ML20028G494

Monitoring of Liquefied Natural Gas & Liquefied Petroleum Gas Shipping & Const Activities on DE River.

ML20028G494
Person / Time
Site:	Hope Creek, 05000355
Issue date:	01/31/1983
From:	Public Service Enterprise Group
To:
Shared Package
ML20028G493	List: ML20028G492 ML20028G494
References
NUDOCS 8302090412
Download: ML20028G494 (46)

Text

-- ,_

s I . .

l h0PE CREEK GENERATING STATION MONITORING OF LNG AND LPG SHIPPING AND CONSTRUCTION ACTIVITIES ON THE DELAWARE RIVER i

l

+

Docket Nos. 50-354 50-355

M P83 4/18 1-df January 1983 1

8302090412 830201 DR ADOCK 05000354 PDR

TABLE OF CONTENTS

1.

SUMMARY

2.- INTRODUCTION

3. YEARLY REPORT 3.1 Total Number of Ships Per Year 3.2 -Construction Activity

4. BI-YEARLY' REPORT' 4.1 Calculation of Conditional Probability 4.1.1 Accident Rate Per Ship Mile 4.1.2 ' Historical Collision Analysis FY79-FY80 4.1.3 Collisions With Fixed Objects 4.1.4 Spills Per Accident 4.1.5 Vapor Cloud Formation Per Spill 4.1.6 Meteorological Factor 4.1.7 Conditional Probability Calculation 4.2 Conclusions 2

1 5

4 l

l

( M P83 4/18 2-df

~ -

l 1

4 -

1.

SUMMARY

l - On November 4, 1974, the Atomic Energy Commission issued -

! Ja Construction Permit (CP) for the construction of the Hope Creek Generating Station in' Salem County, New j Jersey.

l

.The Hope' Creek site is located on the Delaware River estuary near the southern end of an artificial peninsula known as Artific'ial Island.

i Amendment No. 5 to-the Construction Permit requires PSE&G to monitor activities related .to the shipment of liquefied natural gas (LNG) and liquefied petroleum gas

(LPG) on the Delawara River. This monitoring includes

, shipping rates of. liquefied flammable gases and various j types of construction activity. The_ chemicals classi-fled as liquefied flammable gases.are propane, butane, butadiene, vinyl chloride and liquefied natural gas.

The results of this study show that the total condi-tional probability of a flammable vapor cloud reaching -

the Hope Creek plant is 1.5;x 10-7 for 1980 and 1.2 x 7 10-7-for 1981. Both values are approximately night times lower than the 10-6 value established as an~ upper

-limit by the NRC, based on conservative assumptions..

Therefore, the associated-hazards to the health and

-safety of f the publ'ic are. negligible, i

i. 2. INTRODUCTION i

!- The activities of interest are those which might result in an increased rate of shipping or which could cause an

increased probability of accident occurrence.

An increase in the shipping rate would occur either should an existing facility increase its' importation rate or because of the construction-of a new storage or-refining facility.- PSE&G has monitored these' events by keeping-in touch with local authorities, contacting

those refineries and import terminals'on the Delaware ,

i River which are capable of receiving liquid fuels and by reviewing the "Public Notices" issued by the Department of the Army "U.S. Corps of Engineers."

a t

M P83 4/18 3-df i

, ,- - , _ , _ _ . _ . ,,-m_,.,.._,, . _ , , _ _ , . . . , . _ . . - - - - , _ . . - - _ - - . - - , ,

_-,.._,-.,,mm.... , ..

Although there is currently no LNG shipping on the river, there have been proposals to initiate such shipping. All have either been withdrawn or re-jected by the Federal Energy Regulatory Commission (FERC). Consequently, there are currently no out-standing proposals for an LNG facility that could possibly affect the Hope Creek Generating Station.

PSE&G is continuously monitoring the applications received by FERC relating to LNG terminals.

Other types of construction on the Delaware River could also result in increase risk to the Hope Creek Generating Station. It has been estimated (PSE&G, Dockets 50-354 and 50-355 before the NRC, exhibits 9, 10, 11) that a flammable vapor cloud which forms as a result of an accidental spill of a liquid fuel on water could travel up to 12 miles. Therefore, any spill occurring within a distance of 12 miles up or downstream of the nuclear facility has to be analyzed to determine if it presents a potential hazard to the plant. Currently, with the exception of Tower 97,* there are no rammable objects in the vicinity of shipping channel near Hope Creek.

However, in the future, docks, jetties, moorings, piles, or other potential obstructions could be constructed in the river. Therefore, PSE&G is monitoring construction activity of this type.

The calculation of the probability of a flammable cloud reaching the Hope Creek plant is made from a 4

series of conditional probabilities. These proba-bilities involve the expected number of accidents per mile of river transit, the probability that a spill will result given that an accident occurred, the probability that a vapor cloud will form given that there has been a spill, and the meteorological factor.

The NRC has established guidelines for the accept-able upper limit of the probability that the Hope Creek Generating Station will be affected by a flam-mable vapor cloud formed as a result of an accident on the river. In an estimate in-which the factors are determined based on conservative approximations, the number of incidents cannot exceed 10 6 per year.

• Electrical Transmission Line Tower M P83 4/18 4-df

.. . . ._. . ~ - .. -. . .

s HOPE CREEK GENERATING STATION MONITORING OF LNG AND LPG SHIPPING AND CONSTRUCTION ACTIVITIES ON THE DELAWARE RIVER

'T i

3. YEARLY REPORT - 1981 (YR-3) o i

l Docket Nos. 50-354 50-355 M P83 4/18 5.*-df January 1983

s 3.1 Total Number of Ships Per Year The total number of ships passing by the nuclear plant in 1980 and 1981 and carrying each of the designated chemicals was developed by Poten &

Partners, Inc., a shipping consultant. The data was derived from U.S. Coast Guard Records, Import and Export Records, and the U.S. Army Corps of Engineers. The list presented in Table 1 specifies the discharge date, quantity, product, vessel, origin and, where possible, the importer. This information is summarized in Table 2, and graph-ically presented on Figures 1 through 5. As indi-cated, the total number of tankers passing by the plant was lower in 1980 and 1981 than in previous years.

Also, the shipments of vinyl chloride have ceased.

This is due to the fact that the manufacturing facility, located in Puerto Rico, from which the vinyl chloride was being imported, went out of business.

3.2 Construction Activity The construction activity along the Delaware River was monitored through the review of "Public Notices" issued by the U.S. Army Corps of Engi-neers, by contacting various facilities located upstream of the plant (see Table 3 and Figure 6) and the Federal E3ergy Regulation Commission, which is responsible for reviewing and approving any pro-posals for construction of LNG terminals.

9 M P83 4/18 6-df

LIOUEFIELD GAS CARGOS TRANSITING THE DELAWARE RIVER - 1981 TABLE 1 DISCHARGE QUANTITY DATE (METRIC TONS) PRODUCT VESSEL ORIGIN IMPORTER 2/8/81 11,350 Propane Hoegh Sword Kuwait Elf / Sun 2/8/81 7,650 Propane Hoegh Sword Saudi Arabia MSK/ Sun 2/24/81 3,200 Butadiene Garbeta Netherlands Exxon 2/24/81 5,700 Butadiene Garbeta Netherlands Exxon 4

2/24/81 1,900 Butadiene Garbeta United Kingdom Exxon

. 2/24/81 3,400 Butadiene Garbeta France Exxon 3/11/81 6,125 Butane Katrisa Venezuela Gulf Oil

4/18/81 10,410 Butane Mundogas Pacific Venezuela Warren i

6/10/81 United Kingdom 3,970 Butadiene Pascal l 6/20/81 6,777 Butane Clerk Maxwell Venezuela Warren 6/25/81 3,019 Butadiene Nestefox Terneuzen g 7/19/81 20,000 Propane Monge Saudi Arabia Sun i

7/25/81 4,039 Butadiene Pascal Netherlands Paulsboro 7/25/81 1,645 Butadiene Pascal Netherlands 7/25/81 1,244 Butadiene Pascal Netherlands Paulsboro

7/25/81 1,256 Butadiene Pascal Netherlands 7/25/81 1,244 Butadiene Pascal Netherlands 9/9/81 8,245 Butane Devonshire Saudi Arabia Mitsui 9/15/81 5,483 Butane Devonshire Saudi Arabia Mitsui 10/25/81 11,631 Propane Mundogas Pacific Saudi Arabia Mitsui l 11/25/81 13,000 Butane Mundogas Pacific Venezuela Warren 12/10/81 4,500- Butadiene Sine Maersk Netherlands i 12/12/81 1,993 Butadiene Linge Gas Netherlands

, 12/20/81 15,000 Butane Hoegh Skean Saudi Arabia Warren 12/27/81 13,100 Butane Luigi Lagrange Saudi Arabia Warren M P83 4/18 7.*-df

i-

{ ,

Table 2 Sumary of Liquid Gas Ships .

. Total Numhar of Ships -

4 Vigl j Year Propane Butane Butadiene Chloride ING Total 1

i 1977 1 10 10 25 0 46 1

1978 5 10 1 25 0 41

~

1979 1 10 13 0 0 24 1980 2 9 12 0 0 23 1981 3 8 6 0 0 17 1982 1983 1984 1985 1986 1987 1988 1989

, 1990

! 1991 1992 I 1993 1

1994 i 1995 l

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 M P83 4/18 8 *-df l

~ _ . _ _.. , - . . - . . . _ . , _ _ . . . . . . , - . . . _ _ . ~ . , , . , . . . . - . _ , . , _ . . _ , , _ _ . _ _ . . . _ _ . , . . _ - . . -

Table 2 - Contirued Swmary of Liquid Gas Ships

~

• Total Number of Ships Vinyl Year Propane Butane Butadiene Chloride ING Total 2001 2002 2001 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 l 2024 2025 l ~~

2026 2027 2028 2029 2030 M P83 4/18 9.*-df l

l

~

J The focus of our review of construction activity was on the following:

a. Have any LNG or LPG facilities been licensed on the Delaware River?

b. Have any new docking facilities been authorized which might result in an increase in transportation of pro-pane, butane, butadiene, or vinyl chloride?

3. Has the construction of any rammable objects in or near the shipping channel in the 24 mile catchment distance near Hope Creek been proposed or authorized?

An ongoing activity, which has no impact on the nuclear power plant, is the dredging of the river in front of Arti-

ficial Island, on which the plant is located. This activity is sponsored by Public Service Electric and Gas.

3.3 Conclusions Based on our review of "Public Notices" issued by the U.S .

Army Corps of Engineers, responses to inquiry letters sent to the facilities listed in Table 3, and Federal Energy Regulatory Commission notices, we conclude that no new or proposed construction which would have an effect on the cal-culation of the probability of a flammable fuel vapor cloud reaching the Hope Creek Nuclear Generating Station has ap-parently been authorized.

M P83 4/18 10df

l'

• Table 3

, LPG Facilities- located on the Delaware River Atlantic Richfield Company Box 7709 Philadelphia, PA 19101 (215)~339-2632 British Petroleum Oil Company 1 Marcus Hook, PA 19061 (215) 494-3600 British Petroleum Oil Company Paulsboro, NJ 08066 (609) 423-4000 Cities Services Company Box 300 Tulsa, Oklahoma 74102 (918) 586-2211 Getty' Oil Company Delaware City, Delaware 19706 J

Gulf Oil Company Girard Point, Pa. 19145- "

(215) 389-3500 Mantua Chemicals Terminal, Inc.

Crown Point Road Thorofare, New Jersey 08086 (609) 423-5400 j Mobil Oil Company Paulsboro, New-Jersey 08066 (609) 423-1307 i

M P83 4/18 lldf

Table 3 - continued Sun Oil Company-Marcus Hook, Pa. 19061 (215) 447-1244 Texaco Oil Company Eagle. Point, New Jersey _ 08093 (609) 845-8000 t

3 i

i J

l

}

I i

i i

5 i

i i

i n

J t

I

)

i M P83 4/18 12df ,

ew-+ e+-- .~,4.., e.- me--, - n.,y,e w- g , u -

-c-.3m.--,-yw --- rwaya-wr ev ee w +-y -++-e-+n+ --we-me r- eww w'-a-e+-

. FIGURE 1 NUMBER OF SHIPS TRANSPORTING PROPANE 40-35-30-25-W 20 -

Y A

15 -

10 -

5-I.

0, i i i , , i i ,

1975 1980 1985 1990 1995 2000 2005 2010 2015 i YEAR

4 FIGURE 2 NUMBER OF SHIPS' TRANSPORTING BUTANE

'40 35-

\

30-25--

W 20-<

T 3

15 -

10 --

N 5-

.

• i i I i i i 1975 1980 1985 1990 1995 2000 2005 2010 2015 YEAR

~

\

e FIGURE 3 NUMBER OF SHIPS TRANSPORTING BUTADIENE 40 i

35-30-25-k W

20-

[

15 -

10 - 1 l

5-0, , , , i , i i 1975 1980 1985 1990 1995 2000 2005 2010 2015 YEAR

I FIGURE 4 NUMBER OF SHIPS TRANSPORTING VINYL CHLORIDE 40 t 35-30-25- -

4 N

20-

[

15 -

10 -

5 --

0, i , i i i ,

1975 1980 1985 1990 1995 2000 2005 2010 2015 YEAR

FIGURE 5 NUrviBER OF SHIPS TRANSPORTING LNG l l 35-30-25-4 W

20-

[

is 15 -

10 - i 5-i 0- ,  ;  ; i , , i 1975 1980 1985 1990 1995 2000 2005 2010 2015 ,

YEAR I

I-i

FIGURE 6

,as

/

PENNSYLVANIA CEC.D. <

• Amco atriweny Arco htcEIVNC cVLF

• T E R WIN at. REFIN ERY cutr arertviNa CAM 0EN TERWINAL

-GLOUCkSTER A 7 - a CITY smiTesw rrTaottuu suw ott co.-

CHESTER gP/

1%ULSBORO TEXACO EAste towT uswTua ewtuteAts_

/,

.. 8 '" 8 d ' E" O'"

. WILMNGTON e W estL NEW CASTLE 4

/ _

NEW JERSEY

- PENNS NECK 5 i Towen we. sr Q

DEL CITY

'Q 1 SALEM i

cao cawat s/

. A R T,,IF L I.C.I A.L HOPE' CREEK PLANT s

DELAWARE

\ '\

o I. 2 3 k N BAYSIDE nadncat WILES

'\

SMYRNA h DEL. BAY .

LOCATION OF MAJOR RECEIVING TERMINALS a REFINERIES a HOPE CREEK PLANT a TOWER NO.97

~

4 39

4 HOPE CREEK GENERATING STATION

'I MONITORING OF LNG AND LPG SHIPPING AND CONSTRUCTION ACTIVITIES ON THE DELAWARE RIVER

4. BIYEARLY REPORT - 1981

, ( BY R- 2 )

i i

1 i

J i

c Docket Nos. 50-354 50-355 M P83 9/05 1-li January 1983

4.1 CALCULATION OF CONDITIONAL PROBABILITY 4.1.1 Accident Rate Per Ship Mile The number. of accidents occuring per mile traveled by liquid fuel tankers is based on information collected f rom the U.S. Coast Guard Marine Casualty Computer Data and the U.S. Coast Guard accident files. The total number of accidents was divided by the total exposure (average yearly one way trips on the Dela-ware River) as determined from the publication

" Waterborne Commerce of.the United States," to arrive at an accident rate per transit mile in the Delaware Rive r. An analysis of each of these sources of data is presented in the following pages.

U.S. Coast Guard Marine Casuality Computer Data was used to identify all collision incidents on the Dela-ware River over the years 1979-1980 in the following seven categories:

1. a meeting situation

2. a crossing situation

3. an overtaking situation

4. an anchored or moored condition

5. fog

6. docking or undocking operation

7. not otherwise classified Liquefied gas carriers operate in U.S. Coastal Waters under very strict supervision of the U.S. Coast Guard. In the section of river adjacent to the Hope Creek Generating Facility, the tanker will be moving in the channel at all times under Coast Guard es-cort. In particular, in this section of the river, the liquid gas carrier:

. will not be moored

. will not be in area of industrial docks or piers

. will not encounter any area of rock river bottom f

M P83 9/0 5 2-11

. - will not overtake or be overtaken by other ships

. will not meet other ships at bends

. will not meet oncoming ships of a relative speed of greater than 12 knots

. will not itself exceed a speed of 12 knots

. will only transit if visibility is two miles or greater

. will only transit with tug escort

. will be in continuous communications on two radio channels

. will be U.S. Coast Guard supervised Because of the strict operating procedures involving liquefied gas carriers, those accidents which occur in the seven categories listed above were examined, and any which could not occur under the operating conditions mentioned above for liquid gas carriers were excluded. Any incidents which were questionable were included so that the estimate will be conserva-tive, but any accident which did not involve at least one large vessel of over 18 ft. draf t was discarded.

Next, the total number of one way trips in the Dela-ware River (either inbound or outbound)-of large tankers, dry cargo, and passenger ships with a draf t of more than 18 feet was obtained from the " Water-borne Commerce" publication. Each one-way trip con-stitutes a distance of approximately 100 miles, so that the average ship-miles / year is 100 times the total number of one-way trips.

The accident rate per ship mile was calculated by dividing the total number of accidents by the total ship miles. This calculation was carried out for the period under consideration and also for the cumulat-ive accidents since the first estimate was made. The value calulated for a single period is for comparison purposes only; in the calculation of the overall probability, the cumulative value was used since it has more statistical validity because of the larger sample size. Tables 7 and 8 summarize these values.

M P83 9/0 5 3-li

1 4.1.2 HISTORICA_L COLLISION ANALYSIS FY79-FY80 The United States Coast Guard Marine Casualty Computer Data tape was utilized to identify the oc-currence of all collision incidents (the seven cate-gories listed above) on the Delaware River in the yearn FY1979 and FY1980. A total of 42 collision incidents were identified and the U.S. Coast Guard serial number of each incident was recorded. This record of serial number was then utilized to obtain detailed accident reports from the U.S. Coast Guard accident files for each of the 42 incidents. These 42 detailed incident reports were examined and analyzed furthe r.

The 42 events were initially screened with the pur-pose of deleting those accidents only involving rela-tively small vessels that are not representative of large, self-propelled, liquefied gas carriers and could not puncture the LPG gas tanks if they were the striking vessel. These are identified in Table 4 by their assigned Coast Guard case numbers. An attached Appendix summarizes our rationale for their selec-tion, together with our rationale for all other decisions outlined below.

A second screening of the 29 remaining reports had the purpose of identifying those accidents that did not take place on or very near the 100-mile river segment of interest between the entrance of Delaware Bay and Philadelphia. The five incidents described in Table 5 were placed in this category.

The remaining 24 reports, as described in Table 6, involved collisions between ships while at least one was being intentionally moved.

Cases 91724, 92484, 92579, 93285, 94593, 94671, and 03404 involved collisions while one of the vessels was in the process of docking or undocking, or one vessel was moored or anchored. Although a liquefied gas carrier will not be moored, anchared, docked, or undocked in the 24-mile river section of interest, i.e., the catchment distance in which a cargo release might impact the nuclear generating f acility, all of the above cases, with the exception of 93285 and 94671, were conservatively considered collisions while under way and worthy of inclusion in the accident data base.

M P83 9/0 5 4-li

This practice is consistent with previously submitted testimony before the NRC. Case 93285 involvea a motor boat, which is too small to create any daraage when colliding with a liquefied gas carrier, and can therefore be eliminated from the data base. Case 94671 involved a collision in fog, and can be elimi-nated since U.S. Coast Guard regulations do not per-mit movement of liquefied gas carriers in fog conditions.

The remaining cases listed on Table 6 are discussed below. Cases 02782 and 03842 involved passing situations, and were conservatively included in the data base, as was case 05160, where a barge on tow of a tug collided with a vessel wnile being passed and overtaken. Case 05154 was not included, since it involved a collision in a fog situation.

In conclusion, therefore, we conservatively find 8 incidents that should be included in the accident data base: 91724, 92484, 92579, 94593, 03404, 02782, 03842, and 05160. Each is considered relevant and potentially applicable to the analysis of the LNG or LPG tanker spill probability in the 24 mile river segment of interest.

The results of this analysis are summarized in Table 7.

As a reference for the methodology employed above, the following can be consulted:

Supplemental Testimony of Dr. Ashok Kalelkor in 4

response to matters raised by the Atomic Safety and Licensing Appeals Board in ALAB-429, 11 October 1977 Docket Nos. 50-354 50-355 Pages 5-21 M P83 9/05 5-11

,r - , -

--. - -- r ,

. TABLE 4

' INCIDENTS INVOLVING ONLY SMALL VESSELS BY USCG CASE NUMBER

• 91586 01140 91587 02415 91703 02920 92584 05155 92950 05161 93569 94599 94602 t

• These incidents only involved tugs, towboats, non-self i propelled-barges, or other relatively small vessels.

1 M P83 9/05 6-li 9

i

- - - , -- , - , .- ----,-+,,,,--.,,,--.--..n----, - - < - , ~ - , - , - , , , , -

1 i

TABLE 5 INCIDENTS WITH INCORRECT LOCATION Case Number Location 92584 Middle Thorofare Bridge, Wildwood, New Jersey 92850 Middle Thorofare Bridge, Wildwood, New Jersey 93170 Christina River, Delaware.

94603 Cold Springs Fish House, Cape May, New Jersey

05144 He terford Inlet Lt. , New Jersey 4

4 M P83 9/05 7-li i

~

O TABLE 6 COLLISIONS BETWEEN LARGE VESSELS Case No. Description 91724 Tug collided with ship, while docking.

92484 Ship collided with container crane, while docking.

92579 ship collided with container crane, while undocking.

93285 Motor boat collided with moored fishing vessel.

94593 Ship collided with anchored ship.

94671 ship underway lost power and collided with anchored ship, in fog.

02782 Dredge collided with ship which was passing her.

03404 Barge in tow collided with anchored ship.

03842 Barge in tow collided with ship, while being passed in an over-taking situation.

05154 Two tanker ships collided in fog, at Hog Island terminal.

05160 Barge in tow of tug collided with vessel.

a M P83 9/0 5 8-li

TABLE 7 ACCIDENT RATE PER SHIP MILE Ctmalative ative Year Ac d nts dhn [ bb ps e Ctmi{f[

ka 1969-1975 10 10 66321 66321 1.5x10-6 1.5x10-6 1976-1978 7 17 28344 94665 2.5x10 6 1.8x10-6 1979-1980 8 25 14498 109163 5.5x10-6 2.3x10-6 1981-1982 1983-1984 1985-1986 1987-1988 1989-1990 1991-1992 _

1993-1994 1995-1996 1997-1998 1999-2000 2001-2002 2003-2004 2005-2006 2007-2008 2009-2010 2011-2012 2013-2014 2015-2016 _.

2017-2018 2019-2020 _ , _ _

2021-2022 2023-2024 2025-2026 2027-2028 __

2029-2030 M P83 9/05 9.*-11

TABLE 8 DELAWARE RIVER ONE-WAY TRAPPIC IN TANKERS, DRY CARGO, AND PASSENGER SHIPS OF GREATER THAN 18 FT DRAFT (Source: " Waterborne Conmerce")

Year: 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 One-Way Trips: 9744 10151 9258 9553 9858 9086 8671 9559 9300 9485 7789 6709 109163 Average One-Way Trips = 12 = 9100 Each one-way trip representing a distance of about 100 miles, the average ship-miles / year is 100 times the average number of one-way trips. Therefore, this figure is 9.1x105, 4.1.3 Collisions With Fixed Objects The calculation of collision rates with fixed objects (e.g. Tower 97, which is approximately 9 miles up-river from the Hope Creek site - see Figure 6), was handled in the same manner as the accident rate.From the Coast Guard accident casualty data for the previ-ous years, the number of occurrences of rammings of fixed objects involving a ship of over 18 feet draf t was determined. As in the previous analysis, any accidents which could not have happened to a ship foll7 wing U.S. Coast Guard regulations for liquefied gas carriers were eliminated from the data base. The number of one-way transits per year of ships of over 18 feet draf t is the same as was used in the accident rate calculation.

The number of collisions with fixed objects is based on approximately 50 rammable objects

• which large ships could have struck; Tower 97 represents one of them. Thus, to calculate the probability of a colli-sion with Tower 97 per transit, the total number of collisions in the river is divided by 50 and then divided by the total number of one-way transits.

• As counted from NOAA Navigational Charts for the Delaware River. None of the 50 objects, except Tower 97 occur in the 24 miles of interest.

M P83 9/05 10-li

Table 9 sum.narizes the 13 remaining . incidents that involved'large ship collisions with fixed objects.

Case 92596 involved a collision with a jetty in a., fog' situation, and therefore is not included in the data base. Cases 93174, 94601, and 94606 involved move-ment and subsequent contact of a moored ship with.

nearby objects ~due to waves caused by a passing ves-sel. Since a liquefied gas carrier would not be moored or anchored in the mile river _ segment of concern, this sort of accident is wholly unlikely.

For similar reasons, case 02418 involving a ship that broke anchor and collided with the loading pipeline is not included in the data base.

The rest of the cases are considered. applicable.

In conclusion, an examination' of all 13 collisions involving vessels.and fixed objects in the Delaware River (FY79-FY8 0) reveals that, at most, 8 of them are relevant and potentially applicable to the analy-sis of LNG or LPG tanker spill probability.

The results of this analysis are summarzied in Table-10.

M P83 9/05 11-11

TABLE 9 SHIP COLLISIONS WITH FIXED OBJECTS Case Number Object Circumstances 91788 pier docking pilot error 92596 jetty in fog, trying to avoid anchored ship 93174 gangway Moored ship surged due to wake of passing ships, and struck gangway 94601 pier Moored ship surged due to wake of passing ship, and struck pier.

94606 pier Moored ship surged and hit pier, due to nigh winds 02412 pier While docking, ship struck pier 02418 pipeline Moored ship broke anchor and damaged loading pipeline 02709 pier Undocking 03630 pier docking 03827 pier docking 04975 pier docking 05148 pier docking 05150 pier docking M P8 3 9/0 5 12-li

TABLE 10 COLLISIOtC WITH TOWER 97 PER TRANSIT Cumulative Cumulative Total Cumulative One-Way One-Way Collision Collision Year Collision Collisions Trips- Trips Rate Rate 1969-1975 8 8 66321 66321 2.4x10-b 2.4x10-6 1976-1978 11 19 28344 94665 7.8 x10-6 4x10-6 1979-1980 8 27 14498 109163 1.1x10-5 5x10-6 1981-1982 1983-1984 1985-1986 1987-1388 1989-1990 1991-1992 1993-1994 1995-1996 1997-1998 1999-2000 2001-2002 2003-2004 2005-2006 2007-2008 2009-2010 2011-2012 2013-2014 2015-2016 2017-2018 2019-2020 l 2021-2022 2023-2024 l

2025-2026 2027-2028 2029-2030 M P83 9/05 13.*-li

~

4.1.4 Spills Per Accident The probability that a spil) will occur given that there has been an accident was calculated based on a complex energy conservation model developed by V. U.

Minorsky.** The model estimated the depth of pene-tration of the hull of a ship based on the speeds of the colliding ships and the angle of collision.

Conservative estimates of the spill per accident rate a re 0.1 for propane, butane, butadiene and liquefied natural gas carriers,* and 0.01 for vinyl chloride carriers. Vinyl chloride has a much smaller value because it is carried in self contained tanks near the centar of the ship, at a larger distance f rom the hull.

Table 13 presents the probabi3ity values for spills per accident associated with nach type of liquefied gas.

The probability that a storane tar.k on a gas carrier would be ruptured given t hac an accident has occurred was calculated based on a method developed by Vladimir U. Minorsky.** In order to determine if there have been any new developments in this field, PEE &G has contacted Mr. Minorsky et George S. Sharp, Inc., N. Y. Since Mr. Minorsky ha>l rctired the previous year, we discussed this item with an associate of Mr. Minorsky, Mr. Chetank Yang.

In response to our inquiry, Mr. Yang indicated that there have been no new developments in the area of ship damage models which would lead to changes in the probability figures presented in Table 11.

• These values were used in the ASLAB decision al-though the applicant's more rigorous estimate for LNG tankers indicates a much smaller conditional spill probability.

• • V. U. Minorsky " Analysis of Ship Collisions with Reference to Protection of Nuclear Power Plants,"

Journal of Ship Research October 1953.

M P83 9/05 14-li

TABLE 11 SPILLS PER ACCIDENT Year Propane Butane Butadiene C obde IMI 1978 0.1 0.1 0.1 0.01 0.1 1980 0.1 0.1 0.1 0.01 0.1 1982 1984 1986 1988 1990 1992 1994 1996 i

1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 M P83 9/05 15.*-li

=

4.1.5 Vapor Cloud Formation Per Spill Given that there has been a spill of a liquefied gas cargo, a vapor cloud will form onlysif the. fuel is not ignited at the source of the spill.. In a' crash situation which is large enough to release the' cargo, it is expected. that there will be ample sources of

-ignition from severed wires, frictionally heated.

-metal, and-associated sparks. Although a vapor cloud could be expected in less'than one percent of all. ,

spills, a conservative value of 10 percent has been used.

- Table 1:2 presents the' probability values associated

~

.with various liquefied gases.

In order to verify if any new developments in the area of vapor. cloud formation and dispersion have occurred,-PSE&G has contacted the fo.~ lowing sources involved in this type of research:

a. Gas Research Institute Mr. Sami Atallah - Manager, Systems Safety Research

b. Safety.and Engineering Technology Section Arthur D..Little, Inc.

Dr. Elizabeth M. Drake ,

c. Massachusetts Institute of Technology Professor James A. Fay In their response to our inquiry, all three sources of'information indicated that there were no'new de-velopments in this field which would lead-to changes in the probability values presented in Table 12.

I i

i i

M P83 9/05 16-li

TABLE 12 VAPOR CIOUD FORMATION PER SPIII Vinyl Year Propane Butane Butadiene Chloride ING 1978 0.1 0.1 0.1 0.1 0.1 1980 0.1 0.1 0.1 0.1 0.1 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 _

2024 2026 2028 2030 M P83 9/05 17.*-11

~.

4.1.6. Meteorological Factor The meteorological factor is based on distance between the spill and facility, and the mean wind speed and direction. It represents the probability that a vapor cloud formed at a particular location will reach the nuclear plant with a flammable mix-ture of fuel to air. It is not expected to change under ordinary circumstances over the lifetime of the facility and the values which were used are given in Table 13 for the entire 24 mile catchment distance and for Tower 97.

Table 13 - Meteorological Factors

• 24 Mile Ca tc hmen t Tower 97**

Propane 0.31 0.002 Butane 0.32 0.002 Butadiene 0.25 0.002 Vinyl Chloride 0.24 0.002 LNG 0.354 0.002

• For details on how the Meteorological Factor is derived, see answer to Question 3 of Exhibit 10 on Docket Nos. 50-354 and 50-355 dated January 13, 1975.

• • The value of 0.002 is very conservative since it assumes a 10,000 ton spill. Such spill sizes are not considered credible for propane, butane and butadiene since individual tank sizes for such cargoes are considerably smalle r.

4.1.7 Conditional Probability Calculation The probability that a flammable vapor cloud will reach the nuclear facility was calculated for eaca fuel type under consideration (propane, butane, butadiene, vinyl chloride, and LNG). For each fuel type, two separate calculations were made; the first was based on the likelihood of a collision with another ship anywhere within the 24 mile catchmant distance and the second was based on a ramming of Tower 97 nine miles upstream of the nuclear plant.

M P83 9/0 5 18-11

The overall probability of a flammable vapor cloud reaching the nuclear plant is the sum of these eight terms.*

The individual terms were calculated as the product of five other factors. These factors are:

1. Total number of ships per year

2. Accidents per mile; or accidents per passage

3. Spills per accident

4. Vapor cloud per spill

5. Meteorological factor The first of these five factors is determined annually for each fuel type as listed in Table 1.

The remaining four f actors have been conservatively estimated and are described in Sections 4.~1.4, 4.1.5, and 4.1.6. These factors may vary somewhat as a re-sult of a biyearly review of shipping experience, and the method for their calculation is also described in Sections 4.1.4, 4.1.5 and 4.1.6.

The results of all these calculations are presented in Tables 14 and 15, for both ship to ship collisions and collisions with fixed objects.

4.2 Ccnclusions The results provided in Tables 14 and 15 show that the total conditional probability of a flammable vapor cloud reaching the Hope Creek plant is 1.5x10-7 for 1980; and 1.2x10-7 f or 1981. Both values are almost eight times less than the 10-6 value estab-lished as an upper limit by the NRC, based on conservative approximations.

All approximations on this study have been made in a .

conserv3tive manner 17the resultant probabilities of 1.5x10 and 1.2x10 being, therefore, conservative results.

• Since LNG is not shipped on the Delaware, only four cargoes (propane, butane, butadiene, and vinyl chlor-ide) and two spill modes (ship-ship collision and collision with Tower 97) are relevant, leading-to eight probability terms to be added for the total probability.

I M P83 9/05 19-11 n---,, -.---r - - - -

_ ,. - - - - , , - - - ,e ,- - , -

The actual LNG and LPG traffic has decreased from 1979 to 1980 and even more to 1981; no additional rammable objects, mooring or docking sites, or any other facility that might cause a significant change in the probability of a flammable vapor. cloud reach-ing the plant have been built or are planned.for con-struction within the 24 mile catchment distance en the Delaware River; and no proposals for construction of LNG terminals along-the Delaware River have been received by FERC.

Based on the above information, we can conclude that the probability of a flammable vapor cloud reaching the nuclear facility is sufficiently small such that the associated hazards to the health and safety of the public are negligible.-

i M P83 9/05 20-11

TABLE 14

-PROBABILITY OF FLAMMABLE VAPOR CLOUD REACHING NUCLEAR FACILITY (Based on 1980 Data)

No. of Ships Accidents Ranmings Spills Per Vapor Cloud Meteorological Per Year Per Mile Per Pascage Accident Per Spill Factor Probability ING 0 2.3x10-6 0.1 0.1 0.354 0 LPG Propane 2 2.3x10-6 0.1 0 .1 0.31 1.4x10-8 Butane 9 2.3x10-6 0 .1 0 .1 0.32 6.62x10-8

-8 Butadiene 12 2.3x10~ 0.1 0 .1 0.25 ~ 6.9x10

Vinyl Chloride 0 2.3x10-6 0.01 0.1 0.24 0 Tbwer 97

-6 LNG 0 5x10 0.1 0.1 0.002 0 LPG 23 5x10-6 0.1 0.1 0.002 2.3x10-9

-6 Vinyl Chloride 0 5x10 0.01 0.1 0.002 0 Total 1.5x10-7 4

a i

M P83 9/05 21.*-li

9 4

I TABLE 15 PROBABILITY OF FLMMABLE VAPOR CILUD REACHING NUCLEAR FACILITY (Based on 1981 Data)

No. of Ships Accidents Ranmings Spills Per Vapor Cloud Meteorological Per Year Per Mile Per Passage Accident Per Spill Factor Probability

ING 0 2.3x10-6 0.1 0 .1 0.354 0 LPG Propane 3 2.3x10-6 0 .1 0 .1 0.31 -2.2x10-8 Butane 8 2.3x10

-6 0.1 0 .1 0.32 6.x10-8 Butadiene 6 2.3x10-6 0.1 0.1 0.25 3.45x10-8

-6 Vinyl Chloride 0 2.3x10 0.01 0 .1 0.24 0 Tower 97

-6 LNG 0 5x10 0.1 0.1 0.002 0

~

LPG 17 5x10-6 0.1 0.1 0.002 1.7x10-9

-6 Vinyl Chloride 0 5x10 0.01 0.1 0.002 0 1

j Total 1.2x10-7 i

j M P83 9/05 22-li i

I

APPENDIX USCG CASE REVIEWS INTRODUCTION The following sections, denoted.by U.S. Coast Guard Case numbers, discuss our interpretation of each of the 42 reported incidents during the period FY 79-FY 80. Each begins with a transcription of notes taken during the reading of the detailed accident report . files of the U .S.

Coast Guard. Generally, this is followed by supplemental data obtained from a Coast Guard-supplied computerized summary of accidents, and the rationale for our decision to include or not include the incident in the accident data base.

91586 Tow tug collided with barge , due to inattention of tug operator. At C&O Oil Pier, James River, Newport Neus, Virginia. Should not be counted, since it involves only small vessels, and did not even occur in the Delaware River.

91587 Tow tug lost power and collided with bulkhead between Pier 11 and Pier 12 North, Delaware River. Should not be counted, since it involves only a small vessel.

91703 Coal barge collided with Pier 11 South, at Port Richmond, Philadelphia. Should not be counted , since it it involves only a small vessel.

91724 Tow tug collided with ship, while docking at Mobil Oil Co.

Pier, Paulsboro, New Jersey. Although it happened while docking, this case was conservatively included in the data base.

91788 Ship collided with pier, while docking. At Pier 122, Delaware River. Although it happened while docking, this case was conservatively included in data base.

i M P83 9/0S 23-mw

6 92484 t

Ship collided with container crane, while undocking at Berth 5, Packer Ave. Marine Terminal, Philadelphia. Conserva-tively included in data base.

1 92579 Ship collided with container crane, while undocking at Packer Ave., Marine Terminal, Philadelphia. Conservatively included in data base.

92584 Barge in tow by tug collided with anchored dredge barge, in Schuylkill River. Should not be counted, since it involves only small vessels. Also, it did not occur in the Delaware River.

92594 -

Fishing vessel collided with Middle Thorofare Bridge, Wildwood, New Jersey. Should not be counted , since it did not happen in the Delaware-River.

92596 Passenger vessel collided with Roosvelt Inlet Jetty at Lews, Delaware. In fog, trying to avoid an anchored pleasure boat. Should not be counted, since it occurred in a fog situation, when liquefied gas carries are prohibited from traveling.

92850 Fishing vossel collided with Middle. Thorof are Bridge, Wildwood, New Jersey. Should not be counted, since it did not happen in the Delaware River.

92950 Barge in tow by tug collided with dike surrounding Reedy Island Dike South Light, Delaware River. Should not be counted, since it involved only one small vessel.

M P83 9/05 24-mw <

t l

F O

93170 Ship collided with walkway at Wilmington Marine Terminal, Christina River. Should not be counted , since it did not occur in the Delaware River.

93285 Motor boat collided with moored fishing vessel, at Bowers Beach, Delaware. Should not be counted , since it involved a motor boat, which is too small to create any damage when colliding with a liquefied gas carrier.

93569 Barge in tow of tug collided with anchored motor boat, in the New daven Harbor Entrance Channel, on the Long Island Sound. Should not be counted, since it involves only small vesse?s. It also occurred outside the Delaware River.

94593 Ship collided with anchored ship, at Big S tone Anchorage, Delaware Bay. Due to mechanical failure. Although it involved an anchored ship, it was conservatively included in data base.

94999 Tug boat collided with bridge and pier at Hess Oil Docks, Delair, New Jersey. Should not be counted since it involved only one small vessel.

94601 Moored ship was damaged by surging in wake of a passing ship. At Arco, Fort Mifflin Dock, Delaware River. S hould not be counted, since a liquefied gas carrier will not be moored or anchored in the 24 mile river segment of concern.

I 94602 Barge in tow by tug collided with the Tacony Palmira Bridge, Delaware River. Should not be counted, since it involved only one small vessel.

94603 Moored fishing vessel was damaged by drif twood, at Cold Fish House in Cape May, New Jersey. Should not be counted, since it did not occur in the Delaware River.

M P83 9/05 25-mw

t 94606 Moored' ship collided with Pier 1 at Getty Oil Terminal,

. Delaware City. Due to high winds. _Should not be counted, since liquefied gas carriers will not be moored in the river segmer.t of interest.

94671 Ship underway lost power and collided with anchored ship, near Brandywine Light, Delaware Bay, 9 miles northwest of 1

Cape May, New Jersey. It happened at night, in fog conditions. Shr uld not be counted, since it happened in a fog situation.

I 01140 Barge broke anchor due to nigh winds, and collided with dredge. At Marcus llock Anchoraje area, Delawcre River.

Should not be counted, since it involved only small vessels.

02412 Ship, while docking, collided with Pier-E, Port Richmond, Philadelphia. Conservatively included in da ta baso .

02415 Fishing boat- collided with sail boat, east of Cape May Inlet, Cape May, New Jersey. Should not be included since

it involved cnly small vessels.

02418 Moored ship broke anchor, drifted into channel, and damaged oil loading equipment. At Mobil Oil Terminal, Paulsboro, New Jersey. Should not be counted, since a liquefied gas carrier will not be moored on the' river segment of interest.

02709 While undocking, ship collided with dock at Gulf Oil Co. ,

Hog Island, Philadelphia. Conservatively included in data base.

i M P83 9/05 26-mw

-5 -

02782 Dredge . barge collided with ' ship while 'being passed, due to l ,

l suction effect of ship wake. At New Castle Range, Delaware i River. Although a liquefied gas carrier will not be-

! involved in a passing situation, this case was

conservatively included in - the data base.

02920 Moored barge was damaged while surging due to wake of a ship -

r which was docking. S hould not be ' included, since it

_ involved only' a small vessel. At Gulf Oil Terminal, Hog 7

Island, Philadelphia.

03404 Barge in tow by tug collided with anchored ship, at Big Stone Anchorage . Area of Lower Delaware Bay. Although i involving an anchored ship, this case was conservatively.

included in the data base.

I. 03630 i '

Ship, while docking, collided with. dock at-Arco Terminal, l Fort Mifflin, Delaware River. Conservatively included in i

j data base.

03827

{~

Ship under tow by tug collided with pier while docking, and grounded at Northern Metals Terminal, Philadelphia.

( Conservatively included in data base.

t l- 03842

{

Barge in tow by tug collided with ship, while being passed i

and overtaken. At Liston Range, Delaware River. -

Conservatively included in data base , as with-case 02782.

p

! 04975

! While docking,- ship collided'with Pier, at Pier 14, Port

! . Richmond Terminal, Philadelphia. Conservatively included in data base.

I f M P83 9/05 27-mw t

- . -,-.2.. .., - - -- ,- w .. - ..-,,-. - . . - - - , . , , , . - , - . _ . . . . - . . - . .,,,.,,. - -. ,.-

7.

e.

p 05144 Passenger vessel collided with moored- fishing vessel, near

'Heterford' Inlet Light, New Jersey. Should-not be counted,

.since collision occurred outside the Delaware River.

-05148 Tankship in tow by tug - collided with pier while docking.

Conservatively included in data base.

05150 Freight vessel collided with pier, while dccii ng.

Conservatively included in data base .

05154 Two tankers collided in fog , due to pilot error. At Hog Is?and Terminal, Delaware River. Should not be counted, since it occurred in a fog situation.1 05155 Barge suffered damage due to unknown cause. Should not be counted, since it involves only one small vessel.

05160 Barge in tow by tug collided with vessel, due to mechanical failure of vessel. Included in data base.

05161 Barge in tow by tug collided with anchored barge, in a fog condition. Should not be counted, since it involved only small vessels and occurred in a fog situation.

M P83 9/05 28-mw