabilities (See Ref. [8]) connected with the most serious dependent event combinations. This standard provides a methodology for the plant designer in which he can use the range of values from completely dependent events to completely independent events. For completely American National Standard ANSI/ANS2.12-1978 dependent events Pan) = 1, (Eq. 6-3) while for completely independent events P (2/1) = po (tr + te) (Eq. 6-4) The use of a probabilistic approach different from that given in this standard shall be sup- ported by the analysis used to derive that ap- proach. 7. Standard Hazard Combinations ‘The list of standard hazard combinations which results from calculations using the hazards of Section 4, “Natural and External Man-Made Hazards,” the values of Tables A-1 and A-2 and the methodology, equations and exclusion criteria in Section 6, “Methodology for Com- bining Hazards” is given in 7.2 below. The calculations leading to these combins given in Appendix D. 7.1 Criteria for Using List of Standard Combinations. (1) The designer shall use only those hazard combinations which are applicable to the specific plant site, Where the man-made hazard is dependent on a meteorological phenomenon the hazard is only applicable if that specific phenomenon is credible at the site, For example, flammable gas from a pipeline accident is a hazard only if the gas can reach the plant. (2) Standard hazard combinations need not be considered if site specific calculations demon- strate occurrence probabilities equal to or less than 10 per year. - (8) Structural or equipment failures cannot be initiated by events, or combinations of events, against which they have been properly designed. (4) For hazard combinations involving dependent events, it is expected that the designer will specifically evaluate the strength of the dependency to determine to what extent the combination should be considered for design purposes. (5) The resultant loads from these hazard combinations shall be considered as belonging to either the extreme environmental load or the severe environmental load category. uAmerican National Standard ANSW/ANS-212-1978 (6) The hazard combinations list shall be reduced or expanded by a designer, using the methods described in Section 8, “Design Ap. proach for Combining Hazards.” 7.2 List of Standard Hazard Combinations. Natural Hazard Combinations Extreme Environmental Load Category OBE + SPF (River) SPF (River) + 100-year rain storm SPF (River) + 100-year wind Natural and Man-Made Hazard Com- binations.* Extreme Environmental Load Category SSE + military or industrial facility ac- cident SSE + pipeline accident SSE + surface vehicle explosion Probable Maximum Precipitation (PMP) local + aircraft impact Probable Maximum Precipitation (PMP) local + surface vehicle impact Severe Environmental Load Category SPF + surface vehicle impact 100-year tsunami + surface vehicle impact 100-year wind + surface vehicle impact 100-year wind + aircraft impact Man-Made Hazard Combinations None It is noted from the above list that only pairs of events are included. Because the highest bined probability of any of these pairs is on the order of 10 per plant per year, another significant hazard cannot combine in- dependently with any of these pairs to yield a triplet probability greater than 10-° per plant per year. However, a third event may combine dependently with one of the given pair com- binations. These combinations arise from a complete dependency assumption. Appendix E gives an example of evaluating the Conditional. probability of aircraft impect given the oc- fsrence of tind. 2 8 Design Approach for Com- bining Hazards This section discusses an acceptable approach which the designer may use to derive the com- binations of natural hazards and combins of natural hazards with man-made hazards which are appropriate for his specific site and nuclear power generating station design. The approach described in this section is the same as that used to develop the list of standard combinations given in Section 7, “Standard Hazard Combinations.” 8.1 Design Approach Sequence. A flow chart of the design approach sequence is given in Figure 2. 1 Identification of Hazards for Site. A study shall be made of the site and surrounding. area to identify the individual natural hazards and man-made hazards which can occur at the site and which have the potential for causing damage to safety-related structures, safety- related equipment located outside the plant structures (eg. piping and tanks), or the ultimate heat sink. As a minimum, the in. dividual hazards listed in Section 4, “Natural and External Man-Made Hazards,” shall be con- sidered for hazard combination purposes. Those which are not appropriate because of geographical location or siting layout shall be identified and eliminated from further con- jeration. 8.1.2 Review of Usually Excluded Hazards. ‘The individual hazards listed in Table 1 shall be reviewed to determine if any should be added to those identified in 8.1.1, above, because of unique site or plant features. .3 Final List of Hazards. A final list of in- dividual hazards shall be identified for hazard combination purposes. 8.14 Summary of Natural Hazards. For each natural hazard, the designer shall develop data similar to those in Table A-1, and shall in- clude (1) the hazard definition, (2) the probability of the hazard at the site per year, and (3) the estimated duration. Relative to the values given in Table A-1, the designer shall: (a) Use the hazard definitions given in the table.Study to Identity Hazards Natural Hazards v Develop Summary of Hazard Probability, Duration ¥ Compare with Table A-t or or Durat nis Different + — ‘Any one or More Probability ees Final List of Hazards or Use Methodology of Section 6 and Appendix C to Calculate ‘Combinations Develop a Design Basis ‘American National Standard ANSV/ANS-2.12.1978 Section 4 Hazards List Section 4.2 Check ‘Man-Made Hazards [Develop Summary —] of Effect, Frequency, Duration (Compare with Table A-2 Probabilities and Durations are Same Go to Section 7, List of ‘Standard Combinations + Delete Nonapplicable Combinations as a function of probability of occurrence Delete Nonapplicable Combinations based on Evaluation of Consequences SS] Figure 2 List Nonapplicable Combinations and Reasons Flow Chart for Design Approach Sequence 13American National Standard ANSWANS.2.12-1978 (b) Use the probabilities for his specific site or the probability values in the table. () Use the duration times for his specific site or the duration times in the table. Use of values less conservative than those given in Table A-1 may be appropriate for a particular site. However, the basis for such less con- servative values shall be documented. 81.5 Summary of External Man-Made Hazards. For each man-made hazard, a sum- mary shall be developed similar to Table A-2. ‘The summary shall include (1) the effect of the hazard, (2) the probability of occurrence of the hagard per site per year, and (3) the estimated duration, Relative to the values given in Table A-2 the designer shall: (a) Be specific as to the type of man-made hazard applicable to his site.” For example, in- stead of “surface vehicle explosion,” he shall list the specific surface vehicle, such as “barge explosion,” as appropriate. (b) Include the same effects as given in the table unless some of these are not possible due to the nature of the facility. For example, im- pact of a ship with a screenhouse may not be possible due to the presence of sandbars or protective structures in a lake. (©) Use the probability of the hazard for his specific site or the probability values in the table. (d) Use the duration times for his specific site or the duration times in the table. 816 Use of List of Standard Com- binations of Hazards. If the designer's sum- mary contains the hazards given in Tables A-1 and A-2 or a portion thereof, and if his probabilities and durations are the same as in the tables, then he shall use the List of Standard Combinations of Section 7, “Standard Hazard Combinations,” for the site, deleting from con- sideration those combinations which are not ap- plicable to the site. He shall list those which he has deleted, giving his reasons. The methodology ‘which was used to develop the list in Section 7 is described in Appendixes C and D. 8.17 List of Hazard Combinations Ap- plicable to the Site. If the designer's summary contains other hazards or probabilities or durations which are different from those given "Multiple souroes of common hazard should be considered if appropriate, eg. explnions from a barge, railroad or trick. 4 in Tables A-1 and A-2 he shall use the methodology of Appendixes C and D to derive a list of combinations which are applicable to his site. For hazard combinations involving depen- dent events, it is expected that the designer will probabilistically evaluate the strength of the dependency to determine whether the com- bination can be neglected for design purposes. 8.L8 Site Design Conditions. The designer shall use the hazard combinations developed as 4a result of 8.1.6 and 8.1.7 as design conditions for the nuclear power generating station. 8.L9 Example. Appendix E gives an example ‘of how to apply this design approach for a hypothetical site. 9. References [1] Title 10, Code of Federal Regulations, Part 100, “Reactor Site Criteria.” Government Printing Office, Washington, D. [2] American National Standard “Standards for Determining Design Basis Flooding at Power Reactor Sites,” N170-1976 (ANS- 2.8), American Nuclear Society, La Grange Park, IL. [3] SOLOMON, K.A., et ai, “Estimate of the Hazards to a Nuclear Reactor From the Ramdon Impact of Meteorites,” UCLA- ENG-7426, March 1974. [4] MELLOR, M,, “Avalanches,” Cold Regions Science and Engineering, Part III: Engineering, Section A3: Snow Technology, May, 1968, U.S. Army Materiel Command, Cold Regions Research and Engineering Laboratory, Hanover, NH [5] MELLOR, M,, “Dynamics of Snow Avalan- ches,” Cold Regions Research and Engineering Laboratory, Hanover, NH. (For publication in: “Geology and Mechanics of Rockslides and Avalanches.”) [6] General Electric Standard Safety Analysis Report (GESSAR), Volume 2, Section 3.8.4, ‘Tennessee Valley Authority, Knoxville, TN. [7] Final Safety Analysis Report, Tennessee Valley Authority, Knoxville, TN. [8] RAVINDRA, M.K., Sargent & Lundy Engineers, “Load Combinations for Natural ‘and Man-made Hazards in Nuclear Strue- tural Design,” ANS Topical Meeting on ‘Thermal Reactor Safety, Sun Valley, Idaho, August. 1977. ou| | | American National Standard ANSV/ANS-212.1978 Appendix A (This Appendix is not a part of American National Standard, Guidelines for Combining Natural and External Man-Made Hazards at Power Reactor Site, ANSI/ANS2.12-1978, but is inchided for information only) A. Bases for Characterization of Hazards This section presents the methodology for computing the probability of occurrence for each individual hazard. A summary is given in Tables A-1 and A-2. ‘The occurrence frequencies and durations estimated in this appendix may change significantly as more precise forecasting methodologies are developed and as hazard reduction measures are implemented. ‘The designer should be cognizant of, and use as appropriate, the current accepted practice for making these estimates as described, for example, in the latest versions of the relevant American National Standards and US Nuclear Regulatory Commission Regulatory Guides. A. Natural Hazards ‘The following sections describe the characteristics of each natural hazard and the basis for the annual probabilities estimated for these events. ‘The probabilities for the extreme events in each hazard category are not well established. The values chosen for these events are based upon data available in the literature and conservative judgment. In most cases, then, these values represent conservative upper bounds. While they proved not overly restrictive for use in identifying combinations of hazards for design pur- poses, it should be noted that these values may considerably overestimate the actual values for most extreme hazards. Where possible, this standard uses deterministic events (such as the standard project “flood). In this appendix a range of probabilities of occurrence of such events is estimated, and the largest probability in the range is used for determining standard hazard combinations. For purposes of probability calculation and engineering description, the spectrum of parameter values for most natural hazards has been divided into two or more intervals. These intervals define two or ‘more categories for each type of hazard. The consideration of a combination of two types of hazard (such as earthquake and rainstorm) then requires a consideration of all possible combinations of categories selected from each type of hazard. For example, there are two categories of rainstorm and two categories of earthquake so there are four possible combinations of earthquake and rainstorm. ‘The probability associated with the general combination of rainstorm and earthquake should be the sum of the probabilities of the four individual combinations. ‘The calculations used as illustrations in this standard (Appendix D) are based on selecting the most severe of the individual combinations which has a probability greater than 10-° per year. In cases where two or more of the individual combinations might be of equal severity (subject to furthe vestigation of actual consequences by the plant designer), they are considered separately. In other words, the general event combination of rainstorm and earthquake might consist of more than one specific individual combination requiring attention in plant design. It might be possible to have a general hazard combination for which none of the individual hazard ‘combinations has a probability greater than 10-® per year, but the sum of the probabilities of the in- dividual combinations (within the general combination) may be greater than 10°* per year. It might then appear necessary to consider the general combination in the plant design. This standard has chosen, rather, to consider only the individual combinations, for the following ressons: (1) The individual event probabilities and durations were intentionally chosen to be conservative; 6‘American National Standard ANSYANS.2.12.197 thus the calculated probabilities of the individual combinations within a general combination are already overestimates. (2) The number of individual combinations within a general combination is so small (never more than six, and usually only four), and the accuracy with which the individual probability estimates are made is generally so weak, that it is not appropriate to distinguish between the individual probabilities and their sum. (3) A general combination which would be the logical sum (or union) of several individual com- binations would not be defined by precise stress parameters which would be needed by the plant designer. Selecting the most severe of the component parameters (such as the worst earthquake and worst rainstorm) would be unnecessarily conservative. (4) Conservatism accrues from assigning 10°® per plant year as a cutoff because the probabilities been conservatively estimated, and the probability of a 10 CFR 100 release resulting from the iting event is at least an order of magnitude lower. These reasons are explained more fully in Ap- pendix B. Table A-1 Natural Hazards Probability mo OF Occurrence q) Estimated Natural Hazard Hazard Definition” (Per Site Per Yeary Duration® Earthquake (Ground Acceleration) SSE 103 to 108 60 seconds'®) OBB 107 to 109 30 seconds' Flood-River (Flow and Elevation) © DBF 107 2 months SPF 5 x, 105 to 103 "month ® 100 year 10? T'week') Flood-Coastal® Sie ei DBF 107 30 hours Cc SPF 5 x 10 to 103 30 hours 100-year 107 6 hours Flood-Lake DBF 107 12 hours SPF 5 3,10 to 108 12 hours 10-year 1 6 hours Rainstorm (Inches of Water) PMP 104 to 10 1 day 10-year 10 any Tornado (Wind Velocity) DBT a 107 2 minutes Tounami (Runup Elevation) PM Tsunami 104 to 10 2 days 10-year 107 to 105 1 day Wind” (Wind Vel 00-year 10 6 hours Legend: SSE — Safe Shutdown Earthquake OBE — Operating Basis Earthquake DBF — Design Basis Flood SPF — Standard Project Flood PMP — Probable Maximum Precipitation, local DBT — Design Basis Tornado PM Tsunami — Probable Maximum Tsunami 16 oe‘American National Standard ANSIVANS2.12-1978 Notes for Table A-I: (4) Ten-year events are not included forthe purpose of combination since they ae covered by normal design practice assump- (b) Actual rates of oowurrences and durations are site dependent; therefore thave values are estimates. The durations in this standard are teed for probability calculations only and should not be used in the actual design of structures, systems, ot com- Ponents. The hazards with a probability of 10°? are inchided for completeness. (c) Duration of lake flood governed ty duration of wind. {) Duration of coastal flood governed by duration of hurricane or storm. () These durations are typical of large streams. Sites on small streams should be analyzed separately as durations may be shorter. (f) Wind includes hurricane wind effects. For flosting plants the probebilty and estimated duration for wave effects is ‘equivalent to thowe for wind. The durations for the winds in Table A-1 were based on inland sites; for constal sites longer ‘turations should be considered. Windspeed is defined as the fastest mile windspeed with a recurrence interval of 100 years. (g) The values are considered to be the total duration of motion ofthe main shock "The duration af strong motion which is used ‘one-third these time durations A.LI Earthquake ‘The standard procedure defined by Cornell and Merz [1]* calculates the probabilities of exceeding dif- ferent levels of peak ground acceleration at the site. This procedure takes into account various types of earthquake generating sources in the area (faults and seismogenic regions) and their seismic history. In addition, the analysis considers the attenuation of ground motion from the epicenter to the site, the upperbound magnitudes of earthquakes for each source and the correlation of peak ground ac- celeration with the recorded Modified Mercalli intensity. ‘The standard industry practice as required by Title 10, Code of Federal Regulations, Part 100, Ap- pendix A, “Reactor Site Criteria,” is to design for two levels of earthquake: the sefe shutdown earth- quake (SSE) and the operating basis earthquake (OBE). In conformance with the industry practice, these two hazard levels — SSE and OBE — are used herein as shown in Table A-1. The SSE and OBE for a station are not currently selected for specified probabilities of exceeding the earthquake levels. Using the procedure recommended by Reference [1], the probabilities of exceeding the selected SSE and OBE levels for the station can be calculated. These probabilities can be used to derive com- binations of hazards relevant to the station. However, for the purposes of deriving standard hazard combinations, it is necessary to estimate the probabilities of SSE and OBE. Hsieh [2] et al have calculated the probabilities of exceeding different peak ground accelerations at western and eastern U.S. sites. According to these authors, the probability of exceeding a peak ground acceleration of 20% g at a site in the eastern United States (for average site conditions) is 6.5 x 104 per year whereas the probability of exceeding 40% g at a site in the western United States is 9 x 10 per year. A review of the SSE for currently licensed plants has indicated that these are the usual SSE levels for eastern and western United States, respectively. There is considerable variation in the probability estimate of the extreme events such as SSE, depending on the model used and subjective assessment of historical data, upperbound magnitudes of earthquakes, etc. Hence, the probability of ‘occurrence of SSE at any site in the U.S. can only be stated in a rather wide range of 10-* to 10-* per year. A similar review of the OBE levels for currently licensed plants have shown that the OBE level for nuclear plants ranges from 10-* to 10-* per year. : Duration of seismic shaking is a function of wave frequency, amplitude threshold and Richter magnitude. Using world-wide strong motion records, Bolt [3] has derived a table of durations of strong-motion (peak ground acceleration exceeding 0.05g) earthquakes as functions of Richter ‘magnitude and distance from the earthquake source. The longest duration predicted is 35 seconds for an earthquake of magnitude 8.5 at 10 km from the site. For a specific site, the durations of OBE and SSE may be calculated using the table prepared by Bolt when the magnitude and hypocentral distan- ces of these earthquakes are determined. However, for.the purposes of deriving standard hazard com- “References for Appendix A are listed at the end of Appendix A wAmerican National Standard ANSV/ANS-212-1978 binations, it is necessary to estimate the durations of earthquakes. Since the standard hazard com- binations should be applicable for any site in the United States, these duration estimates are un- derstandably conservative upper bounds, ‘Since the OBE acceleration is in the range of 1/3 to 1/2 SSE acceleration, the magnitude of the earthquake to cause this OBE acceleration at 10 km from the fault is between 5.8 and 6.6. Using the duration table of Bolt, the OBE duration for M = 6.6 is obtained as 20 seconds. However, in this stan- dard, the durations of OBE and SSE are assumed to be 30 seconds and 60 seconds, respectively. A.L2 Flood ‘The treatment of flooding in this standard is consistent with ANSI/ANS-2.8-1977 (N170) [4]. As in Ref. [4] flooding is divided into three categories by the type of sites: coastal, lake, and river. Three levels of flooding for each type of site are postulated to be considered for combining with other hazards: an extreme event, a severe event, and a 100-year event, ‘The extreme event forall three types of sites is the design basis flood (DBF) described in Ref. [4]. The procedure to calculate the magnitude and duration of the design basis flood is described in Ref. [4]. ‘The probability of occurrence of the design basis flood is estimated in Ref. [4] to be 10°? per year or less. . - - The severe event flood for a coastal site subject to hurricanes, is the standard project hurricane described by Graham & Nunn. [5] For a coastal site not subject to hurricane and for a site on a lake shore, the severe event flood is not described in the literature and must be addressed on a case-by-case basis. ‘The severe event for a river site is the standard project flood or one-half the flow of the probable maximum flood described in Ref. [4]. The procedure to calculate the magnitude and duration of the standard project flood (SPF) for river sites is described by the U.S. Army Corps of Engineers. [6,7,8] Methods for evaluating this magnitude are described in Ref. [4] for rivers. The probability of oc- currence of an SPF for a river site is estimated by Beard [9] to be 5 x 10-2 to 10° per year. The same probability is assumed for coastal and lake sites in Table A-1. A lesser flood is also considered as a possible high-probability, low-intensity event to combine with other hazards. The 100-year flood is selected for this purpose for all three types of sites. ‘The procedure to calculate the magnitude and duration of a 100-year flood is evaluated from data for ‘each site using the procedure outlined by the U.S. Army Corps of Engineers [10] and from regionalized procedures published by the U.S. Geological Survey. [11] For the purposes of deriving standard hazard combinations, the durations of floods are estimated as shown in Table Ac Evaluation of a flood on a coastal site should include the possibility of coastal erosion during a hurricane, A.L.3 Rainstorm Data to define the rainfall in a site region or for estimating rainfall at a specific site are usually ob- ‘ined from the U.S. National Weather Service, National Oceanic and Atmospheric Administration (NOAA). These references are listed in Ref. [4] (Section 5.3.3.2 “Coincident Storm” for 100-year rain- storm data and Section 6.2 for Probable Maximum Precipitation). ‘The duration of one day for high intensity local rainfall has been used to calculate hazard com- binations. 18‘American National Standard ANSV/ANS-2.12.1978 ‘The probable maximum local rainstorm is arbitrarily assigned a probability range of 10 to 10 per year. This probability is selected as being two to three orders of magnitude more frequent than the Design Basis Flood (DBF) as discussed in A.1.2. In addition, a 100-year rainstorm is listed in Table A-1. The intensity of such a rainstorm can be ob- tained from U.S, National Weather Service publications, referenced above. A.L4 Tornado Ref. [12] recommends risk model for calculating the probabilities of exceeding various tornado wind speeds. A tornado is usually characterized by its maximum horizontal resultant wind speed; other parameters used to describe a tornado are maximum and minimum trafslational wind speeds, velocity profile, pressure drop and rate of pressure drop. For a specific site the probabilities of exceeding various tornadic wind speeds are given in the regional tornado map included in a proposed American National Standard. (See Ref. [12]}. This information can be used to generate site specific hazard com- binations which include tornadoes. For the purposes of deriving standard combinations, only one tornadic wind speed is considered. This one is commonly known as the design basis tornado (DBT), with a threshold probability of 10-7 per year. [13] Intermediate events are not considered since tornadic wind speeds associated with a probability bet- ween 10 to 10 per site per year are about the same as wind speeds based on a normal wind loading with a 100-year recurrence interval. This (100-year wind) loading is considered in combination with other loading as a service load and the allowable stresses (or load factor) are appropriately low (or high) to reflect this fact. The normal wind loading is more critical than an intermediate tornado, and the intermediate tornado need not be considered. ALS Tsunami ‘The methodology for evaluating tsunamis from both distant and local events is described in Ref. [43]. ‘This reference does not describe how the probability of a probable maximum tsunami (PMT) could be calculated. Sufficient historical tsunami data on actual tsunami occurrences are not available to draw statistically sound inferences on the tsunami risk at a site. For reference purposes, the probability of the probable maximum tsunami is assumed to be 10 to 10 per year which lies within the range of probability of safe shutdown earthquakes, For a more frequent event, the 100-year tsunami was selected. ‘The duration of a tsunami depends on the magnitude and distance of the tsunamigenic source and the topography and size of the basin. A distantly generated tsunami generally has a longer duration than a locally generated tsunami of equal magnitude. For a specific site, the durations of the PMT and the 100-year tsunami could be calculated by wave propagation methods. For the purposes of this standard, the duration was estimated by studying the tide gauge records. [14] The tide gauge records that the complete dissipation of a tsunami could require more than 18 hours. Hence the duration was estimated as one day for the 100-year tsunami and 2 days for the PMT. A.L6 Wind Ref. [12] gives a method for determining the design wind velocities for specified mean recurrenc tervals. The procedure involves making a detailed site study incorporating historic wind statistics, site meteorological and topographical features; from this, the extreme windspeed is selected for the site. ‘The winds are taken from existing National Weather Service records supplemented by any on-site data 19American National Standard ANSW/ANS2.12-1978 that are available. This method recognizes that local meteorological and topographical conditions significantly influence the maximum windspeeds at the site. Typical applications of this approach have been demonstrated by ‘Thom. [15] ‘The design wind speed for seismic Category 1 structures of nuclear power plants is usually selected as, a wind speed which has a mean recurrence interval of 100 years. Since the design basis tornado represents an extreme wind loading, no other extreme wind need be selected for the purpose of deriving hazard combinations ‘The duration for the 100-year wind is estimated to be 6 hours, A2 External Man-Made Hazards While natural phenomena tend to vary with frequency of occurrence and severity, man-made hazards are‘usually characterized as having a discrete frequency of occurrence. Thus, estimates have been made for the probability of the man-made events based on available data and probability models. A discussion is provided below for each man-made hazard to show how each of the probability estimates for the man-made hazards was determined, and to provide guidance and references which will assist the designer in evaluating these hazards for a particular site. The designer should carefully analyze his own characteristics before using these values as realistic probabilities for the individual event at a par- ticular site. ‘The information in this Section is summarized in Table A-2. A21 Aircraft Impact ‘The siting of nuclear power plants near commercial or private airports is not a common practice. In those instances where airports are within a five-mile radius of nuclear power plants, probabil evaluation of the potential aircraft impact hazard (missile impact) and fire/smoke hazard is required to be performed by the designer. ‘The aircraft impact probability analysis is based on the following general formula: P=DxAxM where: P = Annual probability an aircraft will impact a plant structure which contains safety related equipment or components, or both. D = A parameter for aircraft large enough to damage structures which contain safety-related equipment or components, or both, based on historical data of crashes involving one or more fatalities (expressed in crash probability per movement per unit area). A = Effective area of included plant structures. 'M = Number of annual movements from the airport. A movement is defined as a landing or a takeoff.* References [16, 17, 18, 19, 20, and 21] give examples of the above calculations. In establishing the parameter “D” inthe gqustion above, historical data is available from the National ‘Transportation Safety Board (NTSB). [22] This data is categorized into small aircraft (weight < 12,500 pounds and large aircraft (weight > 12,500 pounds). The parameter “D” is sensitive to the distance, r, from the plant to the airport. “D” varies approximately as 1/r?. For the impact probability estimates given in the table below for airports within five miles of the plant, Wall [23] has used a value for F of about 2.5 miles. “Ifa principal ar coridor exists near site, a separate probability analysis is performed where the parameter M defines over- Aights as a movement.‘American National Standard ANSI/ANS2.12-1978 Table A-2 © sternal Man-Made Hazards Probability of Occurrence Estimated Effect (per site per year) Duration 1 Airports within 5 miles of plant: Impact (large aircraft) Lt x 10% Instantaneous! Impact (small aircraft) 33 x 105 Instantaneous) Fire (large aircraft) 05 8 hr Fire (small aircraft) 05 30 min Airports more than 5 miles from plant:\® Impact (large aircraft) 4.6 x 107 Instantaneous! Impact (small aircraft) 14 x 105 Instantaneous Fire (large aircraft) 0.510 8 hr Fire (small aircraft) 05 90 min 2, Military or Industrial Facility Accident Blast, Missiles 32 x 105 Instantaneous) 3. Pipeline Accident « Blast, Missiles 81x 108 Tnstantaneous) 4. Retaining Structure Failure Flood or Sudden Drawdown 104 2 hours to 7 days 5. Surface Vehicle Explosion Explosion, Missiles 1x 103 Instantaneous) 6, Surface Vehicle Impact Impact (ship) 4x10 Instantaneous Notes for Table A-2 (a) Actual rates of occurrences and durations are site dependent (b) For purposes of caleulating probabilities of these hazards with other natural or man-made hazards, the hazard duration is sssumed to be 8 seconds. 1 (¢) Conditional probability given that aircraft impact has occurred. (4) For sites greater than 5 miles from sn algpoct, estimates of prabability of impact are required if the site is near « heavily traveled commercial air traffic lane [17] of if there are unusual aireaft operations in the vicinity. [18,19] 2‘American National Standard ANSY/ANS-2.12.1978 ‘The table below gives aircraft impact probabilities from Ref. [23]. This reference assumes a value of A ‘equal to 0.137 square ‘The product of the parameters “D” and “M” in Ref. [34] is based on an analysis of civil-aircraft accident data over several years through 1968, Other probability analyses for specific sites have given similar values, WASH 1400 [24] gives values of 10° to 107 per year. Aireraft Impact Approximate Frequency Maximum Effect (per plant year) Expected Duration Airports within 5 miles of plants Impact (large aircraft) Li x 108 Instantaneous (* Impact (small aircraft) 3.3 x 105 Instantaneous ‘*) Fire (large aircraft) 0.5.) 8 hr. Fire (small aircraft) 0.5) 30 min. Airports more than 5 miles from plant: Impact (large aircraft) 45 x 107 Instantaneous ‘*) Impact (small aircraft) Lt x 105 Instantaneous *) Fire (large aircraft) 0.5 8 hr. Fire (small ai 05 30 min. craft impact has ocurred ‘rom an airport, estimates of prabsbility of impact are required ifthe site is near s heavily ne [25] or if there are unusual aireraft operations in the vicinity. [28, 27] If an aircraft does impact a plant, the probability of a postaccident fire and resultant smoke hazard would be between 0.3 and 0.5. [23] Wall [23] has stated that the range of fuel capacities for small and large aircraft are 100-400 gallons and 3,000 to 50,000 gallons, respectively. The duration of a fire in- volving 200 gallons of aircraft fuel from a small aircraft has been conservatively estimated at 30 minutes. [21] The duration of a fire is largely dependent on the geometry of the fuel spill. For large aircraft, a similar estimate gives a fire duration of eight hours for 32,000 gallons of fuel. A.22 Industrial or Military Facility Accident A fixed industrial or military facility presents several types of potential hazards to a nearby nuclear power plant including fire (and resulting smoke and gases from combustion), explosion (with at- tendant pressure wave, ground shock, and missiles), other missiles, and release of toxic or flammable gases. The principal hazards posed to a nuclear power plant from an off-site fire are the effects on control room habitability and diesel generator combustion air intakes from smoke and combustion gases. The “thermal effects of an offsite fire on plant structures would not generally be significant based on current design practice. The effects of fire on control room habitability would be similar to the effects of a postulated toxic gas release by a nearby industrial facility. Since these effects are considered in- dividually in plant design, and since they are not additive or synergistic with effects of other natural or ‘man-made phenomena, it is not necessary to consider smoke, combustion gas, and toxic gas in com- bination with other hazards. The blast and missile effects of an external explosion could contribute loads to plant structures which are additive with loading due to other phenomena. A review of the available literature [28] shows that explosions of the magnitude required to cause significant blast ef- fects at a nearby nuclear power plant are very infrequent. 22‘American National Standard ANSI/ANS:2.12.1978 Design for external explosion effects become necessary only when the peak positive incident over- pressure exceeds 1.0 psi. Seismic Category I structures can withstand this overpressure. Guidance for calculations of explosion pressure effects is given in Ref. [29]. A safe distance R in feet from the source of explosion can be defined by the relationship R > 45W 1/3 where W is the exploding charge in pounds of TNT. ‘The above relationship suggests that rather large amounts of explosives are required to produce significant blast effects at any distance. Records [28] indicate that there have been about 31 large explosions in the United States since 1910 resulting in fatalities. Of these, 4 were clearly not explosions occurring as a result of a mishap at a fixed facility. Taking 27 large explosions over 64 years (1910-1973), yields an average frequency of 0.42 large explosions per year. ‘The U.S. Bureau of the Census records [30] show that there are about 305,000 industrial facilities in the United States of which 13,127 were associated with chemical and petroleum products. Assuming then that there are about 19,000 fixed facilities in the United States capable of producing significant explosions, the average frequency of significant explosions per fixed facility per year is about 3.2 x 10°. Current practice in nuclear power plant siting avoids sites very near hazards such as a munitions, hazardous chemicals, or petroleum product facilities. However, assuming one facility having potential for a significant accident explosion is near a typical plant site, a conservative estimate of the average annual probability of significant blast effects occurring at a nuclear plant site is about 3.2 x 10%. Studies for a particular license application [31] have shown that the over-pressure effect from an ac- cidental explosion off-site are more limiting than ground shock or explosion-generated missile effects. ‘The duration of explosion overpressure effects has been shown [31] to be less than 5 seconds. ‘The probability of an impact on a nuclear power plant site by a misdirected military missile is con sidered insignificant. It should be noted that 3,2 x 10-5 per plant year is a very conservative estimate of the frequency of significant blast effects at a power reactor site. Most power reactors are located so far from potential sites of large explosions that this probability is considered to be < 10-" per year. In addition, of the 27 “significiant explosions” from fixed facilities in the U.S. between 1910 and 1973, it is doubtful that more than a few were of the magnitude required to produce significant blast effects even if they oc- curred near the site boundary. However, it appears unnecessary to further refine these conservative ‘estimates at this time, since the currently estimated probability coupled with the short duration, is ex- pected to show that these effects should not be considered in combination as an independent event with any other natural or man-made hazards, A.23 Pipeline Accident A pipeline, transporting materials which are in a liquid state under normal temperature and at- mospheric pressure, will not cause any substantial hazard condition even if leakage occurs. However, accidents to pipelines transporting gases under pressure can lead to leaks of natural gas, propane, and other flammable explosive or toxic gases which may have potential consequences to a nuclear plant. ‘The applicable hazards are overpressure due to air blast, thermal load resulting from gas deflagration, missile hazard, and gas concentration within the plant. Most of the parameters involved in deter- ining the effect of a pipeline accident are site-related variables and evaluation must be made on a ‘case-by-case basis. The evaluation involves quantity-distance relationships, site topography, and site meteorology including wind direction, wind speed, and stability class. 23‘American National Standard ANSY/ANS-2.12.1978 ‘The propane-air explosion which occurred at Port Hudson, Mo., on December 9, 1970, as a result of a pipeline break represents a case history of the hazards of fuel transportation. [32] This accident resulted in overpressure of 1.0 psi and 0.5 psi at a radias of about 0.4 miles and 0.7 miles, respectively. ‘Nuclear plants are designed to safety withstand the “Design Basis Tornado.” The detonation which produces a peak overpressure less than the wind pressure from a tornado should not represent hazard. The dynamic wind pressure associated with the design basis tornado i8 2.13, 1.34, and 0.95 psi at a height of 150 feet for maximum wind speeds of 300, 250, and 200 mph, respectively. [12] WASH 1400 [24] gives a gas pipeline failure rate of about 40 leaks/300,000 miles/year. This results in a probability of a pipe rupture being 1.3 x 10°‘ per mile per year. The analysis for the Hartsville Nuclear Plant [33] showed the probability of a pipe rupture equal to or greater than one foot in length was 12 x 104 per mile per year. ‘Tire following is a summary of the pipeline leak parameter estimates compiled from the statistical analysis conducted in that effort: ——_Pipatine Deak Festor_ — Leak rate 0.12 x 10 leaks/mile/year Ignition rate 0.24 ignitions/leak Detonation probability (mean) 0.28 detonation/ignition Ignition time (mean) 6.6 minutes Blast duration 5 tec Fire duration 6 hr Gas contamination duration 6 he As stated before, the effects of pipeline accident are highly site dependent. For the purpose of estimating the upper bound probability of the pipeline accident the following assumptions were made: (1) Both ignition and detonation must occur. (2) Meteorology (wind direction, iwind speed and stability),was such that a gas cloud with the proper mixture to detonate was formed. (Very conservative assumption since the necessary meteorological conditions for such cloud formation are rare at such sites). (3) About one mile of pipeline was of interest for the typical site, ‘The resulting probability of detonation was 8.1 x 10-*iyear. A.2A Retaining Structure Failure (See Section 2, “Definitions”.) Retaining structure failures can result in either abnormally high or low water levels, either of which could be a hazard to the plant. When dam design, construction, maintenance, and inspection are performed according to acceptable modern standards and practices, the probability of failure should be very small in the absence of over- topping or extreme earthquake. A dam failure is defired as any malfunction, displacement, or collapse which results in an uncontrolled release of a significant portion of the reservoir. Historical statistics [34,44] show an average frequency of dam failure of one failure per 10,000 dam-years worldwide (for dams greater than 45 feet in height). Such a large probability, reflects many instances of poor design, construction, maintenance, or supervening environmental hazard. It should be recognized that this is @ conservative assumption because these statistics include dam failures due to floods which are not in- cluded as a hazard in our definition of retaining structure failure. Flooding effects of dam failures are covered in Section A.1.2. For purposes of this standard, a frequency of one retaining structure failure in 10,000 years is used to identify those hazard combinations which are sufficiently large to require »American National Standard ANS/ANS-2.12-1978 consideration by the plant designer. It is up to the plant designer to further investigate the failure probabilities for dams and other water retaining structures in his area to determine whether lower hazard probabilities are justifiable. In performing these studies it is recommended that the designer investigate three categories of cessive failure risk: earthfill dams, dams with high seismic risk, and dams located on poor foundation ‘material. If it ean be demonstrated that the structure is in none of these categories, and that it is sub- jected to regular inspection and maintenance according to accepted professional practice, then a probability significantly smaller than the conservative 10+ per dam-year may be justifiable. A285 Surface Vehicle Explosion ‘A surface vehicle accident can produce several types of potential hazards to a nearby nuclear power plant including fire (and resulting smoke and combustion gases), explosion (with attendant pressure wave, ground shock, and missiles), other missiles, and release of toxic or flammable gases. ‘The principal hazards posed to a nuclear power plant from an off-site fire are the effects on control room habitability and diesel generator combustion air intakes from smoke and combustion gases. The thermal effects of an off-site fire on plant structures would not generally be significant based on current design practice and separation between the plant and transportation route. The effects of fire on control room habitability would be similar to the effects of a postulated toxic gas release by a nearby industrial facility. Since these effects together with separation distance are considered in- dividually in plant design, and since they are not additive or synergistic with effects of other natural or man-made phenomena, smoke, combustion gas, and toxic gas were not considered in combination with other hazards. The blast and missile effects of an external explosion could contribute significant loads to plant structures which are additive with loading due to other phenomena even when considering separation distance. Hence these are the effects considered further in this standard. Surface vehicle explosions (truck, train, barge in river or canal, or ocean ship) present a potential hazard to a nuclear power plant from the standpoint of both blast overpressure and explosion generated missiles ‘To determine whether a surface vehicle explosion is likely enough to be considered to occur simultaneously with any other man-made or natural hazard, the probability of occurrence and the duration of the effects must be estimated. ‘The probability of a surface vehicle explosion which can be hazardous to the plant can be calculated by Po = PaxpexLxf where: By = probabil ord b, = probability of surface vehicle accidert per mile traveled (mi!) conditional probability of surface vehicle explosion, given that accident has taken place. length of path within which explosion of shipment can be a hazard to the plant site (mi.) frequency of shipment (yr-') Several estimates are available for pa for highway, rail, and shipping. The accident rate for trucks on highways has been estimated by the U.S. Atomic Energy Commission [35] to be 2 x 10 per mile. For all railroad accidents the accident rate is about 9 x 10“ per mile. [36] In some applications the qua tity of trackage near the site has been considered to show that a lower accident rate per mile is a propriate. [37,38] However, for the estimates presented here, an accident rate of 9 x 10° per mile will bbe used. For barge shipping in rivers and canals, the accident rate has been estimated to be between ity per year of surface vehicle explosion which may be a hazard to the plant site[American National Standard ANSI/ANS-212-1978 1.8 x 10° per mile and 5.65 x 10° per mile. (35,37,38] For ocean shipping the collision rate has been estimated to be 2 x 10-7 per mile. [40] The conditional probability of explosion, given a surface vehicle accident is taken to be 0.1 for trucks and 0.01 for trains [38], and 0.05 for ships. [37,40] Conservative values for L, the path length of interest, assuming the closest approach distance to be zero, have been calculated [41] assuming that the maximum possible shipments for truck, railroad, and barge or ship are 43,000 Ib., 396,000 Ib. and 5,000 ton of TNT, respectively, and assuming that ex- plosions producing 1 psi or greater peak reflected over pressure are of interest. These path length values are shown in the table below. ‘The frequence of shipments of significant amounts of explosives by surface vehicles varies considerably depending on the transportation route being considered. A value of 400 shipments per year is felt to be 1 conservative upper limit [38,40], and has been used below in calculating the probability of surface vehicle explosions, ‘The following table shows the resulting probability values for surface vehicle explosion which may be ‘a hazard to the plant. For the purpose of combination with other hazards, the highest of the four values (truck explosion), 1 x 109 yr? has been chosen. It should be noted that these are conservative bounding estimates of probability of occurrence of truck, train, and ship explosions that may be a hazard to a plant. Additional refinement can be incorporated if required for a particular site by evaluating P(ffo) and P(c/f), equations 3-1 and 3-2. Values for these parameters will probably vary with distance from the plant over the range of influence, L. Due to the short duration of these events, no further effort to refine these values is required for the purpose of computing the probability of their simultaneous occurrence in combination with other hazards Parameters for Calculation of Surface Vehicle Explosion Probability Parameter Truck Train River Barge Ocean Ship Assumptions: Pa (mi-') 25 x 10% 9x 10% 5x 10% 2x 107 Pe or 0.01 0.05 0.05 L (mi) 1 23 4 4 fir!) 400 400 400 400 ‘Then: Po (rt) 1x10 81 x 105 4x104 16 x 105 A.26 Surface Vehicle Impact Plant areas of concern with regard to surface vehicle impact are safety related structures. ‘An estimate of the probability of a ship impact with the intake structure can be made by using data developed for the Atlantic Generating Station. [42] These studies show that the frequency of strand- ings and dock impacts is about 1 x 107 per ship mile per year. Assuming that the ship traffic past a site is about 400 ships per year, and that the track length during which a ship could go out of control and potentially strike the intake structure is about one mile, then a conservative estimate of the probability of a ship striking the intake structure is about 4 x 10 per site per year. 26 »)‘American National Standard ANSI/ANS-2.12.1978 ‘This estimate is conservative since the ship traffic is not as high as assumed for most sites, and the in- take structure is generally less of target than a dock, being usually on the shoreline, submerged, oF both. Since this probability is already quite low and the duration of the impact load is short, further refinement of this probability is not warranted for the purposes of calculating the probability of hazard combinations. References for Appendix A a P] (3] a) {5] (6] (7 [8] [9] 10] ty 12] 13] CORNELL, C.A. and MERZ, H.A., “Seismic Risk Analysis of Boston,” Journal of the Strue- tural Division, ASCE, Vol. 101, No. ST10, October, 1975, pp. 2027-2043. HSIEH, T, OKRENT, D., and APOSTOLAKIS, G.E, “On the Average Probability Distribution of Peak Ground Acceleration in the U.S. Continent Due to Strong Earthquakes,” UCLA-ENG.-7516, University of California at Los Angeles, March, 1976. BOLT, B.A., “Duration of Strong Ground Motion,” Proceedings of Fifth World Conference on Earthquake Engineering, Vol. 1, pp. 1304-1313, Room, 1973. ANSI/ANS.2.8-1977(N170) “Standards for Determining Design Basis Flooding at Power Reac- tor Sites.” American’ Nuclear Society, La Grange Park, IL. GRAHAM, H.E. and D.E. NUNN, “Meteorological Considerations Pertinent to Standard Project Hurricanes, Atlantie and Gulf Coasts of the United States,” National Hurricane Research Project Report No. 33, U.S. Weather Bureau, Washington, D.C, November 1959. USS. Army Corps of Engineers, “Standard Project Rain-Flood Criteria, Sacramento-San Joaquin Valley, California,” U.S. Engineer District, Sacramento, California, April 1957, Revised 1958, U.S. Army Corps of Engineers. “General Standard Project Rain-Flood Criteria, Southern California Coastal Streams,” The Hydrologic Engineering Center, Sacramento, California, March 1967. UGS. Army Corps of Engineers, “Standard Project Flood Determinations," Em 11102-1411, Revised March 1965, Washington, D.C. BEARD, LEO, “Probabilities of Rare Floods,” Proceedings of the Engineering Foundation Conference, Asilomar Conference Grounds, Pacific Grove, California, September 25-28, 1973. BEARD, LEO, “Statistical Methods in Hydrology,” U. Army Corps of Engineers, Sacramento, California, January 1962. Bulletin No. 17, U.S. Water Resources Council, Hydrology Committee, “Guidelines for Deter- mining Flood Flow Frequency,” March 1976. Proposed American National Standard “Standards for Estimating Tornado, Hurricane, and other Extreme Wind Parameters at Nuclear Reactor Sites,” Draft 2, Revision 0, July 1977, ANS-2.3. Assigned Correspondent: George W. Nicholas; Dames & Moore. MARKEE, EARL H., JR, BECKERLEY, JAMES G., and SANDERS, KE., “Technical Basis for Interim Regional Tornado Criteria,” WASH 1300, USAEC Office of Regulation, May 1974,‘American National Standard ANSANS-2.12-1978 ua) 118) a7 8) 9] (20) 1 (22) (23) (24) (25) (26) (27) (28) (9) WEIEGEL, R.L., “Earthquake Engineering,” Prentice Hall, Inc., Englewood Cliffs, N.J., 1970. THOM, H.CS., New Distribution of Extreme Winds in the United States, Journal of Struc- tural Division, ASCE, Vol. 94, ST7; pp. 1787-1801, July 1968. EISENHUT, D.G., “Reactor Sitings in the Vicinity of Airfields,” Transactions Amer. Nuel. Soc., 16:210-211, 1973. Shoreham Nuclear Power Station, Docket No. 60-322, Amendment No. 3, February 5, 1969. Pickard, Lowe and Associates, Inc., A Study of the Probability of an Aireraft Using Waukegan Memorial Airport Hitting the Zion Station, for Commonwealth Edison Company/Sargent and Lundy Engineers, Washington, D.C., May 5, 1972. Pickard, Lowe and Associates, Inc., A Study of the Probability of An Aircraft Using the Grum- man Peconic River Airport Colliding into the Shoreham Nuclear Power Station, Unit No. !, for Long Island Lighting Company, Washington, D.C., October 1973. Potomac Electric Power Company, Preliminary Safety Analysis Report, Douglas Point ‘Nuclear Generating Station Units 1 and 2, Vol. 10, Appendix D, “A Study of the Probability of fan Aircraft Using Marine Corps Air Station — Quantico Colliding into the Douglas Point Nuclear Generating Station,” August 1973. Northeast Utilities Co PSAR — The Montague Nuclear Power Station, Appendix 2G, “Analysis of the Probability of An Aircraft Using Turners Falls Municipal Airport Impacting the Montague Nuclear Power Station,” 1974. Preliminary Statistical Analysis of Aircraft Accident Data, U.S. Civil Aviation, 1974, National ‘Transportation Safety Board, Report No. NTSB-APA-75- WALL, LB, “Probabilistic Assessment of Aircraft Risk for Nuclear Power Plants,” Nuclear Safety, Vol. 15, No. 3, May-June, 1974. WASH 1400 (NUREG — 75/014), “Reactor Safety Stu US. Commercial Nuclear Power Plants,” U.S. Nuclear Regulatory Commission, October 1975 “Public Service Electric and Gas Company, Atlantic Generating Station Units 1 and 2, Preliminary Safety Analysis Report,” Appendix 2F, Docket No. STB-50-477 and STB-50-478, April 1974, HORNYIK, K., ROBINSON, A.H., and GRUND, J.E., “Evaluation of Aircraft Hazards at Boardman Nuclear Plant Site,” PGE-2001, Portland General Electric Company, May 1973. Pebble Springs Nuclear Plant, Docket No. 50-514 and 50-515, Portland General Electric Com- pany, May 1974. World Almanac and Book of Facts — 1974, Newspaper Enterprise Associ 1974. ion, New York, Department of the Army Technical Manual TM5-1300, “Structures to Resist the Effects of Accidental Explosion,” U.S. Government Printing Office, Washington D.C. ®‘American National Standard ANSU/ANS.2.12-1978 [30] U.S. Bureau of the Census, “Statistical Abstract of the United States — 1974,” 94th Edition, Washington, D.C, 1974 [31] Brunswick Steam Electric Plant, Units 1 and 2, PSAR, Amendment 1, Supplement 1. [32] Report of Investigation 7752, “Detonation of a Flammable Cloud Following a Propane Pipeline Break,” Department of Interior, Bureau of Mines. [33] Mechanics Research Incorporated, Report MRI-2687-75, “Nuclear Power Plant Risk for a Natural Gas Pipeline,” Published August 2, 1974. See also Hartsville Nuclear Plant PSAR, ‘Tennessee Valley Authority, Docket Nos. 50-518, 519, 520, 521. [84] CARLIER, MICHEL, “French Laws and Regulations Regarding Dam Supervision and In- spection.” From the Proceedings of the Engineering Foundation Conference on Inspection, Maintenance, and Rehabilitation of Old Dams, ASCE, 1974. [35] “Environmental Survey of Transportation of Radioactive Materials to and from Nuclear Power Plants,” USAEC, Division of Regulatory Standards, December 1972. [36] Accident Bulletin No. 141, “Summary and Analysis of Accidents on Railroads in the US, 1972,” DOT, Federal Railroad Administration, Office of Safety. [37] Trojan Nuclear Plant FSAR, Amendment 7, Portland General Electric Co., Docket No. 50-344, Portland, Oregon, January 1974. [38] Byron-Braidwood Nuclear Plant PSAR, Amendment 8, Sections 2.2 and Appendix A to Chap- ter 2, Commonwealth Edison Company, Docket No. 50-545-4-5-6-7, October 1974. [39] Brunswick Steam Electric Plants, Units 1 and 2 PSAR, Amendment 1, Supplement 1, Carolina Power and Light Co., Docket No. 50-324-4, Raleigh, NC, September 1968. [40] WALKER, D.H., HARTMAN, M.G,, and ROBBINS, T:R,, “Methods for Estimating Risks to Nuclear Power Plants from Shipping,” Institute of Environmental Sciences Proceedings, Volume I, Page 6, May 14, 1975. [41] Regulatory Guide 1.91, “Evaluation of Explosions Postulated to Occur on Transportation Routes Near Power Plant Sites,” U.S. Atomic Energy Commission, January 1976. [42] Atlantic Generating Station PSAR, Amendment 8, Public Service Electric and Gas Company of New Jersey, August 16, 1974 [43] Prosposed American National Standard “Guidelines for Determining Tsunami Criteria for Power Reactor Sites,” ANS-2.4. Assigned Correspondent: Norman R. Wallace, Bechtel Power Corporation. [44] GRUNER, EDWARD, “The Mechanism of Dam Failure,” Transactions of the Ninth In- ternational Congress on Large Dams, Paris (1967). 29‘American National Standard ANSI/ANS-2.12-1978 Other References Reviewed in the Development of Appendix A Which May be of Interest to a Designer Department of the Army Technical Manual TM5-856-1, “Fundamentals of Protective Design (Non- nuclear).” U.S. Government Printing Office, Washington, D.C. U.S. Atomic Energy Commission, Division of Operational Safety, Safety and Fire Protection Bulletin 10, "The Study of Missiles Resulting from Accidental Explosions — A Manual for Investigators,” U.S. Government Printing Office, Washington, D.C. Regulatory Guide 4.7, “General Site Suitability Criteria for Nuclear Power Stations,” U.S. Atomic Energy Commission, September 1974. GRUND, J.B. and HORNYIK, K,, “Hazards from Aircraft,” Institute of Environmental Sciences 1975 Proceedings, Volume 1, Page 1, April 14, 1975. “Offshore Power Systems, Plant Design Report,” Appendix 2A, Docket No. STN-50-437, May 1973. ‘MC KENZIE, D.W., “Risks to Nuclear Power Plants Due to Natural Gas Pipelines,” Institute of En- vironmental Sciences 1975 Proceedings, Volume I, Page 17, May 14, 1975. National Weather Service, “Precipitation Frequency Atlas of Western United States,” NOAA Atlas 2, Vol. I, Montana; Vol. I1, Wyoming; Vol. III, Colorado; Vol. IV, New Mexico; Vol. V, Idaho; Vol. VI, Utah; Vol. VIL, Nevada; Vol. VIII, Arizona; Vol. IX, Washington; Vol. X, Oregon; Vol. XI, California; Washington, D.C. 1974, WHITE, G.F. and HAAS, J.E., “Assessment of Research on Natural Hazards,” The MIT Press, ‘Massachusetts Institute of Technology, Cambridge, Massachusetts 02142, No. NSF-RA-E-75-001 ORNL-NSIC-117, “Protection of Nuclear Power Plants Against External Disasters,” edited by Wm. B. Cottrell, Oak Ridge National Laboratory, April 1975, Available from National Technical Information Service, U.S. Department of Commerce, Springfield, VA 22151. ‘The Following Monograph Series Available from Program of Technology, Environment and Man Institute of Bekavioral Science University of Colorado Boulder, Colorado #002. FRIEDMAN, DON G., Computer Simulation in Natural Hazard Assessment. 1975. #003 COCHRANE, HAROLD C., Natural Hazards and Their Distributive Effects. 1975, #004 WARRICK, RICHARD A., et al. Drought Hazard in the United States: A Research Assessment. 1975. #005 AYRE, ROBERT S,, et. al. Earthquake and Tsunami Hazards in the United States: A Research Assessment. 1975. #006 WHITE, GILBERT F,, et al. Flood Hazard in the United States: A Research Assessment. 1975. 30 © »#007 #008 #009 #011 #012 #013 #021 ‘American National Standard ANSUANS-2.12-1978 BRINKMANN, WALTRAUD A.R, ef al. Hurricane Hazard in the United States: A Research Assessment. 1975. BAKER, EARL J. and JOE GORDON-FELDMEN MC PHEE. Land Use Management and Regulation in Hazardous Areas: A Research Assessment. 1975. ERICKSEN, NEIL J., Scenario Methodology in Natural Hazards Research. 1975. BRINKMANN, WALTRAUD A.R,, et al. Severe Local Storm Hazard in the United States: A Research Assessment. 1975. WARRICK, RICHARD A, Volcano Hazard in the United States: A Research Assessment. 1975. MILETI, DENNIS S., Natural Hazard Warning Systems in the United States: A Research Assessment. 1975. MILETI, DENNIS S., THOMAS E. DRABEK and J. EUGENE HAAS., Human Systems in Extreme Environments: A Sociological Perspective. 1975. 31American National Standard ANSIANS.2.12-1978 Appendix B (This Appendix is not a part of American National Standard Guidelines for Combining Natural and External Man-Made Hacards at Power Reactor Sites, ANSUANS 212-1978, but is included for information only) Exclusion of Combination of Natural and External Man-Made Hazards on the Basis of Low Likelihood of Occurrence ‘The purpose of this appendix is to provide the basis for the choice of 10 per site per year as cutoff probability, below which combinations of events need not be considered for design purposes. In writing the ANS.2.12 standard, one of the objectives was to develop criteria against which design com- binations of natural and external man-made hazards could be judged. Combinations need not be con- sidered in plant design if the effects of the individual hazard components are not additive, if their com- bined consequences to the safety-related structures, equipment or ultimate heat sink are insignificant, or‘if the likelihood of their simultaneous occurrence is negligibly small. Judgments about the effects being nonadditive or the consequences to safety-related structures, equip- ment or ultimate heat sink being insignificant can be made by straightforward engineering evaluation. ‘The likelihood of occurrence of each individual hazard can be estimated or calculated from the existing data. However, in order to eliminate combinations of events from consideration in design, a judgment must be made as to what constitutes a “negligibly small” likelihood of occurrence. ‘The 10° per site per year cutoff probability (actually annual frequency) is justified by either or both of the following reasons (1) The US. Nuclear Regulatory Commission's Standard Review Plan [1], Chapters 2.2.3, 3.5.1.5, ‘and 3.5.1.6, gives an acceptance criterion of 10-7 per year. It is stated therein that if an individual event is calculated to have a realistic estimate of probability of occurrence and potential exposures in ‘excess of the Title 10, Code of Federal Regulations, Part 100, “Reactor Site Criteria,” guidelines of ap- proximately 107 per year, then the event need not be a design basis event (a design basis event is one against which the design of plant safety-related features is evaluated to assure that the event will have no adverse effect). This standard requires the value of 10°* per site per year as the probability of oc- ‘currence only (not considering consequence) for a hazard combination, below which it need not be con- sidered a design basis event. The probability of the consequences of the hazard combinations con- sidered in this standard would conservatively be es:imated to be at least an order of magnitude less _than the probability of occurrence of the hazard combination alone. Therefore, this standard’s 10-* per site per year value is consistent with the 10-7 per site per year value, ie. ‘S107 is met with P occurrence < 10° and Pconseq, < 10 (2) The USNRC Standard Review Plan 2.2.3 also gives an alternative acceptance criterion of 10-* for the accident (consequence) if the basis for the probability estimates can be shown to be conservative. ‘The calculations in this standard are believed to be sufficiently conservative for this purpose. For the general hazards considered in this standard both these conditions are satisfied, therefore the ac- ceptance criterion of 10°* per year is warranted. However, this may not necessarily be true of all com= binations of events. For some events it may be necessary to actually show that the probability of con- ‘sequences exceeding 10 CFR 100 guidelines is at least an order of magnitude less than the probability of the initiating event. Reference for Appendix B [1] USNRC “Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants, LWR Edition,” September 1975.American National Standard ANSUANS.2.12-1978 Appendix C (This Appendix is not « part of American National Standard Guidelines for Combining Natural and External Man-Made Hazards at Power Reactor Sites, ANSI/ANS 2.121978, but is included for information only) Derivation of Formulae for Computing the Probability of the Simultaneous Occurrence of Rare Independent Events 1 Two-Event Overlap Consider two independent processes which produce events at random. The probability density func- tions for the first arrival times, (vi , v2), of the two events are \je"™, and AyeW2 respectively, where), and ), are the respective arrival rates. Let the events be of duration ty, tz and pick the reference time as one year (Y expressed in the same time units). If the events are sufficiently rare that \iY<«1, the probability of event i occurring during any given year is p; =; Y and the probability of the event occurring more than once in a year will be an order of magnitude smaller. (It might be objected that the probability for the double oc- currence of one type of event should be included, since it will generally be of the same order of magnitude as the probsbility of occurrence for two different types of events. However, such an argument does not apply in the present situation for the following two reasons: (a) The simultaneous occurrence of two natural hazard events of the same type is forbidden by ‘mutual exclusion. For example, two different hurricanes cannot occur simultaneously at the same site. (b) Two man-made events of the same type could occur simultaneously at the same site, but the duration of occurrence for man-made events is so short that this probability would be extremely small. In fact, simultaneous occurrences of any two man-made events fall far below our probability threshold, except for cases where there are strong dependencies between the two man-made events.) We can actually obtain a more elegant derivation of the probabilities if we describe the random processes in terms of the times of conclusion u, uz. (In other words u = t; + vi.) ‘To compute the probability of overlap of events 1 and 2, given that they both occur during a specific year, we consider the distribution of ending times, uj, up over the square shown in Figure C-1. Overlap will occur for all points (uy, up pairs) which satisfy t2 > u2—u1 > —ti. These points will all lie within the strip along the diagonal (which is ur = ua) shown in Figure C-1. For the approximation described above the density of points will be uniform over the square, so the conditional probability of overlap (given that both events occur within the same year) is simply the area of the diagonal strip divided by the total area of the square. The simplest way to compute the area of the diagonal strip is to compute the areas of the triangles (on the upper left and lower right of the diagonal) and subtract from the total area of the square. The result of this calculation is: ‘Area of diagonal, 1 t+ overlap strip = Y4— 3(¥—ty?—(¥— to)? = ¥(ti+ta) — > — (Bq. C-1) Dividing by the area Y? and assuming t, and t, < Y we obtain 2 event | Both events occur hte Pr = 4 1 Gectap | ati! your { Y (Eq C-2) With this, we can compute the total probability of a simultaneous occurrence during a given year, as the product of the independent event occurrence probabilities, times the conditional probability of overlap.American National Standard ANSI/ANS2.12:1978 P= pp, (ty + WY (Bq, C-3) If we do not make the simplifying assumptions, the negative exponential probability density functions ». can still be integrated over the diagonal strip shown in Figure C-1. The cumbersome algebra will not be repeated here, but the answer is: 1 PAUP + pre Pal PreP2tv¥, PoePita!¥ Pi + Po +. eet Joea C4) Pi +P where we have used the substitution p; * With the approximations p, < 1, ti < Y this expression reduces to Equation C-3, 2. Three-Event Overlap ‘The probability space for a three-event overlap is « prism around the main diagonal of the cube shown in Figure C-2. The intersections of this prism with the cube faces are rectangles shown in the opposite diagonal corners of Figure C-2. The six edges of the prism are formed by connecting the correspon- ingly lettered corners of the rectangles lying in the cube faces. The three rectangle intersections in the lower left front corner of the cube can be considered as the bases of three prisms formed by dividing up the main prism by planes determined by the main cube diagonal and each of the three axes. ‘This geometrical relationship can be more easily visualized by the view along the diagonal toward the ‘corner shown in Figure C-3. The dividing planes project along the axes t, U,, uy. and the bases of the three prisms are the three rectangles surrounding the intersection of the three axes. It should be noted that the three axes and the three corners of the rectangles all project out toward the reader if we are viewing from the center of the cube toward the corner. > If we assume that tj < Y we can neglect the portion of the volume which is within the distance t; of the corner. We can then compute the volume of the three prisms as if they each had parallel top and bottom bases. In this case they would be parallelepizeds, and the volume of each would be the area of the base multiplied by the altitude (approximately Y). These bases have areas titz, t2tg, and tity so the total volume of all three prisms is approximately Y(titz + tats + tot) and the probability of a three- ‘event overlap, given that all three events begin in a given year is tite + tata + tat, Atl three events begin} Ee (Eq. C5) within one year Pr |S.event overlap ‘The total probability of three-event overlap per year is then tie + bb + th P= ppp, ve (Eq. C-6) ‘This result can be generalized to the simultaneous overlap of n events using more subtle probabil concepts than the simple geometric arguments given above, Pin event overlap) |P,p,---- Pe (Eq. 0-7)‘American National Standard) ANSI/ANS"2 12-1978 3. Duration of Overlap We can complete the description of the simultaneous occurrence (overlap) with the expected duration of overlap (given the occurrence of an overlap in the first place). For the two-event case we have tt B, (overlap duration | overlap occurrence) = — = 7 (Eq. C-8) 1th For three events we have j titats E, (overlap duration | overlap occurrence) = ———"* ___ (Eq. C-9) tite + tats + tot ‘The generalization to the expected overlap duration for n simultaneous events is now quite obvious. ‘The answer can, however, be expressed in a much simpler form if we merely ask the question, “What is the expected (average) time during the year for which the n event overlap will be occurring?” titata te (duration of n event overlap) = <=> (Eq, C-10)American National Standard ANSI/ANS-2.12.1978 ug = Ur + te urs uy us unui ty FIGURE C-1 TWO-EVENT OVERLAP PROBABILITY SPACE Q 36American National Standard ANSI/ANS-2.12-1978 FIGURE C-2 INTERSECTIONS OF CUBE DIAGONAL PRISMS WITH COORDINATE PLANES 37‘American National Standard ANSI/ANS FIGURE C-3 VIEW OF DIAGONAL CORNER FROM INSIDE THE CUBE, VIEWED ALONG THE DIAGONAL‘American National Standard ANSI/ANS2.12.1978 Appendix D (This Appendix is not « part of American National Standard, Guidelines for Combining Natural and External Man-Made Hazards at Power Reactor Sites, ANSUANS.2.12-1978, but is included for information only). Committee Calculations ‘This appendix presents the results of the Committee's calculations. The hazards listed in Section 4, “Natural and External Man-Made Hazards,” are combined using the methodology, equations and values given in Section 6, “Characterization of Hazards,” and Section 6, “Methodology for Combining Hazards.” The probabilities of occurrence of individual hazards used in deriving the hazard com- binations are the upper limits (ie., higher probability) of the ranges in Table A-1. The entries in Fig. 1D. are the probabilities of simultaneous occurrence of the hazards per year per site assuming the in- dividual hazards to be statistically independent events. For example, the probability of simultaneous ‘occurrence of OBE and 100-year wind is calculated as follows fu +) Pen, See probability of occurrence of OBE = 10 per year probability of occurrence of 100-year wind 10 per year estimated duration of OBE = 0.5 min estimated duration of 100-year wind — 360 min total number of minutes in a year = 965 x 24 x 60 = 525600 0.68 x 10-7 per year Fig. D-2 shows the hazard combinations which have joint probabilities of occurrence of greater than 10° per year per site. These combinations have been developed based on the results presented in Fig. D-1 and the assumed dependency relationships between natural and man-made hazards (Fig. 1). Pe fect probabilistic dependency between the events is assumed. Those boxes which contain entries represent hazard combinations with probabilities of simultaneous occurrence exceeding 10°° per year per site. This could mean one of the following two cases: (a) the probability of simultaneous oc- currence of two statistically independent events is larger than 10-® per year and (b) the probability of the casual event in two simultaneously occurring and perfectly dependent events is larger than 10-* per year. For example, the probability of simultaneous occurrence of OBE and SPF (river) is calculated as 4.1 x 10°® per year, assuming the two events are statistically independent. Among depen- dent events, SSE is assumed to trigger a pipe break leading to an explosion. Therefore the probability of simultaneous occurrence of SSE and pipeline accident is assigned the same value as that of the SSE event itself. The entry in Fig. D-2 for this hazard combination is a probability value of 10-* per year. A blank box means the two hazards have a probability of simultaneous occurrence of 10°* per year per site or less. Some of the governing natural hazard combinations in Fig. D-2 are: 1. OBE + SPF (river) 2. OBE + 100-year river flood 3. SPF (river) + 10-year rainstorm 4. 100-year river flood + 10-year rainstorm 5. SPF (river) + 100-year windAmerican National Standard ANSWANS-2.12-1978 6, 10-year river flood + 100-year wind ‘A further investigation of this list using the exclusion criteria of Section 6.0 indicates: Item Remarks 2 Combination is insignificant when compared with combination OBE + SPF (river) (Sec. 6.1.3). 4 Combination is insignificant when compared with the combination SPF + 100-year rainstorm (Sec. 6.1.3). 6 — Combination is insignificant when compared with the combination SPF + 100-year wind. ‘The remaining natural hazard combinations when edded to the combinations of natural hazards with man-made hazards from Fig. D-2, yield the following standard combinations: List of Standard Hazard Combinations OBE + SPP (river) SPP (river) + 100-year rainstorm SPF (river) + 100-year wind SSE + military or industrial facility accident SSE + pipeline accident SSE + surface vehicle explosion SPF + surface vehicle impact PMP + aircraft impact PMP + surface vehicle impact 100-year tsunami + surface vehicle impact 100-year wind + surface vehicle impact 100-year wind + aircraft impact It is to be noted in the above combinations the term SPF refers equally to the river, coastal and lake flood conditions unless indicated otherwise. For structural design, these hazard combinations have been identified as belonging to two basic load categories: severe environmental and extreme en- vironmental loads. ASCE Manual on Structural Analysis and Design of Nuclear Plant Facilities (Ref. 1) has categorized events such as SSE and DBT as extreme environmental loads. The probability of oc- ‘currence of these extreme events range from 10~ per year to 10-7 per year. The lower intensity events of probabilities of occurrence between 10" and 10 per year such as operating basis wind and OBE are classified as severe environmental loads, Hence in this standard it is judged that the hazard com- binations which have probabilities of occurrence larger than 10-* per-year belong to the severe en- Vironmental load category. The combinations that have probabilities of occurrence between 10 and 10 per year are in the extreme environmental loed category. The ASCE manual (Ref. 1) should be ‘consulted for the design strength limits applicable to these load categories. The final list of standard hhazard combinations is: Natural Hazard Combinations Extreme Environmental Load Category OBE + SPF (river) SPF (river) + 10-year rainstorm SPF (river) + 100-year wind Natural and Man-made Hazard Combinations Extreme Environmental Load Category SSE + military or industrial facility accident SSE + pipeline accident SSE + surface vehicle explosion 40‘American National Standard ANSW/ANS2.12-1978 PMP + aircraft impact PMP + surface vehicle impact Severe Environmental Load Category SPF + surface vehicle impact 100-year tsunami + surface vehicle impact 100-year wind + surface vehicle impact 100-year wind + aircraft impact References for Appendix D: 1. J.D. Stevenson (Editor) “Manual on Structural Design and Analysis of Nuclear Plant Facilities,” ASCE, 1976. aAmerican National Standard ANSV/ANS-212.1978 Barante Hl} Hulud ? Figure D-1 Probability of Occurrence Per Year of Hazard Combinations Assuming Statistical Independence Between Hazards 42 Ls oon‘American National Standard ANSY/ANS-2.12.1978 on wow | 2 tae i u|d Wt Hazard Co 10 Per Year ( Figure D-2 ions with Probability of Occurrence Larger Than 4 the Dependency Relationships Shown in Sec. 5.2)‘American National Standard ANSUANS-2.12.1978 Appendix E (This Appendix is not a part of American National Standard Guidlines for Combining Natural and External Man-Made Hazards at Power Reactor Sites, ANSI/ANS-2 12-1978, but i inckuded for information only) Example of Using Design Approach for Combining Hazards for a Specific Site ‘This appendix is presented as an example of how a designer may use this standard to develop a list of hazard combinations for his site. A hypothetical river site was chosen as the example. Assumptions (1) Light water reactor of current design. (2) Heat sink is an evaporative, natural draft cooling tower with makeup from the river. A pond located at a distance from the river acts as the ultimate heat sink. The pond is located away from the river because of low river level considerations. (3) Commercial traffic uses the river (4) Site studies have been made by the applicant in enough detail to meet licensing requirements. Approach ‘The approach and numbering system used below follows Section 8, “Design Approach for Combining Hazards,” of the standard. Step 8.1.1 Make study of site to identify individual natural and man-made hazards which have poten- tial for damaging the nuclear plant structures, or ultimate heat sink. Consider individual hazards listed in Section 4, “Natural and External Man-Made Hazards,” as a minimum. Conclusion On the basis of the utility's site evaluation study, it was concluded that the following hazards listed in Section 4 exist at this site. Appropriate Natural Hazards Earthquake River Flood Rainstorm Tornado Wind Appropriate Man-made External Hazards ‘Aircraft Impact Surface Vehicle Explosion ‘The hazards listed below from Section 4 are not appropriate to this site for the reasons stated. Nonapplicable Natural Hazards Coastal Flood (not on coast) Lake Flood (not on lake) ‘Tsunami (not on coast) Waves (not on coast) Nonapplicable Man-made External Hazards Industrial or Military Facility Accident (none located within 10 miles) Pipeline Accident (none located within 10 miles)‘American National Standard ANSYANS 2.12-1978 Retaining Structure Failure (upstream dams on river are seismically qualified) Surface Vehicle Impact (ship or barge cannot hit any safety-related structure or ultimate heat sink. Railroad track and roadway into plant not in direct line with safety-related structures.) ‘Step 8.1.2 Review individual hazards in 4.2 to determine whether any should be added to those iden- tified in Step 8.1.1, because of unique site or plant features. Conclusion ‘Toxic gas due to a collision on the river must be considered. Control room habitability for this hazard shall be provided. Does not enter into further hazard combination analyses. il spill due to tank-barge collision on the river near the plant must be considered. Design ultimate heat sink makeup system so that it can be isolated and provide sufficient capacity for duration of fire. ‘Does not enter into further hazard combination analyses. Alll other hazards listed in 4.2 have been considered and it is concluded that they need not be added to the list of hazards. ‘There are no other hazards in addition to those listed in Section 4 and 4.2 for this site. Step 8.13 The final list of individual hazards is therefore: ‘Natural Hazards Earthquake River Flood Rainstorm ‘Tornado Wind Man-made Hazards Aircraft Impact (aircraft crash into plant structures. Airport is less than 5 miles from plant.) Surface Vehicle Explosion (collision of oil tank-barge with another barge in front of plant; 1000 ft. away) Step 8.1.4 Develop a summary similar to Table A-1 for the natural hazards listed in Step 8.1.3. The ‘asterisks mean these values are different from the Table A-1 upper limit values as concluded by the ac- tual site studies.American National Standard ANSIANS2.12:1978 ‘The summary table ii Estimated Natural Hazard Probability Maximum Hazard Definition Per Year Duration Earthquake SSE ws 60 see OBE 10 30 sec. River Flood DBF 107 1 month* ‘SPF 103 2 weeks* 100-yr. 10? 4 days* Rainstorm PMP 10-5* 1 day 100-yr. 0? 1 day ‘Tornado DBT 107 2 min. Wind 100-yr. 10 6 hours Step 8.15 Develop a summary table similar to Table A-2 for the man-made hazards listed in Step 8.1.3, Asterisks mean these values are different from Table A-2 values as concluded by actual site studies. ‘The summary table is: Man-made Hazard Airports within 5 miles of plant Aircraft impact (small aircraft) Fire (small aircraft) Airports more than 5 miles from plant Aircraft impact (large aircraft) Fire (large aircraft) Surface Vehicle Explosion Oil tank-barge collision (blast and missiles) Probability (per year) 10% 05 0) 10% 05 5x10 5 sec. 30 min. 5 sec. ‘Step 8.1.6 The tables above for natural and man-made hazards are not the same as Tables A-1 and A- 2 of Appendix A (as indicated by the asterisks). Therefore this step cannot be followed. Step 8.1.7 Because the tables are different from Tables A-1 and A-2, the standard combinations of Section 7, “Standard Hazard Combinations,” cannot be used for this site. A list of hazard com- binations must be developed specifically for this site. The methodology of Appendixes C and D is used to derive the list. This approach yields probability values for each of the combinations. A figure similar to Figure D-1 in Appendix D would result, except that it would only include hazards specific to this Note: (1) Conditional probability given that airraft impact has ooesrred 46‘American National Standard ANSYANS.2.12:1978 site. However, there is less concern with the exact probability of a combination than there is with those combinations which have a probability of less than 10 per year so they can be eliminated. ‘The values for this site are compared to those in Tables A-l and A-2 to determine if the individual hazards for this site are less or more probable than the values in Table A-1 (upper limit values) and ‘Table A-2. For any of the values which are less probable, the values of Tables A-1 and A-2 can be used as a conservative estimate. This would avoid having to perform the calculations. Alternatively, the ac- tual combined probability can be calculated to see if the less probable hazard results in a combined probability of less than 10° per year and hence need not be considered. For illustration purposes, the comparison method is used. The list below includes only those hazards (having an asterisk in the two summary tables above) which are different from the Table A-1 and A-2 values; all other hazard values are the same. Is actual hazard more probable or less probable than Table A-l or A-2 value which was used to calculate List of Standard Hazard ‘Combinatioi Earthquake SSE Less OBE Less River Flood (SPF) Less Rainstorm (PMP) Less Tornado Same Wind Same Aircraft impact (<5 miles) Less Aircraft fire (<5 miles) Less Aircraft impact (5 miles) Less Aircraft fire (5 miles) Less Surface Vehicle Explosion More ‘The surface vehicle explosion is the only one of the hazards which has a higher probability of oc- currence than the Table A-1 and A-2 values. In performing this comparison, the durations of the in- dividual hazards and the Table A-1 and A-2 durations should also be compared. A 3.step process can now be followed to arrive at the list of hazard combinations appropriate to this site. Step 1. Go to the Section 7 list of standard combinations and delete any which include an individual hazard which is not applicable to this site (tsunami, for example). The remainder apply to this site as a ‘minimum, unless they can be eliminated by Step 2. Step 2. For any one of the individual hazards which has a lower probability of occurrence than the ‘one included in the Section 7 list of standard combinations, the probability of this hazard combining with the other hazards can be calculated. This calculation may eliminate further combinations from the list of standard combinations for this site if they are less than 10-® per year. However, in this exam- ple, it was chosen not to perform the calculations and instead to accept the appropriate Section 7 list as a conservative estimate. Step 3. For any one of the individual hazards which has a higher probability or longer duration than the one included in the Section 7 list of standard combinations (eg., the surface vehicle explosion), the probability of its occurrence with each of the other hazards must be calculated. Equation 6-1 in 6.2 is used for the calculation. 47