ML20009E343

Transcript of Rc Iotti Testimony on Behalf of Util on Bishop Contention 6 & ASLB Question 12 Re Shell Oil Co 6-inch Liquid Petroleum Gas Pipeline

ML20009E343
Person / Time
Site:	Allens Creek File:Houston Lighting and Power Company icon.png
Issue date:	07/20/1981
From:	Iotti R EBASCO SERVICES, INC., HOUSTON LIGHTING & POWER CO.
To:
References
NUDOCS 8107280062
Download: ML20009E343 (12)

Text

._.

July 20, 1981

'R E C E l V E, L' 1

@N v.

2 Jul 21 E t UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION  ; j 3 cr7Ici: 0? A'riICATIO]J s 5 n"20R;3 sBEFORE TE ATOMIC SAFETY AND LICENSING BOARD 4

5 In the Matter of 5 #,

5

• HOUSTON LIGHTING & POWER COMPANY $ Docket No. 50-4 6 g ,

~

(Allens Creek Nuclear Generating 6 p JA 7 Station, Unit No. 1) 5 N[ {

f 6) 6 V l

V/

0 "

TESTIMONY OF ROBERT C. IOTTI, ON BEHALF 0 JUL ~o q \SS\1 T

~

9 HOUSTON LIGHTING & POWER COMPANY ON BISHO #' 3 CONTENTION 6 And BOARD QUESTION 12 \ /> u.5 M [x 10 RELATING TO THE SHELL OIL CO. W -7  ;.

6 INCH LIQUID PETROLEUM GAS PIPELINE \\

11 12 Q. Please state your name and business address and 13 describe your educational and professional experience.

p A. My name is Robert C. Iotti and my business address is Ebasco Services, Inc., 2 World Trade Center, New 15 York, N. Y. I have previously discussed my position and 16 l background in connection with my testimony on Doherty l

17 Contention 47.

18 Q. Dr. Iotti, what is the purpose of your testimony?

19 A. The purpose of my testimony is to address Intervenor 20 Bishop's Contention 6 which alleges that:

21 The rupture of the six-inch liquid petroleum 22 gas pipeline could cause a cloud of explosive gas to travel along depressions to the area of 23 the plant before exploding with such force to gol damage the safety equipment at the plant and the W, 1 l I 8107280062 810720

! PDR ADOCK 05000466 l T PDR

1 2 workers at the plant. For this reason either the pipeline or the plant must be moved.

3 My testimony will also address Board Question 12, 4

which asks the following:

5 Has it been definitely established whether (the LPG] pipeline might carry potentially more 6 dangerous materials such that, following a pipeline rupture, safe shutdown of the plant 7 would be precluded?

8 Q. Dr. Iotti, briefly describe how far the six-inch 9 liquid petroleum gas pipeline is located from the nearest 10 ACNGS Category I structure?

11 A. At its closest point, the line is located approximately 12 8000 feet northeast of the nearest ACNGS Category I structure.

13 Q. What measures has Shell oil Company taken to detect 4

a possible leak in this pipeline?

A. Shell Oil Company is providing a leak detection system capable of detecting a leak of approximately 20 17 bbls/hr in 3 minutes after the leak occurs. Isolation 18 valves in the line will be closed and the pump tripped 19 by the operator after information is received that a leak 20 has occurred. Completion of this isolation can be accom-21 plished in 5 minutes following leak detection. The use 22 of this leak detection system coupled with the ability to 23 isolate makes it extremely unlikely that a rupture occurring 24 e . .

1 2 at the nearest point of ACNGS Category I structures could 3 be accompanied by a release of a quantity of LP gas large 4

enough to present a hazard to the plant safety related structures in the event of a detonation or deflagration.

5 Q. Dr. Iotti, although you state that it is extremely unlikely that a detonation of significant quantities of propane could occur from a rupture of the six-inch liquid 8

petroleum gas pipeline, has the Applicant nonetheless 9

performed an analysis of potential effects of a rupture 10 of the line and the potential effects of detonacion/

11 deflagration of the escaping LP gas?

12 A. Yes. Appendix 2.2-A of the PSAR contains analyses of 13 the potential effects of a rupture of the liquified 14 petroleum gas line and subsequent detonation and/or 15 deflagration of the escaping gas. These analyses utilize 16 several conservative assumptions and demonstrate conclu-l l 17 sively that.the plant's Category I structures will not i

18 suffer any adverse consequences from a line rupture and 19 subsequent detonation and/or deflagration of the emitted 20 LP gas.

21 Q. W uld you briefly describe the gaseous transportation l mechanisms for which these analyses are performed?

22 A. Separate analyses have been performed for each of 23 the four modes in which a flammable cloud of LPG could be 24 1

2 generated and transported toward the plant. The four 3 modes described briefly are:

4

1. Atmospheric dispersion toward the plant of the LP gas released from the postulated break.

5

2. Slumping of the heavier than air cloud 6

formed by the mixing of the LP gas and the surrounding air at the postulated break location accompanied by 8

further mixing with air at the cloud-to-air interface 9

resulting in a fog-like cloud travelling toward the 10 plant along the depression of Allens Creek.

11

3. Initial gravity flow of the LP gas / air 12 cloud followed by atmospheric dispersion and vice 13 versa.

14 4. Gravity flow of nearly pure LP gas / liquid 15 with little mixing with the surrounding air. This 16 results in essentially a pure LP gas stream in the 17 Allens Creek depression, which is postulated to heat 19 up, mix with air, and then be atmospherically dispersed 19 toward the plant.

20 Q. What are the significant assumptions used in each 21 LPG pipeline rupture analysis?

A. Those assumptions which have the greatest impact on 22 the calculations performed are:

23 24 l

. . j l

l 1 l 2 1. The rate of flow of LP gas out of the 3

break, which influences the size of the cloud the.t 4

can be created, is conservatively computed by assuming that a complete severance in the line occurs at the 3

Allens Creek Crossing. (This assumption applies to the analyses performed for the modes numbered 2, 3 and 4 above).

8

2. Double ended rupture of the line occurs 9

instantaneously and at the closest point to plant 10 Category 1 structures (8000 feet). (This assumption 11 applies to the analyses performed for mode number 12 1.)

13 3. The temperature of the atmosphere is 14 assumed to be 72*F. Higher temperatures would lead 15 to higher vaporization of escaping propane, but the 16 flow rate would be less due to higher quality at the 17 exit plane.

18 4. The LP gas is assumed to be propane because 19 of its chemical and physical properties which render 20 it a conservative choice for reasons explained later in my testimony.

21

5. Assumed meteorological conditions were 22 characterized by a Pasquill F inversion with wind speed of 0.8 mps in the direction vi the plant

-s r

1 2 structures. Such conditions occur at the ACNGS site 3 only S% of the time. (This assumption applies to 4 the analyses performed for the modes nutbered 1 and 5 3 above. )

6. A leak detection system is available which 6

is capable of detecting a leak of approximately 20 7

bbls/hr (or about 5% of operating flow) in 3 minutes after the leak occurs. After detection, line shutdown 9

and isolation will be taken in 5 minutes.

10 The break is conservatively calculated to 11 initially discharge 568 lbs/sec for 480 secor.ds.

12 Thereafter due to pipeline operator action in 13 stopping the pipeline pumps and closing valves, flow 14 is predicted at 70 lbs/sec for 515 seconds, then 15 dropping to 30 lbs/sec for another 2750 seconds, at 16 which time flow ceases.

17 To estimate the maximum credible si;a ;f an 18 atmospherically dispersed cloud, an average constant 19 conservative flow of 100 lbs/sec of propane is 20 assumed to issue continuously from the postulated 21 break.

22 Q. F r each analysis that was performed, what were the al ulated effects on the Allens Creek plant?

23 24

1 2 A. For the anrlysis concerned with the previously-3 described fin mode of gaseoun transportation, the calculated effects were:

4 tue assumption of stable atmospheric conditions 5

with low wind speeds led to the prediction of larger detonable /deflagrable clouds than would be predicted if realistic atmospheric conditions were assumed.

8 The calculated maximum size of the detonable cloud 7 ft3 , while the has been determined to be 1.55 x 10 10 maximum deflagrable size has been determined to be 1.96 x 10 ft 3, 12 After detonation, the analysis shows that a 13 peak overpressure of only 0.6 psi would impact on 14 the nearest Category I plant structure. Since all 15 Category I plant structures are designed to withstand f 1

16 a peak overpressure of 2.3 psi, no damage would I occur to the ACNGS Category I structures. l 17 18 For the analysis concerned with the previously-19 described second mode of gaseous transportation, the 20 calculated effects were:

For the analysis of the gravity ficw along the 21 Allens Creek depression with air mixing at the 22 propane / air cloud and air interface, the analysis demonstrates that clouds of flammable concentration 24 1

2 will not extend further than 6,000 feet down the 3 Allens Creek channel, a location which is approximately 4

1400 feet from the nearest Category I plant structure.

Assuming further that the entire cloud is a 5

homogeneous flammable mixture of propane and air, 6

its detonat cn has shown to produce a peak overpressure of only 1.0 psi. Again this detonation would pose 8

no safe'" hazard to the plant.

9 For the remaining two analysis concerned with 10 the previously described third and fourth modes of trans-11 portation the calculated effects were:

12 Detonation of these maximum sized clouds pose 13 no safet'i hazard to the plant. Peak overpressures of 14 less then .6 psi and 1.0 psi have been calculated to 15 result frcm analyses concerned with the third and f.ourth 16 previously described modes respectively.

17 Q. Does the PSAR analysis address the hazards resulting 18 from cloud deflagrations instead of detonations?

19 A. Yes, it does. Using the same analyses that were 20 applied to tne detonable clouds, the conclusion reached 1

is that deflagration would also pose no hazard to the 21 plant Category I structures. l 22 1

Q. Did the PSAR analysis also look at the effect of I 23 missiles generated by the assumed detone n?

l l

1 l

~~

l l

( ..

1 2 A. Yes it did. The hazards presented by missiles have 3 been assessed by computing the total mass of debris observed to be generated by surface explosions of equivalent 4

am unts of TNT for the largest of the detonations postulated 5

by the four analyses previously described. Then the 6

total mass of debris that could strike a plant safety related structure was computed using the assumption that all of this mass would be concentrated in one missile 9

travelling at the maximum air particle velocity.

10 This was done since the open area between the 11 location of the hypothetical clouds and the plant is 12 devoid of structures or objects capable of becoming 13 severe missiles. Under these assumptions the energy of 14 the missile has been computed to be insufficient to cause 15 any damage to any plan + safety-related structure.

16 Q. In the preceding analyses you have stated that the 17 postulated rupture in the line is a complete severance of 18 the line. Could smaller breaks occur and result in 19 hazards to the plant which are possibly more severe than 20 those caused by the very large break?

21 A. No, they could not. Smaller breaks have been examined and, under the same conservative assumptions made in the 22 analyses f the large breaks, result in lower volumetric 23 flow rates and ultimately in smaller, lesser-concentrated

1 2 clouds which, if detonated, would produce lower overpres-3 sures than those resulting from larger breaks.

4 Q. In reference to Board Question 12, propane has been P

5 used as the representative LP gas for performing the analysis. What is the reasoning behind this choice and

_ could the LFG pipeline carry potentially more dangerous i

materials?

8 A. Propane gas was chosen because it represents the 9

most conservative approach in the analysis for the following 10 reasons:

11

1. The LP gas line currently carries batches 12 of gasoline, isopentane, normal butane and isobutane, 13 plus distillate. The only other hazardous products 14 which could reasonably be assumed to be transported 15 in the line are propane and possibly other liquid 16 hydrocarbons. Pound for pound propane has effectively 17 the same TNT mass equivalency of other flammable 18 hydrocarbons such es butane, isobutane, ethylene, 19 butadiene, propylene, etc. If 20 21 If T. V. Eichler, E. S. Napadensky, " Accidental Vapor Phase Explosions on Transportation Routes Near Nuclear Pl. ants,"

NUREG/CR-0075, April 1977.

22 l

23 24 w.

I l

r

1 2 . 2. The flammable and detonable limits of 3 propane are wider than those of the higher hydrocar-4 bons, hence proportionately more propane can be 5 contained in deflagrating or detonating clouds than ther higher hydrocarbons. Thus for the same flow 6

of LP gas out of the break, assuming propane as the 7

LP gas maximizes the quantity of material which can detonate or deflagrate and hence maximizes the 9

hazards to the plant.

10 i 3. For the same line pressure and temperature, 11 the specific volume of the higher hydrocarbons is

• 2 comparable or higher than that of propane, so that 13 the same or more pounds of propane would be contained j 14 in any given segment of the line as other hydrocarbons.

15 Thus again a larger quantity of flammable material l

16 can be postulated to escape from a hypothetical i

17 break in the line by assuming that the line carries 18 propane rather than other higher hydrocarbons, 19 leading again to maximizing the hazards to the 20 plant.

21 4. The initial flow rates out of a break are 22 comparable for propanes and butanes, but would be s1 wer f r the other heavier than air hydrocarbons.

! 23 l Hence, again the assumption of propane leads to 24 maximizing the hazards to the plant. '

i

1 2 5. The amount of energy needed to cause a 3 detonation of a mixture of LP gas and air is lower 4 for propane-air mixtures than for mixtures of air 5

and ther higher hydrocarbons. Assuming that the g LP gas is pure propane increases the likelihood that there will be sources of sufficient energy to cause the hypothetical vapor cloud to ignite or detonate. The only known case of an open air 9

detonation of a vapor cloud involved propane.

10

6. There are several instances of unconfirmed 11 propane-air cloud deflagrations and at least one 12 detonation, against which it was possible to compare 13 results of the analyses.

14 In summary, propane was chosen since it would 15 present the greatest hazard to the plant.

16 Q. What are your conclusions?

17 A. Analyses have been performed, utilizing conservative 18 assumptions, that adequately demonstrate the plant 10 Category I structures at Allens Creek will not suffer 20 adverse consequences due to the postulated LP line 21 rupture and subsequent detonation and/or deflagration of 22 the emitted LP gas.

23 24

_ . . _ . . _ _ _ _ . . . . _ _ _ _ _ _ _ _ . . . _ _ _ . _ . . . . - _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ . _ _ _ . _ _ . . _ . . _ _ _