ML20246D994
| ML20246D994 | |
| Person / Time | |
|---|---|
| Site: | Farley  |
| Issue date: | 05/10/1989 |
| From: | NRC |
| To: | |
| References | |
| NUDOCS 8908280360 | |
| Download: ML20246D994 (81) | |
Text
. _ _
)
...e f*
1 g
a UNITED STATES NUCLEAR REGUIATORY COMMISSION k
ALABAMA POWER COMPANY (APC)
(Hydrogen Gas Intrusion Event at Farley Unit 1)
NRC Building Room 16-B-11 One White Flint North 11555 Rockville Pike Rockville, Maryland Wednesday, May 10, 1989 12:00 p.m.
The above meeting took place, pursuant to notice, when were in attendance, the following:
CHAIRMAN:
JAMES LIEBERMAN, DIRECTOR Office of Enforcement, NRC FROM THE NRC:
I J. TAYLOR J. PARTLOW L. CHANDLER H. WONG F. MERSCHOFF S. VARGA R. WEISMAN W.
HODGES R.
JONES E. ADENSAM A.
THADANI J.
LUEHMAN J. HAYES J. BIRKUT S.
JACKSON FROM ALABAMA POWER COMPANY:
l R. MCDONALD J. HAYES J. WOODARD G. HAIRSTON E. MAKAY Heritage Reporting Corporation (202) 628-4888 l
Wo?
l k f h K 0500034s O 89p;n, g O l l
T i
PNU j,
e
4 3
, y..
2 APPEARANCES (Continu::d)
't-FROM 171RAMA POWER COMPANY:
(Continued)
R. MULLENS N. GATES D. MCKINNEY FROM WESTINGHOUSE ELECTRIC:
R. MAGEE-J. KNOCHEL ALSO PRESENT:
T. CINTULA J. KNOTTS J. MILLER Heritage Reporting Corporation (202) 628-4888 k
3-s o
. - A,4 * '
3 y
l'
'P R O C E E D I'N G S p
j
'2 (12:00 p.m.)-
3:
MR. LIEBERMAN:. This_ meeting is being conducted as l
4 the' result of an enforcement action 88-113, which is an 5
order issued to the Alabama Power Company to pay a civil 6
' penalty arising _-out of an issue involving an accumulation,of 7
hydrogen in the RHR systems at the Farley Nuclear Plant.
8 Since this meeting is being' transcribed, let's begin by 1
9 introducing ourselves for the record.
10 I'm Jim Lieberman, Dir*ctor of the Office of
-11
-Enforcement.
i 12 MR. TAYLOR:
I'm Jim Taylor, Deputy. Executive 13
- Director, a
14.
MR. PARTLOM:
Jim Partlow, NPRe Associate Director 15 for Projects.
16 MR. CHANDLER:
Lawrence Chandler, Assistant 17 General Counsel for the Areas of Enforcement.
18 MR. WONG:
Howard Wong, Deputy Director, Office of 19 Enforcement.
20 MR. MERSCHOFF:
Ellis Merschoff, Deputy Director, 21 Division of Reactor Safety, Region II.
22 MR. HAYES:
Jack Hayes, Back-up Project Manager 23 for Farland.
24 MR. VARGA:
Steve Varga, Division Director, NRR.
25 MR. CINTULA:
Ted Cintula.
Beritage Reporting Corporation (202) 628-4888
4.
'* O 4
st 1
MR. WEISMAN:
Bob Weisman, Attorney, Office of the
'2 General Counsel.
3 MR. HODGES:
Wayne Hodges, Chief Director, Reactor 4
. Systems Branch.
5 MR. JONES:
Bob Jones, Section Chief, Reactor 6
_ Systems Branch.
7 MS. ADENSAM:
Eleanor Adensam, Project Director.
8 MR. WOODARD:
Jack'Woodard, Vice President of 9
Farley Project.
10 MR. MCDONALD:
Pat Mcdonald, Executive Vice 11 President, Alabama Power.
12 MR. MAGEE:
Bob McGee, Westinghouse Electric.
13 MR. MAKAY:
Elmer Makay, I'm a consultant to 14 Alabama Power Company.
15 MR. THADANI:
Ashok Thadani, Assistant Director, 16 NRR.
17 MR. BIRKUT:
John Birkut.
18 MS. JACKSON:
Sue Jackson, NRC Public Affairs.
19 MR. MULLENS:
Rick Mullens, Alabams Power. Company.
20 MR. GATES:
Will Gates, Alabama Power Company.
21 MR. MCKINNEY:
Doug McKinney, Alabama Power 22 Company.
23 MR. KNOCHEL:
Jim Knochel, Westinghouse Electric 24 Corporation.
25 MR. KNOTTS:
Joe Knotts of Bishop, Cooper, Purcell Heritage Reporting Corporation (202) 628-4888
g
- .Y ' h-s:
\\3 5
?r and Reynolds.
l~
L 2
MR. MILLER:
Jim Miller.
11 1:
3 MR. LIEBERMAN::.By way of background, Region II.
24 met'with Alabama Power in an' enforcement-conference-of-t 5
May 25, 1988.
Thereafter on August 3, 1988, a notice of p
6
. violation of proposed civil penalty-in the amount of L
7
$100,000 was issued against Alabama Power.
The violation in 8
this case involved the. failure to maintain two independent 9
ECCS' subsystems operable because the'A-train ECCS subsystems l'0 of both units were inoperable for use in the recirculation 11 mode due to the presence of substantial amounts of hydrogen' 12 gas in the crossover piping from the RER pumps to the 13 centrifugal
- pump sections.
-14 Alabama Power protested this action on October 15 3rd.
On March 28, 1989, the Deputy Executive Director, Jim 16 Taylor, issued an order imposing $75,000 in penalty after^
17 reducing the proposed penalty based on good prior
-18 performance in the area of plant operations at Farley.
19 The response:to this order was originally due 20 April 27th.
Mr. Mcdonald spoke to myself and Jim Taylor and 21 requested an opportunity to provide additior.a1 technical l
22 information for our consideration that may relate to this 23 enforcement action.
We agreed to this meeting because NRC 24 has no desire to go forward with an escalated enforcement t
25 action if the technical underpinnings of the case is in
'i
(
Beritsgo Reporting Corporation (202) 628-4888
]
1
{
[*
9 6
- s. t 1
error.
2 As a matter of procedure, the response of Alabama 3
Power is now due_on.May 29th.
I would expect that we'll l
4 promptly review the mateticls discussed.today and within 5
about two weeks notify you whether our position has' changed.
6 Alabama Power should raalize that we've given extensive 7
consideration to this matter and the past information that
.j 8
has been supplied to us.
?
9 Our position in this matter is a matter of public' i
10 record and was presented in the enclosure to the thrch civil 11 penalty order. For these reasons, this meeting's being 12 transcribed.
Thir is a meeting of technical issues.
We i
L 13 intend to listen today and ask questions for clarification.
14 We don't intend to discuss the civil penalty or the
(
i 15 application of the enforcement policy of the facts of this 16 case.
17 With that background, unless someone from the 18 Staff has any other opening remarks.
19 Pat, why don't you go ahead with your 20 presentation.
21 MR. MCDONALD:
Well, thank you, very much.
We 22 appreciate the opportunity for this meeting very much.
We 23 have been greatly concerned by how the incident has appeared 24 in hindsight, including our own hindsight.
25 We have made an introspective study of the Beritage Reporting Corporation (202) 628-4888 L-_-_-_----__-----
.)
V 1
7 i
s, I
x 11 incident.which, as you know, spanned several years.
Our w
2 objective'today is to share our conclusions from that 3
' introspective study with you.
-l 1
4.
Please help us to communicate effectively with you pt 5
by keeping an open mind, trying to put aside preconceived' 6
thoughts for the time being, and think through.this with us 7
as we tell you about our studies and about the findings.we 8
have to present to you.
9 Also,_if we can present each part of our 10 presentation.we will pass out an agenda, and if we can hold 11
'the comments and questions so we'll have some continuity to 12 each part.- We will have four presentations, one by each of-13 the four people here.
14 Our fifth person, George Harrison, who shows up on 15 the Agenda couldn't get a plane out of Birmingham this 16 morning because of the weather, and so Jack Woodard's going 17 to fill in for him.
18-Our bottom line in this meeting is that we are
. 19 convinced that the safety injection systems were fully 20 operable despite the accumulations of hydrogen.
Please bear 21 in mind that neither we nor Westinghouse have a specific 22 scientific explanation as to why the hydrogen accumulations 23 occurred in our configuration.
But we believe we know in a
.c 24 general sense the physical phenomena involved.
Needless to 25 say, our plant was not designed for something that Heritage Reporting Corporation (202) 628-4888 i
_____________m______.________.m.___
.s '
i
's
'8 1
Westinghouse cannot yet define.
2 As far as we know, we were the first to recognize 3;
this phenomena and report it to the NRC and to our peers in 4
the industry.
We know that several other_ plants have 5
subsequently encountered similar symptoms and I'm sure that 6
you've been informed about those.
7 Before proceeding with the details concerning 8
system operability, let me relate our conclusions concerning i
9 the introspective look at our managerial actions during that 10 period of time.
Our key question in hindsight has been how 11 could we have gone so long.without recognizing the 12 propensity for the hydrogen accumulation and recognizing the 13 phenomena involved.
14 After studying this, we've found that our actions 15 were consistent with prevalent managerial propensities 16 within our industry today.
It was an unexpected complex 17 phenomena in an unexpected location.
Specifically, we have i
18 reviewed several other industry events involving failures to 19 recognize phenomena in an unexpected location that 20 constitutes some type of threat to safety system 21 reliability.
And among these examples, we have identified a 22 common thread.
The common thread is a little hard to 23 describe, cut it consists of two conditions, two 24-preconditions for recognizing a phenomena in an unexpected 25 location.
Heritage Reporting Corporation (202) 628-4888
o e
L 9
1 One condition is that the persons involved must be 2
knowledgeable by prior study of that phenomena, and they 3
must be knowledgeable through experience of working with 4
that phenomena in some means.
5 The second condition which they also need-to 6
recognize a phenomenon in an unexpected location is that 7
when they're faced with a problem for which they have a 8
handy and ready familiar solution that they will probably 9
not discover that phenomenon.
That it's only after we're 10 forced into facing up to it with no alternative solution, do 11 we really then, are we able to discover it, identify it.
12 But those two conditions, and I'll go over those 13 some more because they have some important meaning for us 14.
all for the future.
15 Let me describe four examples and use as a first i
16 example this hydrogen accumulation issue.
Please also let 17 me narrate these descriptions in the first party.
18 Now, I was not directly involved in all the 19 examples as much as some other persons, nor should I be 20 considered necessarily representative of the literally 21 hundreds of people who were involved to some degree.
But 22 nevertheless, I'd like to put it in a personal sense because 23 I have been personally involved in a lot of this, and 24 because I think I can express it better.
25 Concerning the hydrogen accumulation, in thinking Heritage Reporting Corporation (202) 628-4888
- _ _ _ =
4 -
10 1
to our two conditions, I'd like to give you some background 4
2 on how I came into this issue, and how there are many people 3
like me that had typical background in the areas of concern.
4 My background of knowledge through study and experience 5
started at the Naval Academy about 40 years ago.
At the 6
Naval Academy, I learned such things as fluid mechanics, 7
absorption, desorption, entrainment of gases, the theory of 8
how the DA tanks work, the necessity of venting systems.
I 9
remember the lucite models where you fill and vent them, and 10 if you've got any bubbles going out of there, how they 11 collect, and you've got to vent them all.
And if you 12 depressurize or drain any, you've got to go back and vent 13 again.
That was my first study.
14 On my first ship, which was a destroyer, I saw 15 those principles in action.
Of course, one of the things I 16 remember is the operation of the DA tank where we forced 17 desorbed air from the feedwater.
I also remember that on 18 our destroyer, we had a vent from each cooler and a common 19 vent header and that those vents were to collect air that 20 was basically desorbed when the sea water would be heated in 21' the cooler.
That some air would come out of the solution so 22 you had to have a constant venting.
1 23 I also remember that in rough weather, you had to 24 watch the suctions pretty well because you could entrain 25 some air in it, and you had to watch the venting to make Heritage Reporting Corporation (202) 628-4888
e L,
11 l
,e
.1 sure that they operated.
i 2
On my first submarine, which was an old diesel 3
type, I got a lot more examples.
One of the. very more vivid 4
ones that in surfacing a submarine, when you blow low 5
ballast tanks, if you blow too much, the air will come out 6
the bottom and envelop the ship and it'll air bind all your 7
cooling water systems.
So you don't blow too much air, you u
1 8-want to avoid that.
9 One of the things I remember most vividly is my 10 qualification for submarine.
I was given a practical exam 11 of getting a ship underway from a dead stick condition.
Of 12 course, there were safety observers; they didn't trust me 13 that much.
And part of that involved starting up a diesel 14 engine.
And the one that I always remember about a diesel 15 engine that was in my brain indelible was that you've got to 16 vent the fuel oil, you've got to vent the fuel oil.
Well, I
- 17 didn't like venting the fuel oil because I don't like to 18 smell that stuff anyway.
But I had a heavy reinforcement 19 there.
20 Then as I went into the nuclear program in 21 subsequent years in nuclear submarines and so forth, I
22 learned about the addition of hydrogen to control oxidation, 23 and also learned that when you're using hydrogen and you 24 depressurize the system, you've got to vent off the hydrogen 25 very carefully because it'll desorb and it represents an Beritage Reporting Corporation (202) 628-4888
12 1
. explosion hazard.
2 Over those years, I learned many practical lessons 3
including mechanics, including profiles and flow paths, net 4
positive auction head problems, and decapitation.
5 After joining the Farley project, my prior 6
knowledge was reinforced and expanded.
The use of a volume l
7 control tank appeared to be a very clever means of handling 8
let'down, maintaining inventory in the system, putting gas 9
hydrogen in,and taking hydrogen out, injecting and purging, 10 and to maintain a water seal on the systems feeding into the 11 reactor coolant system.
I 12 During my years at Farley, I don't remember any 1
13 events about gas accumulations until up to about 1988 that 14-came out of the industry.
In reviewing, I've only been able i
15 to identify two; one was the Calvert Cliffs incident to 16 where they got air into a service water system; and the 17 second was an Oconee event where they had a different 18 arrangement than our DCT but what in essence it was, it 19 correlated to running a DCT tank dry so you're sucking 20 hydrogen right down the pump.
(
)
21 Now, to start the series of events to our
(
22 accumulation.
During the 1977 and 1983 time frame, we had 23 big problems with charging pumps.
These problems involved 24-broken shafts, seized rotors, one right after another.
We 25 had Dr. Makay living at our plant a good part of the time.
i l
l Beritage Reporting Corporation (202) 628-4888
'13 0
.1
.There-were various design and material deficiencies.
One of 2
~the most significant deficiencies was in the shafts; the 3
shafts,weren't strong enough, they'had to modify the shafts.
'4 During this period of time, we were constantly 5
changing pumps.
It was very expensive.
We always had 6-something wrong.
During that period.of time, we did not 7
~ encounter any significant problems with gas.
There was one 8
exception.
While we were starting,up Unit 2 in about 1980, 9
it was found that the B pump of Unit 2 would not pick up the 10 suction like the other pump.
And when we looked at it, we 11 found that the configuration was different from the other 12 five pumps at the plant.
13 Well, we had a ready solution.
We know how to 14 handle that because if you have a pump that's not picking up 15 the suction, this B pump was the one you'd want to have it.
16 Because it was-an installed spare which could be placed in 17 either tray.
Andsoinordertogetpashourpumpproblems 18 that we have, we would use it in a constant on the run i
19 basis, never make it a standby pump and when we had to start 20 it, well, we fiddled with it and vented at it, and so forth, 21 as necessary to make sure it was reliable.
22 As we proceeded from that period, after we got our 23 pumps fixed, we established an extremely reliable 24 performance trend on these charging pumps and safety 25 injection pumps.
We were very pleased with that.
We still Heritage Paporting Corporation (202) 628-4888
R
-l
= r e
14 1
a1 had this problem-about whatiabout this problem of the B-2-
. charging pump in number two unit.
We had asked our 3
designers.to tell us what the problem was..They.couldn't 4
come up with the problem, but'they came up with a proposal to'put some constant vents in but they couldn't explain why 5
6 the vents were needed, nor could they' explain why they would 7
oven work,'because the balance of pressures wasn't so that 8
we were convinced that they'd work at all.
9 And furthermore, at that time, we had a very
'10 reliable arrangement and.the modification involved some air 11' operated valves which'would make things more complex, and we-12 thought would be a significant reduction in the reliability ~
13
.of our system.
So we carried that along trying to look for 1-4 a good solution.
15 That happened, then we had another event that 1 46 happened in 1987 and it happened on a 1-A charging pump.
17 And that is, we found some gas in the system.- And we went 18 back and looked at it and it coincided with operating the 19 pump in the system with a decrease in the VCV pressure of at i
20 least five pounds.
And so when we looked at that and 21 analyzed it, we thought that that could pretty well fully 22 explain how that gas got in there.
23 Now, let's come up from that time to February 26, 24 1988.
I received a call that day from the Farley plant that 25 says that we were venting the 1-A RHR suction -- we really Beritage Reporting Corporation (202) 628-4888 1
v v
15 ys N
<;l
-weren'tLventing the'l-A RER kuction, we were trying to get a-
.2
- samploifrom it because there was a variation'in the boron 3;
-concentration in the primary loop and they couldn't figure 4
.out exactly.why-it varied'as much as it did.
So they were.
5.
going around taking boron samples.
-6L And when'they tried to get a' samples from the high 7
point in the RHR suction,fthey got gas out.
8 MR. LIEBERMAN:
RHR to charging pumps?
9-MR. MCDONALD:: Right..They got gas out.
So I got 10
.this call reporting that this gas came out and I says, oh, 11' my gosh, we've really screwed up somewhere.
And=I said,-
12 what happened.
Well, I don't know.
I says, well, we've got 13 to find out.
Yes, we've got to find out.
14 So over the next several hours, they went:back 15 again.-
And gee, they'd gotten all the gas out once, and' 16' they got gas out again.
And this time, I said, well, what 17 is it.
Well,. heck, I don't know what it is.
Well, we've.
18 got to find out.
We've got to take a sample of it and find 19 out what it was.
20 So then the next report I got back was, it's 21 hydrogen.
Oh.
So after getting that report, I called in an 22 engineer or something who had a drawing, sat down, and 23 within 30 minutes, I recognized that we were, that the 24 problem was probably a phenomena of desorption.
A 25 desorption caused by the decrease in pressure along the Beritage Reporting Corporation (202) 628-4888 s '.
l:*
16 1
tortuous flow path of the water from the VCT to the pumps, 2
one in which the gas would be desorbed and then entrained, 3
plus a possible influence of a Bernoullis effect crossing 4
the RER suction line, or rather where it intersects with the 5
RER' flow path, you have a pipe off the top and a Bernoullis 6
effect probably combined with that desorption pressure and 7
in some way got gas up into the system.
8 Now, I say that and there was one other 9
possibility and that is entrainment of hydrogen from the 10 VCT.
But we viewed that, and we've always viewed that as 11 pretty remote because of the size of the tank and the size 12 of the flow and that sort of thing.
13 Now, to this day, Westinghouse and we cannot 14 explain why that the pressures that should be in that design 15 system by tortuous paths should drop and gas should come out 16' of solution.
And everyone I've talked to has really been 17 extremely surprised at that.
And I was surprised.
We 18 really felt like, God, we've discovered something.
19 So we knew we had a problem.
From there on, we 20 went -- that was on the 26th that we first encountered it on 21 about the 28th until we confirme ' it was hydrogen, and then 22 after we checked that whole system out, we decided we'd 23 better review the configuration of the other unit.
And on 24 the next day, we found it in the same pump and only in the 25 same pump suction on unit 2.
There were some similarities Heritage Reporting Corporation
-(202) 628-4888
s' 17-
. s.-
~
1 auui there were~ some. differences and Jack will go into that.
l 2
So we then-of course recognized what we had.
We l
L 3
took precautions, corrective action to avoid problems with 4'
it, and continued on.
So in our introspective study, I want
.i 5
to apply this thread among four events.
And the other 6
events are going to be a North Anna plug incident, emergency 7
core cooling crack that we had at Farley, and a pressurizer.
8 surge line problem.
9 The two' conditions that constitute this thread is, 10.
one, to have prior study of the physical sciences involved 11 in the phenomenon.
Two, is to be familiar with it by seeing 12 practical applications.
And the second one, the second of 13 the conditions is to not have a ready solution to the 14 problem perceived.
See, we had had a ready solution to that 15 B pump.
We got instant reliability.- And the common thread 16 that I'm going to show on all these, that if you don't have 17 a ready solution and you are forced like we were forced on 18 the 26th of February to recognize that the root cause may be 19 phenomena-related, then we can recognize it.
And of course, 20 we did.
21 Nowh let me go from that event to the next event, 22 and let's take that as the North Anna plug incident.
About 23 1980, I knew and essentially the entire industry knew that 24 in Canal I-600, no matter how well it was stress-relieved by 25 heating, would in time at Ph temperatures we were using Heritage Reporting Corporation (202) 628-4888
-m--___
m
_.--_.____m-__...._.m_
L 18 I
would develop stress corrosion cracking from the primary 2
site.
That was extremely well established in the lab and 3
backed up by what we were seeing in the field.
In fact we, like many other' people in the industry, preemptively plugged 4
5 the first-and second rows of our steam generators.
6 Concurrently with this action taken by the steam 7
generator owners group, now the steam generator owners group 8
has been running a long time and I've been on their 9
executive committee for about ten years and at every 10-meeting, we'd discuss the Canal I-600 problem and when it 11 became obvious it was a problem we couldn't boat, we rapidly 12 tried to get I-690, we immediately tried to get I-Canal-690 13 qualified.
We went to great expense, much effort, review 14 over time.
And so we knew this about I-canal-600 for ten 15 years.
16 And we talked about it every meeting, all the 17 industry representatives.
We were in a steam generators 18 group meeting this last winter during the time when the 19 North Anna plug occurred.
We had Jack Wilson, who was the 20 steam generator rep from Virginia Power, to tell us about 21 what they were just finding on the plug event.
And when he 22 described it and was talking about this plug with something 23 like this, well, what was it made of?
Nobody knew.
During 24 all those meetings that we had in the steam generator owners 25 group with the thousands of people in the industry knowing Heritage Reporting Corporation (202) 628-4888
s 4
19-s'
+'1
-the 600 would fail, me, and'many.other people, never thought
+-
2 to ask, what are you making the plugs of.
3 We asked about what design is it.
This is despite 4
the fact that many utilities had leaking plugs.
We never 5
had one.
But many utilities had leaking plugs so they'd 6
take them out and put one back in.
So here is a case to 7
where we have a phenomenon-related root cause that the 8
phenomenon was extremely well known in the. tube application, 9
but in that simple wedge-like device that you put in there, 10 that when you put it in you have to expand it, which puta 11 lots of stresses in it, doomed to fail sooner or later at 12 the pH's involved, we never even thought to ask what they 13
'were made of and the designers never oven thought to realize 14 that they would fail.
15 Okay.
What does that show about these two 16 conditions.
One, hell, we knew the phenomena back and 17 forwards and we were experiencing it everyday, a whole bunch 18 of people.
Two, we had ready solutions for the problem with 19 plugs.
You pull one out and put an improved design in.
We 20 never looked at the phenomenally related cause.
21 Okay, let's go to another one.
In late 1987, we 22 were starting up our Farley Unit 2, I think it was.
Was it 23 Unit 1 or 2 on that crack?
'87 when the ECC cracked, pipe 24 cracked -- Unit 2.
We started up Unit 2 and we were in the 25 process of start-up when we started to get indications of Heritage Reporting Corporation (202) 628-4888
i**
1 20-1.
leakage in containment.
And so we stopped, went down, shut 2
down,.depressurized and found that the emergency core 3
cooling injection line right near the juncture to the l-4 reactor coolant system just downetream of a check valve had 5
developed a circumferential crack.
Gee, what could cause 6
it.
We had no idea.
What could cause it.
7 So we went down the check list trying to figure B
out what caused it, and finally we instrumented the line and l
l 9
heated up again, and we found that the high pressure 1
10 emergency cooling system, which is at a higher pressure than 11 the RCS, had a small leakage from the. higher pressure to the 1
12 lower pressure RCS.
The leakage was leaking past a boron 13 type condition valve and it would leak just enough that it 14 would. build up pressure behind that check valve, and then L
15 burp the check valve.
And the check valve was sitting there 16 burping about every 30 seconds or a minute, and when it 17 burped, it gave a stress,' thermal stress.
We found it.
18 Now, gosh, why did that have to be a surprise.
We 19 knew from the steam generator nozzle problem, which has to 20 do with thermal stresses, that when you mix one water with 21 another, we knew it'd cause thermal stresses.
None of us, 22 including me, the plant, the industry, had recognized that a 23 small leak from a high pressure injection system, if it 24 snuck by, could cause that kind of a problem.
25 Now, looking back at it, you know, we had no tech Beritage Reporting Corporation (202) 628-4888
4 4
21
., 1 spec requirements, we had no internal requirements to leak 2
check or double valve or anything else to make virtually 3
certain that there would be no leakage.
Why didn't we 4
discover it earlier?
We didn't discover it until we got hit 5
between the eyes with no alternative solution.
We had to 6
find it.
That's the same criteria applied again.
7 Let's take one more.
Let's go back to the steam 8
generator nozzle problem for where the nozzles carry both 9
auxiliary feedwater and main feedwater.
And the cracking 10 that started in probably the late 70s, and was recognized.
11 And it was recognized that it was due to some thermal 12 cycles.
But it was thought by most people that it was 13 because you would have a slug of cold water and it would get 14:
hot and change temperatures, and you'd have another slug.
15 But it was surprisingly found that that wasn't the I
16 case at all.
That we were having a stratification in that 17 line and the stratification would be in layers, and that 18 would create tremendous stresses on that nozzle and would 19 cause cracking.
We knew that.
We put improved nozzles in 20 the Farley steam generators.
The whole industry went 21 through knowing this problem.
So we knew about the so-22 called striping phenomena.
Everybody knew about it.
23 Did we think to look for that phenomena in some 24 other location?
If there was one place in our whole plant 25 who would be one which we would think of first for any kind i
Beritage Reporting Corporation (202) 628-4888
22 s.
l' of that type of striping, where would we think of it now?
2 We all know.
Obviously the pressure line.
But did we think 3
of it?
No, we didn't think of it until the trojan pressure 4
line was found deformed.
5 So let me, I could go on.
I could go on to steam 6
generator hideout and the effects of it and the damage it 7
does to tubes if you con't recognize it.
I could talk about 8
erosion corrosion.
But each one of these comes as a result 9
of two conditions coming together.
You won't find them 10 unless they come together.
And they are the conditions that 11 I've repeated to you several times.
12 So looking back on this, in summary, from our i
13 introspective study of our managerial actions with this l
14 hydrogen issue, we believe that our performance was typical 15 of the industry.
However, now that we have more insight 16 into the recognition of phenomena related root causes, 17 phenomena related root causes, we and our industry should be 18 able to more effectively avoid surprises like those we've 19 had.
20 I'd like to now leave that issue of introspection 21 from a managerial point of view and turn the meeting over to 22 Jack Woodward who will talk about the system description of 23 our ACTS.
24 MR. JONES:
Before we go on, I thought we were 25 going to have questions after each part.
l l
Beritage Reporting Corporation (202) 628-4888
Yes, certainly.
2 MR. JONES:
Okay.
In listening to you describe 3
the actions taken following the discovery of a lot of gases 4
or air at the time of your initial -- on the 26th of 5
February.
6 MR. MCDONALD:
Let me interject something for the 7
record.
The dates he gave are incorrect.
The correct dates 8
of discovery, initial discovery are in the LER and --
9 MR. JONES:
Okay, that's what I was going to ask.
10 MR. MCDONALD:
And I knew that.
He's trying to 11 make a point and he wasn't remembering the right dates.
The 12 initial date of discovery as reflected in the LER was Mkrch 13 lat.
14 MR. JONES:
I just wanted to know whether it was 15-preliminary information.
16 MR. MCDONALD:
Looking backward, once you find 17 initial discovery, then you start looking backward to find 18 out where all the symptoms came in.
The first physical 19 symptom was on 2-26.
We didn't really get the thing 20 solidified and recognize where it was discovered in both 21 units until the first.
22 MR. JONES:
My reason for asking was to ascertain 23
- thether that was new information --
24 MR. MCDONALD:
Okay.
You know what I thought you 25 were asking --
Heritage Reporting Corporation (202) 628-4888
.g L
24 l
s l-
,1 MR. JONES:
-- or is the LER the correct dates.
I 2
MR. MCDONALD:
-- what I thought you were asking, l
3 in the cited violation, you didn't say the 26th of February, 4
you said 6th of February.
5 MR. PARTLOM:
How do I tie your thesis and 6
phenomenon in -- as I understand it in catch up reading, 7
some engineer in 1979 predicted this phenomenon.
8 MR. MCDONALD:
Yes.
Yes.
Now, that --
9 MR. PARTLOW:
And in the ensuing years, you had 10 some plans to put in some modifications to address it, and 11 you didn't.
How do I tie that in --
12 MR. MCDONALD:
Okay.
I'm glad you asked it 13 because I've enjoyed in this study, because you know, again, 14 looking back in hindsight, you say, good grief, how do we.do 15 these things.
Well, you know, on design changes for 16 example, we have design change meetings with management 17 every year and we go down the list, and we go over this 18 list, and we discuss this design change, is it ready.
No, I 19 don't think we've got it yet, it doesn't look good.
20 And it started way back then, but the changes that 21 that young man raade, in my opinion, he was a fluid man and 22 he was looking at the system like you wanted a vent at 23 virtually every hypo.
He had essentially no theory behind 24 it.
When we looked at what he was saying, we viewed it as 25 something that's explainable in the test program.
For Beritage Reporting Corporation (207.) $28-4888
r
's.
v -
si' 25 1
example,Ehe was giving his for a test run where they had 2
filled the system during start-up with nitrogen.
f 3
And it was in a test program.
So you expect in a I
4 test program,'you always have your start up people-and
]
5 they've got 5,000 ways the plant should have been built.
i 6
And we felt that he saia those and it hadn't been in 7
perspective of what the whole system was doing and how it 8
would operate normally.
9 And we went back over that.
You know, you kick 10 yourself trying to figure out why you missed them.
And see, 11 he really didn't give a phenomenon either.
Juxi here's one 12 thing typical.
Right after this event happened, I've told 13 this story about suddenly the light bulb going on within 30 14 minutes after. hearing it was gas.
When I told that to him, 15 what did you say?
Never thought of it.
And he worked on 16 our pumps.. We had an NRC team that was down looking at the 17 pumps, because we were having so much problems.
18 Other utilities can hardly believe.
Why?
Because 19 what you set down with a pencil on paper, it doesn't come 20 out by the numbers.
Now, obviously, there has to be 21 explanations.
But another thing obvious is in designing 22 these systems, there was no spec, for example, on the pumps.
23 There was no spec on how much entrained gas or anything you 24 can pump.
It's just a really unexpected phenomenon turning 25 up.
Heritage Reporting Corporation (202) 628-4888
j I
12 6 l
j 1-And we sal' we know what it is.
We don't know.
i I
2 exactly what it is, but we know enough to'be careful and to-3 have our system function where it's very high' reliability l
4 now.
5 MR. WOODARD:
I might add to that, there was never 6
a symptom of gas accumulation.in the safety injection supply 7
line in the A train in the operating plant until 2-16-88.
8 All of the symptoms.that we saw of gas had to do 9
with gas in the suction piping and pumps, which is a 10 separate issue.
11 We-further searched for gas with Dr. Makay in the 12 late 1970s as a contributing cause to shaft. breakage.
And 13 we did find evidence of the sorption, as you probably would 14 in any pump to some degree, but not to the degree that led 15 to shaft breakage.
16 Then later on we discovered shaft breakage was due 17 to a design and heat treat phenomenon.
And that was 18 corrected.
19 The agenda, Part 2, systems description, at this 20 point our objective is to explain to you that the pump would 21 operate, accepting the gas in the line.
That is our 22 principal objective.
23 Our purpose is to explain the charging pumps to 24 you, what they are used for and where the gas was in the 25 system.
And you can stop me any time.
Because if you don't Beritage Reporting Corporation (202) 628-4888
+
1.. ;
- f
j.?
27
~,
1 understand it, now is the best time to talk about it.
2 We have three charging pumps in our plant.
You 3
may find this a little different than other plants you've 4
looked at, other TWRs..Some have separate charging pumps 5
?from safety injection pumps.
And these are dual purpose 6
. pumps.
7 One purpose they have is for normal. charging, 8
which allows you to maintain the inventory and the rapid 9
coolant system.
You do that by taking suction'on the volume 10 control tanks going through the pumps to the reactor coolant 11 system through the normal charging line as well as the 12 reactor coolant pump seals.
13' The pumps are arranged in two trains'-- A train 3
1 41 and B train.
A pump is A train, C pump is C train.
B pump 15 can be aligned to swiiag from either to either train.
16 The other purpose of the charging system is
-17 emergency core cooling, a part of the emergency core cooling 18 system.
19 It has two subpurposes under that:
one for 20 injection and the other for recirculation phase.
21 Injection would begin at the receipt of the safety 22 injection signal.
At that point, a pump in each train would 23 start.
Also, the RHR pumps would start.
The RHR pump 24 suction would be lined up to the RMST and they would begin 25 to either inject water into the reactor coolant system or I
Beritage Reporting Corporation (202) 628-4888 j
I' I
7,,
J 28 1
recirculate, depending on-the pressure of the coolant
{
2-system.
3 The charging pumps, at this point of initiation of 4
safety injection, would be taking the auction from the 5
volume. control tank snd then the RNST suctions would open.
6.
When the RMST suction became fully open, then the-7 charging pump suction to VCT would close, giving you-suction 8
from the RMST through the charging pumps into the reactor 9
coolant system, injection phase.
10 on down the line, when you get into the ECTS 11 application of charging for recirculation, which by the way 12' would occur for the event of concern for this particular 13 scenario of gas accumulation, the area of concern is limited-14 to the A train; it is limited to one to four inch size 15 breaks where you maintain the reactor coolant system 16 pressure up, im at, it does not drop all the way down to 17 where RER can come in and provide you immediate core 18 cooling.
You are reliant upon charging to provide core 19 cooling on this limited break.
And it occurs about two to 20 four hours into the event.
It occurs means that is where 21 you start needing to supply charging suction through recire.
22 Okay.
Recirculation phase.
23-At this point, the charging pumps continue to i
24 operate.
They are never shut down.
The point that you need 25 to go on recirculation, meaning you are running out of RNST Heritage Reporting Corporation (202) 628-4888
_ = _ _
L I
29 l
,I water and you need to start recirculating the coolant from 2
the containment sump, back to the core for cooling purposes 3
with the reactor coolant system still at some pressure above 4
the RER discharge. pressure.
5 At this point you shut off the RER pumps.
You 6
open the RER pumps' discharge valve to the charging pump
'7 suction.
You start the RHR pumps.
And at this moment, the 8
charging pumps -- remember they are still running -- are 9
taking suction from the RMST as well as to start receiving 10 water from RER which is lined up -- RHR you close the pump 11 down form RNST and open the suction to containment sump.
12 Then you start the pumps and then you are injecting water 13 two ways into the charging pump at that moment, until the 14 RHR pumps build up enough pressure.
And at that point they 15 are solely supplying water to the charging pumps.
That is 16 the recirculation phase.
And that is the phase of concern, 17 because at the point that you start the RER pump on the A 18 train, you purge the gas, the accumulated gas through that 19 line to the suction of a charging pump or A train charging 20 pump.
21 Are there questions about that?
22 MR. PARTLOW:
When you start the RER pump?
23 MR. WOODARD:
When do you?
24 MR. PARTLOW:
No.
You said that happens when you 25 start the RER pump?
I Heritage Reporting Corporation (202) 628-4888 1
c.,
\\ m.
30 l'
MR. WOODARD:
Yes..
Because you hau previously -
2 had to stop the RHR pump in order to line up this suction 3
from the RMST.to the containment sump and realign the
.4 discharge to the charging pump suction.
5 But the charging pumps continue to run, take 6
suction from both places'until that pressure builds up.
7 At this point I would like to show you, take you 8
to the field and.show you where the gas is.
9-See, I told you the gas is in the A train auction, 10 but as you know, it does not look like that.
So let's take 11 a look at what it really looks like.
12 If I could have the human easel, and move this out 13 to the foreground here.
14 (Viewgraph) 15 MR. WOODARD:
The refueling water storage tank is 16 here.
500,000 gallons.
Containment is over here.
And this 17.
is the part of the auxiliary building that this pocket runs 18 through.
19 The charging pumps are located here.
The RER 20 pumps are located here.
21 Remember, on the recirculation phase, we are 22 supplying the charging suction from the RHR discharge on the 23 A train.
24 The 56 cubic feet in question is located between 25 this elbow and this elbow in this run pipe.
But the run Heritage Reporting Corpora 0lon (202) 628-4888 i
o 31 i
~>
,'1 comes from charging pump, from RHR discharge, up through 2
here, over'here to charging pump suction.
The gas is 3
located between here and over here.
4 Now, at the moment you start the RER pump, this 5
gas becomes compressed down into a secticn about this far, 6
1.e. about 12 cubic feet.
So the 56 cubic feet becomes 12 7
cubic feet under that pressure.
8.
Also, when they are started, the flow 9
characteristics are such that the bottom. edge of the pipe as 10-the water shoots kind of underneath, if you have seen any 11 dynamic videos of gas or water going through a pipe, the 12 water really on the bottom leads and the vapor kind of hangs 13-up at the top.
14 MR. PARTLOW:
Are you saying there never is a 15 complete gaseous slug?
16 MR. WOODARD:
At the point -
yes, let me have the 17 other one.
18 Bob is going to talk a little bit more about that.
19 He is our fluid man.
20 But I am only describing this -- here is that same 21 line looking at it from the other direction from the other 22 side, on the containment side.
23 Here is the 56 cubic feet from here around to 24 here.
When these pumps ara started,'and now you are 25 supplying gas, I mean water and gas, from here as well as Beritage Reporting Corporation (202) 628-4888 1
j4*
i 32-1, u.-
t
.1 water'from the RNST.directly, this water compresses this gar 2
but it.also begins to separate, the gas going up tofthe top 3
and the water at the bottom.
4 So the point I am-leaving off the discussion i
5 is'right here, where it iF' Compressed up to here and you 6
have some gas at the top, some. water at the bottom.
7 And he is going'to talk about it when it hits this-8 band.
9 (viewgraph) 10 MR. WOODARD:
Before I highlight the key points 11 and turn this over to Bo% are there any questions about tho' 12 system descriptions?
13 (No response) 14 MR. WOODARD:
Okay.
The key points I am trying to 15 make-are the charging pumps operate continuously during the 16 event, the gas volume of 56 cubic' feet is compressed to 12 17 cubic ~ feet in the RHR line, although it does not represent 18 an exact slug face to face with the gac, the gas' initially 19 is at 33-foot elevation, it's up high in that line and it 20 has to fall-through to 104-foot elevation to get to the 21 charging pump suction.
I.e, it has quite a tortuous path to 22 go.
23 The gas accumulation event is limited to one train 24 of ECCS, limited to the recirculation mode.
It is limited 25 to small breaks of 1 to 4 inches and it doesn't occur until Heritage Reporting Corporation (202) 628-4888
- - _ ~ _. ~ _. _ - -, _ _. _ - -. _.
,1 33 1.
2 to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> into the event.
2 MR. HAYES:
Jack, are you going to discuss the 12 3
feet?
You just mentioned the separation and the, 4
essentially the vapor and the liquid interface, so it seems 5
to me that the 12 feet calculation must be a fairly ticklish 6
one.
7 MR. WOODARD:
No, I am just showing that you've 8
got what seems to be a very large volume of gas and in 9
effect when the pressure builds up you begin to disrupt 10 that, two things happen.
And I am trying to lay the 11 groundwork for Bob.
12 Two things happen.
One, the volume of 13 consi
- a. lon is reduced markedly from 56 feet to 12 feet.
14 Secondly, the liquid and vapor begin to not have a 15 distinct barrier.
You begin to start having turbulence.
16 And then he is going to pick it up from there.
17 MR. PARTLOM:
That 12 cubic feet is based on 18 pressurization up to what kind of pressure?
19 MR. WOODARD:
150 pounds.
20 MR. PARTLOW:
So you are basically assuming pumps 21
'come up to full speed and full pressure?
22 MR. WOODARD:
Yes.
23 MR. MERSCHOFF:
Say Jack, when you vented the pipe 24 you fop.nd 56 cubic feet under the ambient conditions.
25 Is there any basis for assuming that most, that Heritage Reporting Corporation (202) 628-4888
J s.
[.
r.s.,
34 i
v
,1 you could not' add'100, 200, that that point in time is'
'2 characteristic of the worst case?
'3 MR. MOODARD:
Yes.
It hasn't.been touched ~in a I
e 4-very long period of time.
And the subsequent tests that i
5
. ere done days later reveal smaller volumes of gas.
w 6
Itcan't prove ~it to you theoretically.
.I can 7.
. empirically show you data after the thing had been sitting 8'
there running for very long periods of time.
9-MR. MERSCHOFF:- It seemsTthe geometry of the 10 system would be amenable to substantially more.
11 MR. WOODARD:
But there is a reason you recall 12 can't exceed -- theoretically I think 80 some cubic feet is-13 the. maximum we figured you could theoretically get in thel 14
. pipe under this condition.
And it was of interest to us 15 that the actual maximum number was around 56.
16 MR. MCDONALD:
But it involves the flow paths and 17
-the gas won't collect in the flow paths because it would be 18 entrained if it comes out after it desorbs.
19 MR. WOODARD:
See, it isn't going to go past the
-20 highest point.
21 Secondly, it is not going to go past, the gas 22 volume is not going to go past where the pump suction is-23 pulling it back.
I 24 Theoretically, that gives you an 87 cubic feet 25 maximum.
The actual maximum after a long period of Beritage Reporting Corporation (202} 628-4888
= _ _ _ _ _ _ _ _ _ _ _
y s.
35 I4
, 11 operation was'56 cubic feet.
2 MR -MCDONALD:
75.
3 MR. WOODARD:
757 Sorry.
75.
So maximum is not 1
4
.much more than the actual we found.
f 5
-Anything else on system description?
6 MR. PARTLOM:
I found something here that saya 1A 7
and 2A piping are self bent.
8 MR. WOODARD:
1A and 2A piping do not lend 9- ~ themselves to accumulations in the auction piping, which is 10 a separate, a separate problem that we had was accumulation 11 in the-B pump when it set still and the other pumps were 12 running, and their piping configurations were different in 13 that they ha'd locally a high point near the pump suction.
14 And the A pump configurations in the suction of the piping 15 were not.auch that they had a local high point, but they did 16 have in the A train back up the line way bgek outside the 17 room that tied into that RER to charging pump suction.
18 Anything else?
19 (No response) 20 MR. WOODARD:
Okay.
At this point, I would like 21 to' turn it over to Bob Magee, senior engineer in the fluid 22 systems design with Westinghouse Electric Corporation.
And 23~
he will discuss further our system interaction.
24; (vievgraph) 25 MR. MAGEE:
Patrick said I am in the area of fluid Heritage Reporting Corporation (202) 628-4888
- _ = _ _ _ _ - _ - _ _ _ _ - _ _ - _ - _ _.
- n.
~ w :, -
'3 6 '
1
' systems.-.specifically-I'am'the lead fluid systems engineer-2 for Plant.Farley.
I have been in that position'for in
'3 Lexcess of.five years.
I am very' familiar with the system 4
requirements of the ECCS system-and how it would play into 5
the accident analysis.
6 More specifically, we are looking at a subsystem 7
of the ECCS, that is, the high head safety.Anjection system 8
and how that supports the overall requirement.
9 I want to just reiterate again what my discussion 10
.is going to be,about, and what Dr. Makay is going to get
.11 into.
12 We are going to talk about the issue of-13 operability.
I think that is the key issus here, whether 14 the system is operable or not.
15:
I am going to address from the systems and what 16 the meaning of operability is, what is required of the 17 system.
18 Dr. Makay then is going to get~up and talk about 19 how the pump,.given the conditions that we have seen, will 20 operate and can it support the overall requirement of that 21 system.
22 Now, the initial Westinghouse evaluation goes back 23 to we were notified of the observation of the plant 24 identified that very weekend that it happened.
25 (Viewgraph)
Heritage Reporting Corporation (202) 628-4888
s-
~'
37
)
i
-1
-MR..MAGEE:
Monday' morning, came in, and the first E
2 thing that the Alabama Power Company wanted was to say hey, 3
what can we do about operation of this plant?
Is there a 4
problem?
Should we be operating this plant?
r 5
We had to meet the immediate and urgent needs of 4
'i 6
the utility.
Principally that letter of 3-4-88 was to identify what we can do as far as now that we have knowledge 7
8 that there was hydrogen gas accumulating in the RER charging 9
pump suction head.
10 Me identified a venting procedure and program 11 where we said that if a plant observed that procedure wo i
12 could concur with future operation of that plant, of the 13 plant.
14 The time restrictions dictated that we do a verj 15 conservative analysis.
There was no detailed engineering 16 evaluation done at that point.
17 Again, we are focusing on what do we have to do 18 right now to address future operation or continued operation 19 of the plant.
20 After we got through that fire drill, the next 21 thing that was of interest to myself personally was okay, we 22
.got hydrogen up there.
We have a source of hydrogen in 23 volume control tanks.
Was that where it was coming from; 24 was it desorbing as it progressed from the VCT down to the 25 suction of the charging pump?
Heritage Reporting Corporation (202) 628-4888
e 38 1
^
l' I.went through, put pencil to paper, went through 2
a calculation and I was unable to show that it would happen.
3 Since that time, I have sat down and done a couple 4
of-other times and have reached the same conclusion every 5
time.
6 Now, I am not saying this to suggest to you that I
7 it is not happening.
Obviously we'are getting hydrogen up 8
there.
But the traditional analytical approach would not 9
and cannot predict that you are going to get the hydrogen 10 coming out of solutions.
11 From that point, most of our followup evaluation 12 has centered on again determination of the system and pump 13 operability.
14 (Viewgraph) 15 MR. MAGEE:
Now, when we get to looking at 16 operability, the first question is the very general 17 definition of how we define operability.
18 Operability is understood to be the ability to 19 perform its function.
20 What is required of this system during this period 21 of interest?
22 Now, unlike determining what flow rate 23 requirements during the injection mode are, you can go to an
{
24 accident analysis, or the FSAR, and it is all plainly in i
l l
25 there.
I Heritage Reporting Corporation (202) 628-4888
- ,a.
4 39
{
w f1 We do not'do a similar evaluation through the
'2 recirculation mode because it has'always been recognized s
3 that the injection mode is the most limiting case as far as j
4
. outflow requirements.
5 In addition, if you go to the plant technical
~
y, 6
specifications, they do not identify a flow rate requirement.
7 with a high-head safety injection system during the 8
recirculation mode.
9 We don't have nice written.out flow requirements.
10 Westinghouse-has always considered the requirement of.a high 11 head safety injection' system during.the recirculation' mode 12 to be equal to or to be able to demonstrate that you get 13 sufficient flow to meet boiloff so that you can prevent core 14 uncovering.
15:
We performed evaluation to determine what those i
16 flow requirements were.
17 MR. MCDONALD:
Would you explain that in the 18 recirculation mode that out of pump head and everything you 19 can't possibly meet the tech spec for the other modes?
20 MR. MAGEE:
Okay.
Let me jump ahead here.
21 Surveillance requirement 452I.
I am paraphrasing 22 it a little bit here but it says that the ECCS subsystem 23 shall be demonstrated operable but performing a balanced 24 test and verifying the following flow rates.
25 And for the high head safety injection system it Beritage Reporting Corporation (202) 628-4888 l
1
N 40 1
says greater than or equal to 193 GPM for branch line.
2 In the case of a low head safety injection system, which is also covered under this surveillance requirement',
3 it says greater than or equal to 3981 GPM total injection 4
5 flow.
6 The origin of these numbers goes back to the 7
accident analysis for the injection mode.
And we performed 8
a flow rate calculation using design basis considerations.
9 We assumed suction from the RMST, minimum level.
We assumed 10 a degraded pump curve, loss of one train.
At zero RCS 11 pressure these are the ficw rates that were calculated.
l 12 These flow rates then went into the accident analysis to 13 demonstrate that everything was acceptable.
14 Now, when you look at the one flow rate, I use 15 this to cite'this case in the case of a low head safety 16 injection system, the 3981 GPM total injection.
17 When you go on to recirculation, the plant does a 18 couple of things.
19 one, they are no longer taking suction from the 20 RWST, which is a higher elevation, so they' lose some suction 21 head there.
22 In addition, they close some of their cross 23 connect lines in the low head safety injection system.
l l
24 This has in effect basically increasing the 1
25 overall system resistance that the low head safety injection Heritage Reporting Corporation (202) 628-4888
.=
,,e 77 f
- 1 1
system sees.
2 The other thing they do is they line up for the 3
' charging pumps.
So some of the flow is no longer directly injected into the RCS from a low head safety injection.
It 4
5 is being diverted to the charging system.
6 The bottom line is that you could not apply this 7.
flow rate, the 3981, you could not demonstrate using design 8
basis assumptions that you can meet that flow rate.
You 9
will fall below that flow rate in Plant Farley.
In 10 addition, any other Westinghouse plant, you are going to 11 show the same thing.
12 (Viewgraph) 13 MR. MAGEE:
Now, I am talking about the low head 14 system.because here is an example-of where this toch spec as 15 a whole does not apply to the injection mode.
I'm sorry.
16 To circulation.
17 Getting back to what the calculated flow 18 requirements were, what we came up with, Case 1 looked at 19 steady state flow requirements to prevent a not loss of 20 inventory in the core.
21 At two hours, which we have said that is the 22 minimum time that you are going to get the time of interest, 23 where the high head safety injection system is required.
24 I am looking at ANS 71 plus 20 percent decay head 25 loads, roughly 340 gallons per minute.
l l
Heritage Reporting Corporation l
(202) 628-4888 l
t I
1 L _ _ _- __
oO.,
42 1
At that point, minimum charging flow rate is 2
expected to be 550 GPM.
3
.Me looked at another case and recognized when you 4
.get to' recirculation you have already recovered the core.
5 You have some volume of water sitting above;it.
So you said 6
okay, now how long could we last with total interruption of 7
flow, meaning you have this. inventory sitting up there, you 8
have your decay heat, at boil off, but there is a period of 9
time before which to uncover.the core.
10 Again taking the two hours as a reference you are 11 out to about 348 seconds.
12 Again, this was a conservative analysis.
It took-13 no credit for the fact that you are actually going to get 14 spill inflow from the hot legs coming in also.
~
15^ '
(viewgraph) 16 MR. MAGEE:
The final case, which is maybe the 17 most relevant here is to say you might get some degradation, 18 you are not going to get total degradation.
How long could 19 we live with that, before this inventory would boil off?
20 Two points of reference, you can see.
First of 21 all, let me apologize.
This is illegible.
This is 20, 40, 22 60 minutes here.
\\
23 Here is my six-minute point that I showed you i
j 24 before.
Talked about 550 GPMs by just taking a point of
{
25 reference to say that 275 which would be 50 percent, you're Beritage Reporting Corporation (202) 628-4888 L-____--_-
p.
.O' 43 s'
1 out to about 30 minutes.
2 So the safe operating range is really anything 3
below this curve.
And altho' ugh we are going to recognize 4
that when gas goes through, it can affect performance and 5
you can get degradation, in no way are you going to be up 6
here..You are going to be running below this curve.
And in 7
fact, for a relatively, we believe, short period of time.
8 So our conclusions:
9 One is that even with some degradation, that is 10 acceptable, because the flow rate requirements are greatly 11 reduced when you reach recirculation.
12 (Viewgraph) 13 MR. MAGEE:
In addition, based on the input that I 14 received from Dr. Makay and from the evaluation of what flow 15 rate requirements a're, he quoted that the high head safety 16 injection system was capable of performing the required 17 function.
18 Now, I have talked to you about what the system 19 requirements are.
And Dr. Makay is going to get up and talk 20 a little bit about pump performance and whether it can meet 21 these.
22 Just maybe to set some groundwork for Dr. Makay, 23 one of the important things on looking at pump performance 24 is what conditions is the pump going to see.
25 We believe that the pump is going to see a two-Beritage Reporting Corporation (202) 628-4888 i
l l
J
[t.
' Jb.
's 44 w
1
' phase mixture. At no' time is it going to see a solid slug.
2~
This gas is sitting up at the 133 foot elevation.
I 3
The charging pump is running, taking suction from the RMST.
4 The isolation. valve is open and the RHR pump is 5
-turned on.
6 What happens?
As the pump comes up to speed, the 7
first thing is that you have really two suction paths for
.8 that pump.
One is -- let me correct this.
This really 9
should be saying you are taking suction from the RWOC.
In 10 addition, as your RER pump starts to develop a little head 11 you start to take suction from there.
12 So you are.getting a mixing from two suction 13 sources.
14 Think about the compression of this what is going 15 to happen up at this 133 foot elevation. --Now, I have done, 16 I have looked at, had the opportunity to do some translucent 17 modeling.
You have a pipe you turn on a pump.
What does 18 not happen is that you do not get this solid flow of water 19 pushing the gas down this horizontal pipe section.
20 What in fact you have happen is the water tends to 21 slide under the gas.
You have a wave front on the bottom 22 that pushes.
23 Now, this is important because what you are 24 setting up really is a separated two phase flow up at the 25 133-foot elevation.
Heritage Reporting Corporation (202) 628-4888 a-______-______
)
9 if 45 1-Now, that is important because then when you start 1
2 to develop your velocities and push this solution towards 3
the pump, you have fittings and elbows, you have tortuous 4
paths.
All this is tending, you go around that first elbow, 5
you are going to get a torque.
It's going to torque that 6
thing and you are going to get mixing.
You go past the 7
couple, you go past some connecting lines.
The RMST line 8
which is in a horizontal section, it tends to spill flow too 9
towards this two phase mixture and add moisture to that, or 10 water.
11 In addition, you have a phenomenon whereby the 12 pump suction provides mixing.
And Dr. Makay is going to get 13 into that.
14 The important point here is that you do get two-15' phase mixing.
And that is important because Dr. Makay is 16 going to tell you why if you have moisture, if you have 17 water going in there, you have lubrication.
18 That concludes what I have to say.
19 MS. ADENSOM:
I just have one question.
20 How long do you estimate the charging pump is 21 taking suction from both --
22 MR. MAGEE:
It would be a period of time where 23 actually you've got this gas at 20 pounds pressure. It takes 24 a while -- we are talking seconds.
But in the short time, 25 it takes a while until you -- the pump doesn't come right up Beritage Reporting Corporation (202) 628-4888
46 1
to speed because the pump is attempting to compress its gas 2
and it is not until it compresses it to a point that the 3
head overcomes the elevation head of the RMST that you would 4
shut that check valve.
5 However, the vertical distance, there is some 6
period of time that even after you push this through, that 7
you would start to see the first indications of gas.
I 8
believe that the RMST line is still open at that point when 9
you start to see the first indications.
10 The other point I want to make, that even after 11 you close that valve from the RMST, you still have all that I
12 piping downstream of that check valve filled with water.
It 13 is going to be like a spilling cup as the gas passes by it.
14 It is going to keep trickling in there.
The gas 15 is going to want to, some of the gas is going to want to 16 find its own vacuum there.
17 MR. HODGES:
How big is the piping?
18 MR. MAGEE:
Eight-inch piping.
19 MR. EAYES:
You mentioned the difficulty in 20 calculating how the nitrogen got there.
But where do you 21 think it came from?
22 MR. MAGEE:
I believe the source is from the 23 volume control tank.
I believe that local low pressure 24 zones can cause the gas to come out.
You know, we do a 25 traditional approach but a traditional approach does not Seritage Reporting Corporation (202) 628-4888
[
L,, ai p'
i..'
1 o:
L s
47 4
l' look at every elbow, every tee in the local phenomena --
]
I 2
HR. MCDONALD:
Steve, it almost has to.
It's l
3 there, even though you don't protect it, you know that it j
4 is.
5 MR. HAYES:
It's the only source of hydrogen that 6
you know is a volume control --
7 MR. MCDONALD:
Oh, yes.
8 MR. HERSCHOFF:
The hydrodynamic flow regime you 9
described is very complex and I am trying to understand if 10 you based your opinions on anything more rigorous than just 11 a gut feeling and some observations of the lucite model.
12 Is there anything more substantial involved?
13 MR. MAGEE:
I think the observations that we have 14 made are very important.
I think what you are suggesting ~is 15 that have we done a complex three-dimensional modeling of 16 this, which would be required.
17 No.
I don't, we have not done that.
18 MR. HODGES:
Have you looked at a situation where 19 the same size pipe and comparable flow rate to see that the 20 flow regime that you are talking about is indeed the flov 21 regime that would exist beyond the elbow?
22 MR. WOODARD:
In the modeling that I am referring 23 to we actually used smaller pipe but we modeled the 24.
velocities in that pipe.
So it was a scale model.
25 MR. TAYLOR:
This was a special model you ran?
Beritage Reporting Corporation (202) 628-4888
48
+
1 HR. WOODARD:
This was'a model after the fact to 2
show that the venting rate that we had recommended to 3
Alabama Power was acceptable.
4 MR. TAYLOR:
The venting rate was acceptable?
5 MR. WOODARD:
Yes.
6 MR. HODGES:
What scaling criteria did you use?
7 What scaling model?
8 MR. WOODARD:
It was based on both, we look at 9
both Froud and Reynolds.
10 MR. HODGES:
What did you finally scale on?
11 MR. WOODARD:
I'm sorry, but I was not involved in 12 principally setting up the modeling.
13 MR. HODGES:
You can't scale something like that 14 on both of them at the same time.
15 MR. WOODARD:
That is right.
16 MR. JONES:
You said that tech specs in the FSAR 17 did not specify lease or qual requirements and then you 18 specified what Westinghouse believes the requirements are.
19 But you have never documented that anywhere.
Is that 20 correct?
21 MR. WOODARD:
That has always been our 22 understanding.
23 Have we ever documented?
I know we have had 24 correspondence in various cases where plants have looked at 25 what the requirements are and that has always been it.
Beritage Reporting Corporation (202) 628-4888
l 49 l
1 HR. JONES:
And there's always been documented 2-cos. despondence back and forth between the NRC and 3
Westinghouse or the affected utility when a change was being 4
made to a flow requirement?
q 5
MR. WOODARD:
When asked about what the flow-6 requirements are this has always been our interpretation.
a l
7 MR. JONES:
To the best of my knowledge, we have 8
never expected that when we switch over to a recirc. mode 9
and you go on a piggyback operation, we would in fact expect l'
that the flow through the HBI pumps would go up, because the 11 suction source is at a higher pressure.
12 MR. WOODARD:
That is absolutely --
13 MR. JONES:
We have never been under the 14 understanding that you would expect a system flow to go down 15 and when we have done reviews of analyses and looked at 16 system performance for FSARs, it is based on these assumed 17 lineups and performance requirements, not degrnded 18 performance in the systems due to other effects.
19 But you've never documented something smaller.
20 MR. MCDONALD:
There was an important fact here.
21 You said degraded requirements.
The pump cannot meet that 22 spec when it is in the real line condition by any normal, by 23 full performance.
24 MR. JONES:
The pump surveillance testing that you 25 do is to demonstrate that the pump is capable of fulfilling i
l i
Beritage Reporting Corporation (202) 628-4888 i
\\
O!
fl L
50 W
'.1-
'its function..by' showing;it is developing: the. appropriate
~
2 flow rate.
And'that tells us whether the function has 3
degraded or not for compliance with ASME testing.
4.
MR;.MhGEE:
Let me make that.-- you are absolutely-5.
true about the high pressure safety injection system.
.You.
6:
piggyback ~it'with the RER system.
In that case,.you're 7
going to get more flow because you are getting a suction 8
boost.
l, '
9 So if I can' meet 193 during'the ejection mode,.I-10 ma going to by. coincidence bet more than 193 when.I go to 11 recirculation.
That's-true.
12
'But the same toch spec talks about the low head I
13 safety injection system.
And I realize that is not the pump 14 in issue here..But this, I'm citing a case of where this 15-entire spec-was not intended for the injection mode, because 16 this flow could not meet, or the recirculation mode, because 17 this pump could not meet that flow rate during the 18 recirculation mode.
19 MR. JONES:
That spec tells me that that system is 20' performing in its normal fashion.
When we look and evaluate 21 the system we are expecting that it will perform in their 22 normal performance behaviors in their normal lineupe.
23 So yes, it's not surprising that if you are 24 diverting flow to other systems that indeed the flow that 25 comes to the specific system could possibly go down.
But Beritage Reporting Corporation (202) 628-4888
r 51 1
that is known when you've done the FSAR review.
It is not a 2
known that when you shift over to a recirc or piggyback 3
operation that you are going deeper -- in fact, it has 4
always been assumed it would increase.
5 MR. PARTLOW.
This presumption of what would 6
happen, this is'part of the new information from a year ago?
7 Is that right?
You didn't present this a year ago?
8 MR. MCDONALD:
We are presenting new information 9
here in terms of more detailed information from Bob here.
10 We didn't present the information in this detail.
11 MR. LIEBERMAN:
Okay.
Pat, do you have extra 12 slides?
13 MR. WOODARD:
We will provide you a copy of 14.
everything that we use here.
15' MR. MAGEE:
Okay.
At this point I would like to 16 introduce Dr. Elemer Makay.
~
17 Dr. Makay is a mechanical engineering graduate of 18 the University of Budapest.
And he has his Ph.D. in 19 mechanical engineering from the University of Pennsylvania.
20 He has extensive design and testing experience 21 with centrifugal charging pumps.
22 By 1973 he had run over 1,000 tests on a variety 23 of pumps of this type.
24 Since 1973 he has headed his own consulting firm l
25 specializing in hands on trouble shooting and modification Beritage Reporting Corporation (202) 628-4888 L L -
c p
l' I
k.-
52
- ~.-
- 1-
.of power: plant pumps.
2 He has modified over 600 multistage centrifugal y
3-
, pumps.for flow instability.
g 4
He is an internationally recognized expert in the 5'
field and has worked both for industry and for Oak Ridge 6'
National Laboratory,which was under-contract to the Nuclear 7
' Regulatory Commission.
8 Dr. Makay started working with us at Farley on 9'
pumps in the late 1970s.
He worked with us on circulating 10 water pumps, steam generator feed pumps, and of course the 11 centrifugal charging pumps, which is the subject today.
Dr.
12 Makay.
13-NR. KCKAY:
I guess first'I would like to mention 14 that if you don't understand, please don't hesitate'to stop.
15 me because my wonderful Transylvanian Hungarian English may 16 be difficult for you to understand occasionally.
If 17 something is not clear to you, please, you don't hurt my 18 feelings.
I learned my English from scratch and I'm used to 19 not being understood.
1 20 The mkjority of.my work, as Jack said, is trouble 21 shooting, mainly multi-stage concern.
Most of them are 22 larger.
Also, I am directing quite a bit of actual 23 laboratory research and development work and I'm doing quite 24 a few surveys as you will see on some of the slides because 25 the three of them -- I have to have all three to supply me i
Beritage Heporting Corporation (202) 628-4888
53
{
a e
1 with directional work, have to verify that at the nuclear 2
power station,-so the three go hand in hand.
3 But I would like to point out that I really didn't i
4 come here to represent Alabama Power Company or the NRC or 5
the electric parties.
My principal always is that when I go 6
out in the-field, I represent the pump -- and not you 7
because you may be biased by "you" I mean'whoever, but the 8
pump isn't.
Pumps don't' speak Transylvanian English.
They 9
don't speak English, but they understand -- if you really 10 work with pumps, you will end up understanding the pump.
11 Speak up?
12 THE REPORTER:
Please use the microphone.
13 MR. KCKAY:
Can I put it over here because I will 14 need both of my hands.
15 THE REPORTER:
Yes.
16 (Viewgraphs displayed.)
17 MR. KCKAY:
First, I would like to.show a couple 18 of literature, front pages, to show you that although I work 19 the majority of my time is with the utilities, this was my 20 earliest work.
I worked with Oak Ridge for the NRC.
This 21 is on nuclear pumps.
It is a design handbook.
Although the 22 title suggests that we are talking only about reactor 23 coolant pumps, I was at that time told by Oak Ridge that 24 write the hydraulic section for any kind of pump as a pump 25 handbook.
So, this was my first engagement at that time.
Beritage Reporting Corporation (202) 628-4888 I
c; g ;
E' A :
~
O 54 1: ^ Atomic Energy Commission and Oak Ridge.
-2 The second one is a more recent -- that was 1972.
3 This one is Electric' Power.Research Institute., Now this vas.
4-
-- this'is survey of feed pump outages.
What do'I mean by 5
feed pumps?
The majority of pumps to me is just a small 6.
feed pu g.. This charging pump taaybe 500 horsepower and the-
'7 largest' boiler feed pump of tho'same type we have out in the
~
8 field is 70,000 horsepover.
So, we go-from small to big.
9 The advantage of testing and researching the big 10 ones, for example, for'yoa, some of the hydraulic phenomenon 11 don't create. big enough forces for you to notics them.-
12 Don't worry in the 40,000 horsepower boiler feed pump, that.
'13 will show clearly.
14 This-is a little bit more recent.
And this is 15' also for Oak Ridge and this is for you.
The reason I am.
16 showing this because the aux feed pumps basically are almost 17 like the charging pump.
Some of them are opposing power, 18 some of them are in-line powers, but it is basically the 19 same pump we are talking about.
20 This is just an interesting -- I'm sure you are up 21 to date on this.
This is an ASME NRC symposium that will be 22 in August in Washington, D.C.
23 Next.
And I am one of the key speakers on this 24 symposium.
And you can see the subject is low flow 25 instabilities.
And charging pumps, nuclear outfeed pumps seritage Reporting Corporation (202) 628-4888 1
.., u s.
- 55'.
1 and so on,. hydraulically ^ induced unstable forces and low 2
.i flow' operating mode. 'That's the main subject of my paper.
3' And that is: basically what.we.are facing here with the 4.
charging pump when the'aubject comes up of having gaa or gas 5
. mixture hitting the auction, the first stage of the pump.
6.
Now, the next one is just to-show from 1964 to 7
1970, I was doing design research, testing at the major pump 8
manufacturer.
I was restricted.to their own design.
9'
-Since 1970 to present, we had a very elaborate, 10 very adequate'research lab, the biggest research lab on the 11 whole pump manufacturers, including ASBS in Germany and 12 Switzerland.
13 In 1970, when I departed from the pump 14 manufacturer, I started to extend my research lab.
- And,
'15 today, my research lab goes from the east coast to the west 16 coast,-from Texas to Canada.
Some pumps in Alaska, some in i
17 Germany, some in Switzerland, because we need the 1
18 information in a disputed case when we have to pin down:
y 19 What do you base your hypothesis or assumptions are?
20 "Well, for example," like Mcdonald mentioned shaft 21 breakage.
That was one of my little special projects with 22 EPRI and we had to collect hundreds of reports of shaft 23 breakage to really conclude why the shafts broke.
I 24 Now, during the occasion of gas or mixture coming 25 to the pump, the flow is a very complex phenomenon.
In a seritage Reporting Corporation (202) 628-4888 L
i 56 l'
complex ~ phenomena like this we usually like to have flow 2-visualization testings.
L 3'
In this one I just wanted to show some of my 4
research projects, hydraulic instability, axial thrust, 5
system instability, they go hand in hand.
They are always l
6 heavy on the agenda.
7 Now, this is a picture which you can find it in 8
several of my publications.
The flow is so complex that 9
oven to draw a picture is difficult.
And the very first 10 time I saw the visualization testing, inlet at this portion, 11 the outlet discharge, I was shocked.
I couldn't believe 12 that it was that messy.
13 So, I started to work on this, what, two decades 14 ago, because large response, high energy response to these 15 ferces is fantastic.
In the charging pump, we are facing 16 two areas of importance.
One is the inlet.- And one is the 17 discharge.
See how the flow shoots back into the discharge 18 behind the impeller blade, if the geometry is like this.
i 19 And there are two dimensions here.
20 Thic area here on the side, this area here as it 21 is creates very large disturbance in this area and the same 22 on the other aide.
23 Now, this vortex is recirculation flow, this also 24 has a component in the tangential and it is a very strong 25 one.
Now, I worked on one visualization test leak and when I
l 1
Heritage Reporting Corporation (202) 628-4888
_______________A
(
1 57 L
L
- 1 we go down to low flow with these pumps, this vortex can go' l'
2 as far as 10, 20, 30 feet vortexing and really -- all I can l
3 describe it to you is it always is a mess.
And also it has 4
a low frequency component that fluctuates.
5 MR. TAYLOR:
You mean it extends for that 6
distance?
7 MR. KCKAY:
Pardon?
8 MR. TAYLOR:
The vortex in fact extends as much as 9
10 feet?
l 10 MR. KCKAY:
More than that.
11 There is one test that we decided to put, change 12 the plexiglass suction pipe instead of 5 feet to 10 feet and 13-then 20 feet and the vortex at one point did go into the T'
14 steel piping.
So, it is more than that.
1 F.
And that test is so convincing that if you would 16 see it, you would have no question about it.
If I describe 17 it, you eay, " hell, I don't know."
But if you see it, it is 18 just absolutely convincing.
19 So, there I emphasize these two areas because 20 this, the low flows, I need that mixing action.
But when 21 the supposed gas hits the pump I don't want this and this 22 area, the wearing areas -- this is the same manufacturer as 23 the charging pump.
Essentially it's a boiler feed pump and 24 500 horsepower, is 25,000 horsepower and the forces were 25 fantastic.
Every time this pump failed, the failure was Beritage Reporting Corporation (202) 628-4888
V 4
58 l
C 1
miserable and very dramatic.
l 2
Here's that area when the secondary flow at low I
3 flow shoots back into this area which I named " Gap A".
And 4
this Gap A serves several purposes.
One is it is a filter.
5 It avoids the flow shooting back into anything at low 6
frequency pulsation.
7 (2) It isolates the two areas.
This type of 8
velocity, surprisingly, that was not the purpose of this 9
originally.
It straightened out the axial thrust of the 10 pump perfectly.
I'll show you that.
11 We are using this as a low frequency filter.
It 12 is actually a separator and a filter.
The dimensions on it 13 are very small.
While most older designs and some present i
14 designs, this distance is huge here.
So, we permit flow 15 disturbance.
16 The other distance is a narrow vein to diffuser 17 vein.
Imagine this is the rotating impeller vane.
It 18 rotates and here is the diffuser as the impeller passes by 19 and you've got the shock.
And that shock also penetrates 20 behind this area here but more important goes through the 21 whole system.
i 22 Now, there I have -- this is where the original 23 dimension was on this and I changed it to this.
And that of 24 course eliminates that shock because that's simply I pulled 25 the two parts apart and now if there is an amount of Beritage Reporting Corporation (202) 628-4888
'4 4
l 59
'1 distance beyond that, I don't do too much to the shock.
But r
l 2
when the gap is very small, then I double it or triple it, l
I 3
the change.is phenomenal:
shock-wise, noise-wise, j
4 vibration-wise.
l
{
l 5
So, the discharge is important for us.
When the 6
' gas hits the pump I entrap liquid over here.
These j
7 tolerances, these distances of the wearings are very small:
{
8 14-16 mil.
.014,.016, 9
Sometimes I take these pumps apart and I find 10 10 mil.
But it is a very important surface because this is 11 where is friction starts.
And if I don't have liquid in it, 12 in a new pump, for a certain period of time, I start to heat 13 it up and I burn out the pump.
It ceases.
'14 So, the question is how long do I need it and I, 15 on my own, I concluded that I have enough liquid over here 16 that if number of seconds it takes to push that gas through 17 I have not damaged the pump yet.
18 Okay.
If that distance is big, when the gas hits 19 the impeller, I lose the liquids -- I don't entrap it.
20 There is a whirlpool overlap.
And this section is 21 enlarged here.
Between this, I had to chop off for some 22 hydraulic reason, flow guidance.
The stationary and 23 rotating side plates I have small Gap A from here to here.
24 And this distance divide by the gap is called the overlap.
25 And we did lots of research on it.
We had a big Heritage Reporting Corporation
'(202) 628-4888
60
'EPRI research projectLand~this is my pet.
We had to 1
2 optimise:
What ratio do'I need for this purpose, that 3
purpose and so forth.
And the third one,.what fringe 4'
benefit came in that I didn't think of it.
I think.it's 5
like this pump.
Why did I not think of it.
I should have.
6 It provides some very nice damping to the pump at 7
a certain ration.
If less than a certain ratio,'the length 8
of the gap is less, I get no damping.
I still get the 9
filtering effect, but 1x) damping.
If I go up to a ratio of, 10
.say,.8 -- ASBS in Germany was willing to go up to 16 for me 11 for research purpose, and at Sulzer in Switzerland, my other 12 EPRI project, we are up to 8 with them so we have data from 13 0,
1, 2,
4, 6,'8 up'to 16.
14 Then we went too big, I begin to lose too much in 15 efficiency.
So, there is an optimum how far I may go.
But 16 there is a very nico damping effect.
What does it mean?
17 Reliability.
18
- Now, I made the flow stabis.
I show you three 19 examples.
Start at'the top.
The original head curve of 20 that' multi-staged pump, when the Gap A was big and Gap B was 21 incorrect, and when I modified it, it became beautiful and I
22 at about this point the whole power station became unstable.
23 The pump was unstable, the system was unstable.
And this i
24 one removed the instability, the changes, i
25 This is another one.
This is a bad one and Heritage Reporcing Corporation (202) 628-4888 1
- - _ = -
1
U:
o r
61 o
1 because'these'are boiler feed pumps, just for interest I w
=2 show.you--- this'is the Limerick reactor feed pump.
- Now, 3.
Limerick' reactor feed pump.after. modification, well, we went 4
down to 50 percent flow at Limerick.we had problems.. We had 5
stability problems.
And now when I finally got permission 6
to modify the fuel pumps, we went down on high speed, four p
7 speed testing to 15 percent flow and the pump is beautiful, 8
proving that it does work.
.i 9
There were many unexpected benefits coming from 10 this like, originally, if you calculate axial thrust, you 11_
should get something like this but you never get it.
12 When this geometry is incorrect, the smallest 13 amount of change you give to the impeller or the rotor, 14 misalignment'of the diffuser or change of dimension, in 15 three cases here the. red, the green and then-the' black.
16 This is a fairly big pump and the forces were so 17 big that it took'a.few seconds to fail this pump.
There was 18 no -- and the thrust suddenly shifted from here to here.
19.
Percent flow and thrust -- see, it doesn't repeat.
Smallest 20 amount of change made the system completely different.
And 21 this poor power station has had several big failures due to 22 this because the axial thrust suddenly went and took 23 everything with it.
24 After modification on this pump, we do have th.;
i 25 because we did get it because we are fully controlling the Beritage Reporting Corporation (202) 628-4888
{
Lu___________--_------_-----_-------------
.w i
62
]
1
. flow behind theEside plate.
l a'
b 2
This was an unexpected benefit but we welcomed'it.
i3 3
Additional unexpected low benefits'were.that this 4
is low frequency vibration is-due to. Gap A.
This is the
~
5 vane passing frequency that's due to ' gap irr the small -
6' distance from vane to vane.
7 This or it could be anywhere between.45 to.478 8
and I' find one multi-staged pump that that frequency ratio 9
being
.9.
But up to that' point I only find a high of 78.
10 I only~use the combination of the two and when we 11 introduced these changes, these components completely 12 disappear.
So, the pump reliability increases, vibration 13 comes down and this is an additional fringe benefit.
'14 Next.
~
15 Most presently operating multi-station boiler feed 16 type pumps are design with uncontrolled geometry up there.
17 And the diffuser type very easy to convert like the one I 18 showed you with the side plates.
The volute type is more 19 difficult because the gap is huge and you need lots of 20 welding and lots of modification.
21 Now, this is -- it took us a long time to persuade 22 the pump manufacturers that this is indeed right and not 23 only that but it is very important for reliability and 24 stabilizing the pressure.
As I started to point out the 25 pumps that I personally modified -- if you modify one or a Beritage Reporticy Corporation (202) 628-4888
63 1
pump manufacturer modifies one, it is not on my chart.
2 These are what I personally was involved.
And the numbers, 3
as you can see, people began to believe that what I'm saying 4
is needed.
And this is only 1986.
5 These are the same geometry.
6 And this is in connection with the impeller I, 7
this part.
That's a special one.
8 I would like to present two simple examples.
One 9
is the Palo Verde aux feed pump test that I was asked 10 jointly by NRC and Public Service Arizona to run a test 11 because they wanted to find out what happens to the pump 12 when they lose auction.
13 And now if you have a pump failure and you cannot 14 control it, but for any purpose you want to fail it, I have 15 to control it.
So, I said, I agree to running the test if 16 they permit me to fully -- and I said this is my vibration 17 chart:
vibration amplitude, velocity, acceleration and 18 initial data points that we took on the pump was down in 19 this area which is desirable.
And I said, I would like you 20 fellows to agree that we do not purposely go into the damage 21 level or the destruction level.
22 So, that instrument for temperatures, pressure 23 pulsations and vibrations and then let me judge how far will 24 they go.
And I said I don't want to go into the damage 25 level.
1 Heritage Reporting Corporation (202) 628-4888 1
1 i
f 1
E_ _--
l
(L l
64 1
When we got'down to the breakdown point, it took 2
us 8 minutes and it was purposely because time was 3
important.
We went down from this point to here, which is 4
.the pump here is in complete fully developed cavitated state 5
which is a very poor mixture of very little water and lots 15 of bubbles.
7 And I had the pump down here at this point for 8
three minutes.
And NRC requested to take the pump apart and 9
inspect it because damage is possible.
And I pleaded with 10 them and they allowed me to prove that the pump is not 11-damaged without taking it apart.
12 I took the pump back up here, coming up, this is 13 my matching point to this.
And coming down to recire, this 14 is my matching point.
And I said if I can match the head 15 exactly, and no same vibration as before, will you believe 16 me that we did not damage the pump?
And they agreed to it.
17 Now, the other type, the other one is starting of 18 the fire station and suddenly they go too far full so we 19 lose suction and we lose MPSH and the pressure broke and 20 completed.
21
.And this is not measured in the control room.
22 This is flow.
Flow with the startup pump which is a small 23 like the charging pump and when they went up to 145, 150 24 percent flow by mistake, the pressure -- the head broke down 25 and the pump was running dry.
Beritage Reportir 7 Corporation (202) 62b-4888
65 1
Here is the maximum amount of time they are e
2:
running dry'is 5.5 minutes.
Here's another one 4 minutes 3-and 2 minutes, half a minute and this is early in the 4
' morning from 5:00 a.m. to 6:00 a.m.
So, this is a one' hour 5
of constant struggle.
6-I got to this station and I said, put the big pump 7
on because this pump wasn't designed for this.
8 This is a very good proof what can happen to the 9
pump if gas or mixture will hit the pump.
These pumps are 10 horsepower-wise much bigger and this pump had no business 11 lasting 5.5 minutes.
12 There I had a chance.
This is the accident.
The 11 3 other one was purposely planned to see what would happen"to 14 the pump if I run it almost dry for 3 minutes.
15 MR. PARTLOW:
Is what we're'getting at here -- do 16 I understand that Westinghouse at one time said six cubic 17 feet slug of gas pa'asing through a pump could result in a 18 catastrophic pump failure.
19 Are you saying no, that's not necessarily true?
20 Is that where we are going?
21 MR. KCKAY:
I am trying to show you a few 22 examples.
I don't know what that was based on.
Of course, 23 you may be able to answer.
24 I work with pumps.
And pump systems, but mainly 25 pumps.
And I am trying to show similar examples where the Beritage Reporting Corporation (202) 628-4888
66 I
results were definitely contradictory to that.
2 MR. PARTLOW:
Okay.
3 MR. WOODNARD:
Excuse me.
Let me interject where 4
the six cubic feet came from.
5 When we were 1:amediately reacting and I would 6
like, Bob, your concurrence on this one.
We were 7
immediately reacting to the problem.
And the day we 8
understood what was happening, we called Bechtel and 9
Westinghouse that evening and got them involved.
And we 10 asked for early on guidance on what kind of volume could 11 that pump take in their engineering judgment for short-term-12 reaction.
i 13 And that's where the six feet came from.
14.
MR. MCDONALD:
The statement was the six cubic 15 feet would not cause catastrophic -- or less than six cubic 16 feet.
17 MR. KCKAY:
During these testings --
18 MR. MCDONALD:
That was based on the venting 19 program.
We said keep it to less than six cubic feet and we 20 don't have a problem.
And that was, again, that was the 21 very hurry-up evaluation that we did.
22 MR. KCKAY:
Now, these pumps I show you.
They 23 didn't fail because when the full cavitation or almost empty 24 drum -- the flow came into the -- the ground flow came into 25 the section, I had enough entrapped water here that I was Heritage Reporting Corporation (202) 628-4888 E_-_---__-
_-9
' o 67 4
l' able to keep this surface lubricated.
2 Also, this is the time to.show three more items.
1 3
I'11 just name them.
4 MR. TAYLOR:
Does Farley pumps have such gap 5
configurations?
6 MR. KCKAY:
Yec, sir.
See, we started in the 7
early stage was the-Connecticut Yankee that was a 13-stage 8
combination.
And I got involved.
And that was the time 9
when Westinghouse was debating what to do and we went to a 10 new design.
11 Now, talking about these surfaces.
It had 11-12 stages in this pump, so you had 22 surfaces to worry about..
13 Then you have larger areas, you have a balancing drum.
And.
14 that balancing drum is about this long and that has the same 15 clearance in it.
16 We have to maintain liquid out here in the 17 balancing drum which is directly connected to this area 18 here.
And the third area is the seals.
Now, the seals in 19 this case have their own injection system and their own 20 cooler.
21 So, therefore, it has a better chance than the 22 pump with the big gap on the side.
23 What is the next one?
Okay, put that on.
24 We have two items here.
One is the mixing at the 25 impeller I.
You will see some of our testings.
When you go Beritage Reporting Corporation (202) 628-4888
-o 68 a
t 1
down to low flow,-the mixing is very strong.
So the -- on I
2 those test, impeller I effect is very convincing.
I called 3
'the impeller IV inlet.
So, you have a good strcng mixing 4
action.
5 Elbows in the suction piping adds to flow of 6
mixing, hence, it adds to the mixing.
7 Summarizing a couple of effects.
What I am saying 8
here is the charging pumps, aux feed pumps, boiler feed 9
pumps and if the boiler feed pumps are -- 20, 30, 40,000 10 horsepower, where the friction forces are huge, then a tiny.
11 little pump like this doesn't really worry me.
12 These tiny pumps we use in the lab tests, basic 13 hydraulic desk pumps.
And basically all high speed, high 14 energy multi-stage pumps -- so, the proper -- I say modified I
15 because it shouldn't have been modifi-ad by me.
It should l
16 have been 30, 40 years ago that way.
17 I found, by the way, an article in 1941 in 18 Switzerland that Mr. Richey who cwns a small pump company 19 wrote a one page report in that journal in Switzerland about 20 my Gap A and B.
21 And then some of my competitors ran computerized 22 study and they found it.
Sulzer company in Switzerland 23 found it because they wanted to prove that one of their men 24 invented it.
But they found it.
25 So, the proper -- I talk about A and B vaguely Beritage Reporting Corporation (202) 628-4888 1
f,
~
V 69 4.
,, 1 improve' low flow stability.-
This'was the main itom why I 2
went after these components and --
3 Therefore, reliability, also the amount of damping
'4 we.cantadd to the rotor helps tremendously because they use 5
the vibrations.so when the gas' hits the pump the vibration 6
at that point is still damped well by the gap.
7 isternal pump clearances now, behind the impeller-8 side plate, they retain water fos pump internal lubrication, 9
namely those wear-surfaces and the cooling.
So, lubrication 10 and cooling for a much longer period of time than if the 11 dimensions are incorrect.
12 Of course, stabilizing axial thrust is again very 13-important because it just means no failure or much less 4
14J chance to fail.
-15^
And then the last one is, of course, a separate.
16 These are the discharge of the impeller, the diffuser.
And 17 the fifth one is the low flow condition promotes suction 18 turbulence which aids mixing between hydrogen and water.
19 MR. JONES:
When you say low flow conditions, what 20' are you saying?
21 HR. KCKAY:
Let me put the Palo Verde pump.
22 I design a pump for BEP, best efficiency point.
23 And we just call it 100 percent from flow.
And there is a 24 certain amount of recirculation.
Recirculation so that you 25 permit because each one has its own limitations, how far you Beritage Reporting Corporation (202) 628-4888
'~
-4,
,9 70' L
- 'l' may go for. reliable, good operation.
So, we do specify a-121 certain amount.
In this case a impeller 200 to 220 gallon E per minute.-
14 Low flow would be if the' design is hydraulically.
5 unstable then wherever that dip in the curve begins, that's 6
part of the low flow.
If you eliminate it, low flow.is 7:
where we approach recirc flow or where the vibration level-8 or pressure pulsation begins to go up..
9 So, that means 100 percent here, say, for larger 10 boiler feed pumps, you're going to go below 25 percent.
11
' Some of the pumps in connection with your coolant in 8804, 12 you are finding some pumps with very little,'very small
'13 recirculation. flow.
That's what I mean by low flow.
So.
c 14' Lthe. low percent flow.
15 MR. WEISMAN:
The conditions we're talki.ng about 16' the pump handling.
One, you've got to worry about the 17-preferential location of the gas in the pump.
You have to 18
' worry'about dynamic forces because it's not going to be a 19 steady load.
So, you worry about that affect on the pump.
20 You also worry about that effect on the clapping.
And 21 operability means once the vent flow, the pump, the piping, 22
.the whole arrangement is still intact and can perform 23 reasonable to what it should perform.
.24 I have never heard any discussion of what you get 25 here was a two phase flow as opposed to all the water -- I Beritage Reporting Corporation (202) 628-4888
'O
- e 71 I
haven't heard any discussion on the dynamic forces on the 2
other piping.
Would you care to address any of that?
3 MR. KCKAY:
Well, the two examples I am showing 4
here, they are addressing the first one.
And my 5
instrumentation chart and my vibration charts shows that 6
very carefully I set a certain length limit, what my 7
judgment at that point'I don't damage the pump yet, the 8
dynamic forces are not big enough for the rotor to -- enough 9
to create forces.
And that I measured for five locations.
10 One axial, two on the two ends.
And that is an indirect, 11 the amount of vibration I will create in the pump by doing 12 this test for directly reflecting it to the dynamic forces.
13 HR. WEISMAN:
I have been involved with pump testa 14 where you had simply two phase mix, you go into the pump.
15 And the load on the top can be rather large.
I j
16 And we saw large p4. ping moving --
~ 17 MR. KCKAY:
I think you are talking about small 18 pumps.
19 MR. WEISMAN:
What I'm talking about here is the 20 case of a surging flow causing the problem.
Not the 21 internal recirculation.
Not a low flow.
But the fact that f
22 the flow was not steady and the pump test that was running, 23 the pump was acting somewhat like a check valve and you have 24
-- it was a way for a pump to fill out a circuit of water 25 and then wait for a while longer for the pump to fill up.
Heritage Reporting Corporation (202) 628-4888 m
L___---__------------------
~
r; j
1
- =
l l
72
)
s 1-So, it was'a very dynamic surging situation.
2 My problem is I don't have the confidence that the 3-pump'would operate.
I'm looking for a way to have 4
confidence it will operate not just we think it will operate i
5 because we're not -- in an accident, whether or not it will 6
operate, we want to have a high probability.it will operate.
7 MR. KCKAY:
See, at this point, I've broken the 8
head down almost completely.
That is a very poor two-phase 9
full condition.
And the vibration, that's the one.
Now, I 10 do lots of testing out there.
I didn't bring more than two 11 examples with me.
12 The other one was very dramatic because that's a I
13 big power station and the startup pump was lost completely.
14-And I was next to it and I didn't have to run away.
C 15' I have seen many of them where I was scared to 16 death and I wished my-knees wouldn't be stiff and I could 17 run away.
18 But we are talking about the design condition here 19 internally where we are minimizing these instabilities.
20 MR. WEISMAN:
I am trying to relate it back to the 21 Farley pump.
Alternatively, based upon the theory of a high 22 confidence it will work, why run the test because you'd be 23 confident you're not going to tear the pump up.
24 MR. WOODWARD:
I understand what you're saying.
25 There are several factors that come to mind immediately.
Beritage Reporting Corporation (202) 628-4888
e:
73 s
1 The.first two have to do with the examples he gave.
He did
.2 not directly state that the pump was damaged.
3 He stated the pump was not damaged.
He did not 4
comment on the suction piping.-
In the two examples you 5
gave, did you know of damage to the auction piping?
6-He indicates no.
7 With regard to the Farley Plant, itself, three 8
things come to mind.
One of the main points Dr. McKay has 9
tried to point out is the fact that when you make the 10 modifications internal to the pump, you improve the
{
11 stability of the pump at these low flow conditions, i.e.,
12 less Pibration.
Okay?
l 13 Secondly, both -- well, secondly and thirdly,'both units have experienced cavitation of the pumps during 14 15 operation as a result of gas accumulation in the suction 16 piping.
I refer to that as a separate problem when we first 17 started, which is often confused with this problem we are 18 addressing.
19 But, basically, the effect is the same when you 20 get down to the suction.
Because if you are running a pump 21 with a two phase flow, and we have not observed any piping 22 damage in any of those instances where we had gas 23 accumulations in the suction and then started the pump and 24 the amps -- just like you would expect.
You know, the amps l
25 doesn't just drop -- they're not just dropped to zero and Heritage Reporting Corporation (202) 628-4888
a -
74 p
1 sit there.
They move around.
And they move around 2
reflecting a slug of water and a little gas and a little 3
water of varying portions, meaning non-homogenous flow which 4
reflects the situation you're talking ebout.
5 So, we actually have -- both of the units at 6
Farley have experienced this as a result of gas in the 7
suction piping and starting the pumps that would have some 8
two phase flow.
9 MR. KCKAY:
Every pump manufacturer we test we 10 test every new pump for practically every pump we send out 11 to power stations for MPSH at various points.
12 Like this qualifies as a MPSH.
13 We go all the way back here and do the same.
Low 14 flow to very high flow.
15 My test of the Sacramento pumps which is at Rancho 16 Seco, we had three pumps and we did this.
I have three test 17 pumps:
1, pump 2.
They broke down differently, but we 18 broke them all completely down. And they are very similar 19 pumps and the way we proved that we didn't damage the pump, 20 come back to these two points and retested them for 21 performance.
22 I see -- doing troubleshooting, obviously, I have 23 to see lots of damages.
And I could show you horror 24 pictures.
But what I am trying to say in this case is that 25 the' conditions are such that the pump -- can I have the next Heritage Reporting Corporation (202) 628-4888
]
(
i j
75
)
w 1
slide?'
i
)
2 MR. MERSCHOFF; Can I ask you, the one case you 3
mentioned about thrust oscillations?
4 MR. MAKAY:
Yes?
i i
l 5
MR. MERSCHOFF:
The Farley configuration is
-6 different than any I've heard described so far and that, as 7
Mayne said, the not positive suction head is being driven by 8
another pump.
We've heard explanations that it's not a 9
single slug of gas.
Rather, you have slugs of water i
10 impinging on the suction side of this pump in some 11 periodicity.
Which would in intuitively induce thrust 12 oscillations.
And I haven't heard anything yet address 13 that.
14 MR. HAKAY:
Yes, but I don't think it will go that 15 way.
Because the mixing action of first station is rather 16 high, it's a very strong action.
So I don't think it will 17 oscillate.
And then the damping added by the proper 18 geometry at the exits and the detaining liquid on the side 19 plates, that's very important.
20 MR. MERSCHOFF:
But is my conclusion correct that 21 none of the examples presented to date would model the 22 actual Farley situation in that the net positive suction 23 head two phase flow is being driven by an outside source?
24 MR. MAKAY:
Well, the second one I showed you 25 where the head breaks down, the reason for the pump to Beritage Reporting Corporation (202) 628-4888
p e
[
76 I
restart because from the *, liquid is doing the same thing, 2
the
- is coming down.
3 MR. WOODARD:
All our B pumps are typically driven 4
by' condensate pumps or condensate booster pumps.
5 MR. MAKAY:
And the hydroganerator, it's a 6
combination.
Anyhow, you can discuss the subject.
This is 7
my opinion based on my many years of-design testing and 8
field trouble shooting, it is my conclusion that the Farley 9
A Train would have operated without failure for the duration 10 of the subject condition.
That's my conclusion.
11 Yes, ma' am?
12 M3. ADENSAM:
Perhaps on this at this point, I 13 heard you say earlier that these conditions would be 14 acceptable in terms of cooling and lubrication as long as 15 these modifications of the proper dimensions had been made.
16 Have the Farley pumps been modifiedN Were they modified?
17 MR. MAKAY:
They were designed this way.
- Because, 18 you see, when I was sitting in on the design review i
19 meetings, this was after the Connecticut Yankee.
The 20 Connecticut Yankee already -- by that time, and I was part 21 of the review board that set down the specific and hand 22 picked dimensions.
And those dimensions were picked 23 purposely.
Now, Bill and myself, we took on -- about a year 24 and a half ago -- about '86, we purposely went down to the 25 Westinghouse shop just to make sure that they made all the Beritage Reporting Corporation (202) 628-4888
e
\\..
'.t 77 1
particular dimensions and checked that particular road work.
2 If there is such a thing as a pump failed, broke, I don't 3
know what they put back-in.
But we purposely made a trip 4
from Philadelphia,'and some from Alabama,-and we did a 5
thorough inspection to the components.
6 Thank you for your patience.
7 HR. WOODARD:
Just prior to giving the conclusion, 8
I would like to point out something in the NRC letter to us 9
of March 28, 1989 that is incorrect, significantly 10 incorrect.
In the Appendix on page 5, at the bottom of the 11 page, it says, "further, once the problem was discovered by 12 the licensee on Unit 1 on February 6, 1988, the licensee.did 13 not check Unit 2 for the same condition until February 25.
14 Therefore, despite the fact that the licensee did ultimate'ly 15 discover and report the problem, no mitigation for 16 identification and reporting is warranted."
17 The problem was not discovered on February 6, 18 1988.
The problem was in fact discovered as a problem on 19 March lat.
And looking back, the first symptoms we saw on 20 it were February 26th.
21 MR. HERSCHOFF:
If I can add, the inspection 22 report reflects that date as well, that is wrong in there, 23 is a typo.
It should have been the 26th.
24 MR. WOODARD:
In conclusion, gas accumulation is 25 limited to one train of the ECCS on each unit, and is i
Beritage Reporting Corporation (202) 628-4888
p 78 L
l' further limited to the recirculation phase of safety 2
injection.
Westinghouse concludes that the hydrogen gas-3 would arrive at the suction side of the charging pump in a 4
gas water mixture under the assumed conditions.
5 Dr. Makay has stated that the charging pump 6
continues to operate during the postulated event, based on 7
adequate lubrication and cooling being maintained in the 8
pump internals until normal flow is reestablished.
9 Therefore, the ECCS, the entire ECCS will remain operable 10 during the postulated event.
11 Mr. Taylor, that concludes our presentation.
12 MR. LIEBERMAN:
Thank you.
13.
MR. JONES:
Can I ask one more question.
14 One thing that's not been addressed, if we assume 15 that the pumps will continue to operate, and Dr. Makay has 16 said that the pumps will probably, there may be a chance 17 that the pumps will be oscillating in the flow, that the 18 amps of the pumps could be going up and down.
What is the 19 operator going to do, what is he instructed to do, limiting 20 precautions from the pump if he's running in this mode and 21 the pump has large vibrations indicating large amperage 22 problems, will he be instructed by his procedures to turn 25 off the pumps like he did when he was doing his testing 24 about a year earlier on the 1A pump when he had amperage 25 oscillations during the testing?
Heritage Reporting Corporation (202) 628-4888
r he, c f
3 79-s c
1-HR. WOODARD:
We actually asked operators'this as,
~
- 2-a result of this incident, because it was one of the 3
questions-.that came to our mind.. As I. remember, the i -
4
. overwhelming majority of them said, if I.were.in normal.
5
-condition starting up a pump say from a refueling, a non-l
'6
' accident. condition, and I started the pump and.the amps 7
started to oscillate, I'd shut it off and go report it.
8 They said, however, if I was in an accident 9
condition, I wouldn't touch it.
10 MR. MCDONALD:
Jim, I think this is an important 11 point from what we have as detailed.
I'think somehow'you 12 should agree with the analysis of Dr. Makay that this needs' e-13' to be told to people, because we're under the strong: opinion 14 that it would be a mistake to turn that pump off in.an-15 accident situation.
16 MR. LIEBERMAN:
Well, we thank you very much and 17 we will' consider this information and be back in touch.
1 18 MR. WOODARD:
Hold up, please.
19 MR. MCDONALD:
I might have made -- once I got a 20 prompt here.
We plugged the first rows in each unit of our 21 steam _ generators, not two rows.
And I have people back here 22 checking what I said, 23 We want to thank you.
It's been a long 24 presentation and we've taken you a lot of details.
But to 25 each and every one of you, we appreciate your patience and Beritage Reporting Corporation (202) 628-4888 I.a
f'.
o.
,7 80
^
- 4, 1
the opportunity to talk with.you.
2 MR. TAYLOR:
Well, it's been interesting. Thank 3-you very much.
4 (whereupon,'at 2:25 p.m., the meeting was 5
adjourned.)
7 8
I 9
10' 11
(
12 13 14 15 16 17 18 19 20 21 22 23 i
24 l
25 Heritage Reporting Corporation (202) 628-4888 1.
.R <
- 4; o
- .
y 9 %
1 CERTIFICATE 2
3 This is to certify that.the attached proceedings before the 4
United States Nuclear Regulatory Commission in the matter
-5 of:
AIABAPA PCHER CO@ANY (APC) 6' Names
. g g a Gas Intrusion Event at Farley Unit 1 7-8 Docket Number:
88-113 9
Placet Wile, Maryland 10 Date:
May 10, 1989 11 were held as herein appears, and that this is the original 12 transcript thereof for the fl.le of the United States Nuclear
'f3 Regulatory Commission taken stenographically by me and, 14 thereafter reduced to typewriting by me or under the 15 direction of the court reporting company,.and that the 16 transcript is a true and accurate record of the foregoing 17 proceedings.
18
/s/
19 (Signature typed:}
Peter M. Dixon 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 Heritage Reporting Corporation (202) 628-4888
-