ML13302B394

1990 Offsite Hazards Update,Summary Rept

ML13302B394
Person / Time
Site:	San Onofre
Issue date:	02/28/1991
From:	Southern California Edison Co
To:
Shared Package
ML13302B396	List: ML13302B394 ML20070S447
References
NUDOCS 9104020294
Download: ML13302B394 (13)

Text

SOUTHERN CALIFORNIA EDISON SAN ONOFRE NUCLEAR GENERATING STATION 1990 OFFSITE HAZARDS UPDATE Summary Report Prepared by:

ERIN Engineering and Research, Inc.

2175 N. California Blvd., Suite 625 Walnut Creek, CA 94596 (415) 943-7077 February 1991 9104020294 910228 PDR ADOCK 05000:36.1 R

PDR

TABLE OF CONTENTS Section INTRODUCTION.....

1 METHODOLOGY.....

HAZARD IDENTIFICATION.....

2.

1 HAZARD EVALUATION 3.............2 ACCEPTANCE CRITERIA................

5

SUMMARY

OF RESULTS.

6 TOXIC CHEMICALS......

....6 SHIPMENT SIZE...........................

6 FLAMMABLE AND EXPLOSIVE MATERIALS.........

REFERENCES...........................

9 1

INTRODUCTION The 1990 Offsite Hazards Analysis Update documents SCE's review of hazards posed to the San Onofre Nuclear Generating Station (SONGS)

Units 2 and 3 from transportation of hazardous materials by the facility.

The 1991 report represents the third tri-annual update of hazardous materials transported by the SONGS site on Interstate Highway 5 and the Atchison, Topeka and Santa Fe Railway.

The following summarizes the purpose of this study, methodology used to perform the study, and results obtained.

The United States Nuclear Regulatory Commission requires that a utility filing an application for authority to operate a nuclear power plant include in the application information regarding potential offsite hazards which could affect safe operation of the plant. The hazards to be specifically evaluated include the potential release of toxic gas materials in close proximity to the plant in which air intake could potentially jeopardize continued occupancy of the control room.

In addition, consideration must be given to the potential release of explosive chemical mixtures in which detonation could adversely 'affect the plant's physical structure and, consequently, plant safety.

For San Onofre Units 2 and 3, the analysis of offsite hazards is documented in Section 2.2.3 of the San Onofre Nuclear Generating Station Units 2 and 3 Updated Final Safety Analysis Report (UFSAR).

Upon original submittal and review of this analysis, the NRC noted the dependence on supporting data regarding the size and frequency of hazardous cargo shipments.

As a result of the potential variability in transport rates, cargoes, accident rates, and shipment sizes, the NRC requires the San Onofre hazards analysis to be updated every three years by Technical Specification 6.9.1.14.

The first offsite hazards analysis update was performed in 1984. The analysis was updated again in 1987.

The purpose of this report is to document the nature, frequency, and size of hazardous material transported near the San Onofre site in 1990 and update the quantitative analysis of off-site hazard frequency based on these findings.

METHODOLOGY The methodology for evaluation of the potential risk to the safe operation of the San Onofre Nuclear Generating Station Units 2 and 3 from the shipment of hazardous cargo past the plant is described in the San Onofre Nuclear Generating Station Units 2 and 3 UFSAR Section 2.2.3.

The evaluation consists of determining hazards posed by truck traffic on Interstate Highway 5 and railway traffic on the Atchison, Topeka and Santa Fe Railway. The highway is located within 560 feet and the railway within 490 feet of the closest safety-related plant structure.

Those types of hazards resulting from the shipment of hazardous materials past the plant are categorized as follows:

individual toxic gases which could reach the control room in concentrations which exceed their toxic limit and could cause the control room to become uninhabitable, 1

solids, liquids or gases (including transient vapor clouds) which could explode and create an overpressure greater than the design allowable at plant safety-related structures which could affect plant safety, and solids, liquids or gases which could ignite and yield high temperatures or generate other hazardous chemical bearing clouds which could also affect plant safety.

HAZARD IDENTIFICATION The update of the offsite hazards analysis was based on information obtained from: 1) the railroad and military authorities believed to be transporting hazardous materials, and 2) a roadside survey of truck traffic. In both cases, questionnaires were used to obtain specific information related to the type of material being transported, as well as shipment size and frequency.

The roadside survey was conducted over-a two week period from September 17-30, 1990, at the truck weigh stations located on either side of Interstate Highway 5 (1-5) approximately three miles south of San Onofre.

There are no highway entrances or exits between the weigh stations and the plant site.

Survey personnel were stationed at the weigh stations to determine truck content when trucks passed through the weigh stations. All trucks entering the weigh stations were considered. Those trucks displaying a hazardous cargo placard were detained briefly and questioned regarding the types and quantities of materials being transported. A separate data sheet was used to record the pertinent information on each truck displaying a hazardous cargo placard.

Data obtained during the two week survey was extrapolated to estimate the annual number of shipments of each hazardous material. The survey was performed during periods when the weigh station was open.

To account for hazardous material shipments when the weigh station was closed and estimate the annual number of shipments a correction factor was calculated based on observed truck traffic densities during daytime, evening, nighttime, and weekend hours.

It was estimated that for each hazardous material shipment observed during the two week survey that approximately 40 shipments would occur over the course of one year.

This estimated annual frequency was used in the calculation of hazardous material risk.

In order to update the railroad and military hazards, the appropriate authorities were contacted through correspondence.

Letters were sent to the following agencies:

United States Marine Corps - Camp Pendleton, CA Department of the Navy - Seal Beach, CA Department of the Navy - San Diego, CA Department of the Navy - Concord, CA Atchison, Topeka and Santa Fe Railroad - Topeka, Kansas 2

Information relating to shipment of hazardous and explosive material being shipped by truck or rail was requested.

Agencies were requested to provide annual shipment numbers as well as average and maximum net explosive weights for explosive material shipments.

Phone conversations were held as necessary to supplement and clarify written correspondence.

HAZARD EVALUATION The risk or.frequency of hazard associated with each of the three categories of hazard is evaluated using the following assumed relationship:

Risk - Frequency of Transport x Probability of Consequence where the probability of consequence is equal to the product of the event probabilities required for the consequence to occur.

The individual event probabilities contributing to the toxic gas hazard include 1) the probability of an accident for specific types of hazardous material per highway or railway distance, 2) the exposure distance along the highway or railway where the accident could lead to a hazard at the plant, 3) the probability of a spill given that an accident has occurred and 4) the probability of the wind blowing from the accident site to the plant.

The individual event probabilities contributing to the explosion hazard include

1) the probability of an accident for specific types of hazardous material per highway or railway distance, 2) the exposure distance along the highway or railway where the accident could lead to a hazard at the plant, 3) the probability of a spill given that an accident has occurred, 4) the probability of an explosion given that a spill has occurred, and 5) the probability of a boiling liquid expanding vapor explosion (BLEVE) given a spill.

The individual event probabilities contributing to the flammable cloud at the plant hazard include 1) the probability of an accident for specific types of hazardous material per highway or railway distance, 2) the exposure distance along the highway or railway where the accident could lead to a hazard at the plant, 3) the probability of a spill given that an accident has occurred, 4) the probability of a vapor cloud fire given that ignition occurred,

5) the probability of a vapor cloud explosion given the ignition occurred, and 6) the probability of the wind blowing from the accident site to the plant.

The hazard frequency is determined for each offsite event potentially capable of leading to a degradation of plant safety.

The risk associated with each type of hazard is assumed to be directly proportional to the frequency of shipment of each hazardous.material.

Thus, the risk can be updated by multiplying the original risk by the ratio of updated hazardous shipment frequency to original hazardous shipment frequency. The frequency of each type of hazardous event is then compared to the guidelines in the Standard Review Plan (SRP) Section 2.2.3 to determine whether the event should be considered a design basis event.

3

Approximately 100 new potentially hazardous materials were identified in the survey. An in-depth review of each material was conducted in order to assess whether or not an accident involving these materials might be capable of producing an explosive, fire or toxic gas hazard.

The review identified the chemical composition and physical state of the material as well as a description of its use and/or characteristics.

These characteristics permitted an identification of the degree of hazard as described below:

NONE -

No Harm via Exposures of Normal Use LOW Causes Some Discomfort MOD Can Cause Considerable Discomfort HIGH -

Incapacitating and Poisonous Including Potential for Explosive Impacts The identification of material characteristics permitted the categorization of potential new hazardous materials into one of the following groups:

1. Identified as non-hazardous; deleted from further consideration;

2. Evaluated to be non-hazardous; or

3. Categorization by physical properties to be similar to a previously evaluated hazard.

Only 20 of the new hazardous materials were determined to potentially pose a hazard to the facility.

Evaluation of the physical properties for these materials allowed them to be categorized with previously analyzed hazards. The surveyed number of shipments of each material was added to the similar material shipment number obtained from the survey. Thus, the shipment frequency of each previously evaluated material was increased to include any similar substances.

Solid materials were excluded from further consideration as inhalants due to their poor dispersion capabilities. In this manner, all of the new potentially hazardous materials which were identified were dispositioned.

The revised shipment frequencies (including similar substances) were then divided by the frequencies found in the original study to determine the ratio of the new shipment frequency to the old frequency. This ratio was then multiplied by the original hazard frequency to obtain the updated hazard frequency.

In the case of gases, identified commodities could either be grouped with an existing analyzed toxic gas or were otherwise shown to be non-hazardous.

The revised hazard from toxic gas was calculated by multiplying the original hazard probability by the ratio of the 1990 survey shipment frequency divided by the original shipment frequency for each material.

4

In the case of the explosion hazard, the shipment size affects the percentage of shipments capable of causing an excessive overpressure at the plant.

All substances capable of creating an explosive overpressure in excess of the predetermined value of 7 psi at the nearest safety-related plant structure were evaluated together as a single event and the result compared with the SRP acceptance criteria. As part of this study SCE is reconfirming that all safety related structures were considered in the overpressure evaluation.

We are planning to complete this effort by December 31, 1991.

The 1990 Offsite Hazards Update also employed the use of site specific truck accident data.

The 1981 Science Application Inc.

(SAIC) analysis and two subsequent updates used truck accident data from a survey conducted in the Denver, Colorado area. Site specific truck accident data taken from 1982 through 1989 in a ten mile section of Interstate Highway 5, five miles north and south of San Onofre, was used for this analysis. The truck accident frequency obtained using site specific data was 1.4 x 10-6 truck accidents per truck mile per year.

Other factors affecting the frequency of plant hazard such as conditional spill probability, source emission rates, dispersion modeling, etc., were not changed during this update.

ACCEPTANCE CRITERIA The SRP states that initiating events leading to potential consequences in excess of 10 CFR Part 100 exposure guidelines should be estimated using assumptions that are as representative of the specific site as practicable.

Accordingly, the expected rate of occurrence of potential exposures in excess of the 10 CFR Part 100 guidelines must not exceed 10-7 per year based upon a realistic analysis, or 106 per year based on a conservative analysis.

Any events exceeding these guidelines are classified as design basis events. Each hazard classified as a design basis event is reviewed to determine that the effects of the event on the safety features of the plant have been adequately accommodated in the design of the plant.

The SONGS Offsite Hazards Analysis is judged to be a conservative evaluation of plant hazards, based on the application of the following conservative assumptions:

1.

All explosions from flammable liquid spills from tank trucks were assumed to be fuel-air detonations, yielding the maximum possible pressure.

2.

For hydrogen and acetylene, it was conservatively assumed that the maximum yield is 100% of the TNT equivalent weight.

3.

Total release of all hydrogen and acetylene gas from all cylinders in an accident was assumed.

5

4.

Release statistics used did not distinguish between the more likely mechanism of a small rupture or crack (resulting in a minimal leakage and/or leak rate) and the less likely severe rupture (which presents the more significant hazards to the plant).

5.

Effective length of highway did not include realistic appraisal of dilution at the site (i.e., effects of ground roughness and topography).

6.

No credit was taken for control room air volume dilution or mask breathing devices.

The hazards analysis methodology assumes the operators are disabled in the event that a sufficient concentration of toxic material exists at the control room air intake. This is conservative in that it does not credit mixing.of the toxic material with the control room air volume which will act to dilute the toxic concentration.

In addition, operators have access to protective breathing devices which will limit exposure and reduce the likelihood of being disabled.

Based on the application of these conservative assumptions, the threshold value for classification of design basis events is 10.6 per year.

Any hazard exceeding this value will be evaluated to determine that the effects of the event on plant safety features have been adequately accommodated in the plant design.

SUMMARY

OF RESULTS TOXIC CHEMICALS Table 1 illustrates the results of the updated chemical hazard frequency. The estimated number of annual shipments of hazardous chemicals and associated frequency of plant hazards are provided.

The frequency of hazard for each toxic material is less than lx10 6 per year, with the exception of propane and anhydrous ammonia.

The threshold of lx10'6 per year is applied based on the use of conservative assumptions in the analysis. Monitoring and automatic control room isolation provisions are provided for propane, ammonia and chlorine by the Control Room Toxic-Gas Isolation System.

Thus, the largest unprotected chemical risk is from formaldehyde at 3.9E-7 per year.

This risk value allows a 2.5 safet factor in shipment frequency prior to reaching the threshold value of 1.0 x 10 per year.

SHIPMENT SIZE The maximum shipment size for individual hazardous material shipments was reviewed with respect to shipqent sizes evaluated in the baseline offsite hazards analyses performe in 1981 J for toxic materials and a separate study also performed in 1981 for explosive and flammable materials. Increased shipment sizes were found for the following hazardous materials:

6

FREQUENCY OF HAZARDOUS ANALYZED SHIPMENT 1990 OBSERVED HAZARD MATERIAL SIZE SHIPMENT SIZE (per year)

Uquid Hydrogen 8500 gal 15,000 gal 1.3x108 Carbon Dioxide 1500 gal 5,300 gal 2.4x10 Nitrogen (gas) 226 sd 4.86x106 d

<1.0x10 Nitrogen (liquid) 8000 gal 11,700 gal 5.1x10 9 Naphtha 2000 gal 10,000 gal 3.9x10 Isopropyl Alcohol 4500 gal 7,500 gal 8.6x10_

Standard Solvent 4500 gal 6,000 gal 9.1x108 Perchloroethylene 4000 gal 5,550 gal 7.6xi0 Methylene Chloride 2000 gal 4,000 gal 5.1x109 Each material was evaluated on a case-by-case basis to determine the impact on plant risk.

The increased shipment sizes were found to have either negligible or no impact on plant risk.

For each of the materials the plant risk was calculated to be well below the 10-per jear threshold. The actual risk values were all found to be less than 1.0 x 10 per year.

FLAMMABLE AND EXPLOSIVE MATERIALS Table 2 illustrates the results of the updated flammable cloud and explosive hazards.

Explosive materials identified in the survey include Liquified Petrolium Gas (LPG), liquid hydrogen and acetylene. These materials also have properties which pose a flammable vapor cloud hazard. As such, the chemicals were evaluated for their contribution in both explosive and flammable vapor cloud risks. Other compressed and liquified gases were also considered for evaluation of the risk of a flammable gas mixture reaching the control room HVAC air intake.

Compressed hydrogen and Liquid Natural Gas (LNG) were identified in the 1987 offsite hazards update but were not observed in the 1990 survey. In addition, no chemicals other than LPG, liquid hydrogen and acetylene were observed in the 1990 survey which could pose a flammable cloud hazard.

The contribution of these materials to the flammable cloud hazard was calculated to be 4.9 x 10-7 per Xear.

The resulting frequency of explosive hazards was calculated to be 7.7 x 10 per year. Each of these values is less than the 1 x 10-6 per year threshold.

It is noted that the original NUS offsite hazards analysis performed in 1979 3 evaluated the potential hazard of all flammable liquids found.

The original analysis determined that only formaldehyde, gasoline, and xylene were capable of causing a peak positive normal reflected overpressure (from explosion) at the nearest plant safety-related structure exceeding 3.0 psi.

Since the capability of safety-related structures at the plant to withstand overpressure conditions was re-evaluated and determined to be 7.0 psi, the potential hazard capability from spills of formaldehyde, gasoline, and xylene was re-evaluated.

Based on the 7

1990 survey, these materials are not shipped in quantities which could create an overpressure in excess of the 7.0 psi limit.

Therefore, the risk from this particular hazard is zero.

In order to evaluate the explosive hazard from shipment of military ordnance on 1-5 by San Onofre, questionnaires were sent to the U.S.

Marine Corps, Camp Pendleton, California, and Department of the Navy, Naval Weapons Station, Concord, California.

Based on their responses, the maximum weight of any military shipment of explosives past San Onofre Nuclear Generating Station Units 2 and 3 for 1987 through 1990 was less than 14,600 pounds.

Assuming this maximum weight shipment is composed of materials having a pound per equivalence with TNT (conservative assumption), the peak positive normal reflected overpressure at the plant site produced by the surface detonation of such a shipment at a distance of 560 feet (distance to closest safety-related structure, the Fuel Handling Building) is approximately 4.7 psi. Since the safety-related structures of the plant have been determined to be capable of withstanding overpressures of up to 7.0 psi, an explosion of a shipment of military ordinance on 1-5 does not pose a hazard to plant safety.

The Atchison, Topeka and Santa Fe railway reported a substantial increase in shipments of Liquified Petroleum Gas (LPG) from the previous update in 1987.

LPG is classified as both explosive and flammable. The increase in shipments from 124 in 1987 to 1499 in 1990 does not result in the plant hazard exceeding the SRP acceptance criteria of 1.0x10 6 per year for either explosive hazards or flammable hazards. The increase in shipments from the 1987 update was a result of a contractual agreement with Santa Fe railway to supply LPG to a private manufacturer in the San Diego area.

The current hazard margin allows approximately a 50% increase in shipment frequency before additional analyses are warranted.

In summary, the frequencies of potential hazards related to the shipment of hazardous materials on the highway and railway adjacent to the plant including toxic, explosive and flammable cloud hazards were all calculated to be either less than lx106 per year, determined not to be a hazard on some other defined basis, or protective equipment already exists to mitigate the consequences of the hazard. Based on these results no new design basis events from offsite hazards need to be considered for the San Onofre Units 2 and 3.

8

REFERENCES

1.

"Analysis of the Probability of Toxic Gas Hazard for the San Onofre Nuclear Generating Station as a Result of Truck Accidents Near the Plant,"

February 28, 1981, E. A. Hughes, A. J. Unione, and R. R. Fullwood, SAIC report.

2.

"Analysis of Explosive Vapor Cloud Hazards for Rail and Highway Transportation Routes Near the San Onofre Nuclear Generating Station Units 2 and 3 Using a Best Estimate Analysis," R. H. Broadhurst, and M. C.

Cheok, Supplement 1, April 17, 1981.

3.

"Analysis of Explosive Vapor Cloud and Missile Hazards for Rail and Highway Transportation Routes Near the San Onofre Nuclear Generating Station Units 2 and 3," R. H. Broadhurst and C. Y. Li, NUS-3367.

9

Table 1 1990 CHEMICAL HAZARD FREQUENCY Chemical Estm ted Anual Shpment Freqmny J Frequen of plant Hazard (per ye)

Acetone 4.1E46 Acetylene 472 8.5E4 Ammonia 363 2.3E4 Argon 472 6.5E 5.8E-9 Benzene_

0

< 1.0E49 3 0 3 1.2 E.7 Bu Acetals 157 1.0E4 Carbon Doxide 2,240 2.4E Chlorine 707 8.4E-7 Crude Oil 1,061.9E 01e601 Oil 2387.8 Formalehyde 157 3.9E-7 Gasoline 8,725 3.5E-7 Hydraulic 0

1.0E-9 Hydrochloric Acid 29 1.4E4 Hydrogen 230 1.3E4 o

.=1,336 8.68E4 Jet Fuel Z437 9.3E4 Methyl Bromide 238 1.4E4 Methyl Ethyl Ketone 314 2.0E4 Metylene Chloride 79 5.1 Motor Oi 500 3.9E.

M dc Acid 79 E

Naphtha 354 Nrogen 2.751 5.1E-0 Perchlorelu 116 7.684 Propane 2,12 1.4406 SuluriAcid 314 6.60 Toluene (Soleno 1,415 9.1E4 Vikane 157 1.0E4 Xylene 39 4.5E4 10

Table 2 1990 UPDATED FREQUENCY OF FLAMMABLE CLOUD AND EXPLOSIVE HAZARDS 1990 Frequency of Flammable Vapor 1990 Frequency of Substance 1990 Equivalent Cloud at Plant 7 psi Overpressure Annual Shipments (per year)

(per year)

LPG - Hwy 2555 1.1E-7 7.5E-8 LPG - Rail 1499 3.4E-7 6.8E-7 Liquid Hydrogen 236 9.1 E-9 6.8E-9 Compressed NO

<1.0E-9

<1.0E-9 Hydrogen Acetylene 472 2.7E-8 1.8E-9 LNG NO

<1.0E-9

<1.E-9 Total:

4.9E-7 7.7E-7 Notes:

• NO" indicates no shipments of the material were observed during the survey.

11