ML19308B854
| ML19308B854 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/23/1979 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001170447 | |
| Download: ML19308B854 (32) | |
Text
. _ _. _ _ _ _ _ _ _
t.
3 July 1979 Calle([foiiiiTimiiisiil(IE:RIII),,about his memo expressing concerns about the ranie' an'd^ sided oflyecorderj;[for post-accident monitoring.
-- His concern was that low range temperature was not sufficiently low to ensure compliance w/ temp / pressure limits.
-- Also concerned that the high range of some instruments did not comply with RG 1.97.
-- Matter was referred to NRR.
-- NRR response was that the matter would be handled as part of the generic resolution of implementation of RG 1.97 (i.e.,
Task Action Plan A-34).
b S001170
.i
6 July 1979
.....~
.. _. ~
.. Talked,to Richard _Hartfieldlabout the,! Nuclear Plant Reliability Data;
' ' ~ ~ '
(Syst_em!.
-- It's a voluntary system.
There is considerable inconsistency in what is a safety system and what is a failure. As pesult, it is about worthless.
Major emphasis is component failure. There doesn't appear to be much of an effort to assess scenarios.
L J
11 July 1979 Talked to"i [ Q T @ [
d Roger Floyd in his testimony to the IE interviewer said that on at least two instances they had a reactor trip that ended up with a dry pressurizar.
-- He thinks that at TMI they shut off the RCPs because they were i
only drawing about 100 amps (normally draw about 600 amps) which would indicate that they were completely steambound.
I l
j i
a J
f l
r i
i i
1 4
-n-r--
~e-
, +,
s,
.---.m e,e-p
--, -, ~..
v-~~,
g y,,.u,
-,g
s D'
23 July 1979
- <.. ~. - _. _,
CallediChip.Perkinslat>I, Conn;Yankeel;re any documents concerning 1968 inci-dent at Conn Yankee. He'll see what he can find and call back.
s 20 June 1979 Called [@[Ifr~amil(RSB:NRR) concerning the[INNi/EaEliiM
- He's not sure why he wrote the memo but he thinks it's a com-bination of the Davis Besse incident, the draft Michelsol report that he had received from Ebersole,and the Pebble Springs response to Ebersole's question (NUREG-0560 App W).
- He wrote the memo as a matter of technical interest w/o identifying it with a scenario. He couldn't at the time figure out how the pzr would cool off fast enough to be a problem.
Didn't think about the PORV sticking.
- The manometer problem may not be applicable to TMI because the seal could have been blown out as the water or steam moved into or out of the pzr. ;Iowever, it's possible that steam went throu the surge line and thru the water and out the PORV w/o affecting the prz level.
- He found out aoout the Davis Besse incident at a meeting set up by IE.
The meeting was for information (IE/NRR have such meetings 2-4 times a year) concerning the DB incident. The major emphasis was on the quench tank rupture disk failure and the damage.to insulation. The emphasis seemed to be more related to system interaction than any specific problem. He doesn't remember who set up the meeting.
- Jerry Mazetis may have been involved and may have visited the site.
- He's not sure why he didn't make a bigger point of it (eg refer to operating reactors).
He wasn't that concerned about it and he assumed (w/o any basis) that the operators know about the problem. They had duccessfully handled it at DB.
4 '%
a i
5 July 1979 o-u---
-- s,,,,,,
Called SanfortC[stee))with Bill Parler m
Parler was at a meeting with Israel, Novak and Denwood Ross (NRR) j corcerning some questions that Wayne Lanning had about Davis Besse. Parler understood Israel to say that there were a lot l
more relevant issues, including menos between Israel and Novak, concerning Davis Besse.
The impression was that these were j
issues that had not previously been brought to light.
Israel claims he didn't say any such thing. He says that Ross may have said something in the context that there were more.important issues than the ones that Lanning was raising.
t
-- Israel says that the only memo he wrote about Davis Besse is the now famous Israel /Novak memo.
Parler and Lanning both agree that Israel was quite clear
.in his statement and implication.
l t
22 June 1979 Calledig@Mif. IE:RV (Formerly IE:RI)
- Called him about the March 31, 1978 memo ![ENhkN) where he raised a concern that the PORV failed open on loss of control power.
- His major concern was that the valve had no indicat on so it's possible to have an unannunciated electrically initiated LOCA.
This LOCA only required a single failure of one inverter.
- NOTE: The valve doesn't fail e,en on loss of power.
The bistable power comes from the invertar and deenergizes to open the valve.
The power fc the valve comes from o separate power supply.
- NOTE: There's no way to install direct reading position indication i
on the valve because it's not a rising stem valve.
- He was concerned that they'd have a LOCA w/o recognizing it. He didn't go far enough in his analysis of the incident to ret ]nize the potential problem with pressurizer level.
- He was also concerned that they'd had several inadvertant Na0H in-jections and it was. causing chemistry / chemical problems in the core.
- Sternberg says that it was policy in IE not to use inflamatory words in correspondence with IE:HQ.
- he felt that he worded the memo as strongly as he could considering the problems between IE: Region and IE:HQ.
- he orginally wanted to say that his concern was an "un-annunciated LOCA." But he took it out because it was too infl amatory.
- He originally recommended that the matter be transferred to NRR but took it out because he was concerned about causing problems b/ prempting IE:HQ's perogative to decide what issues sho:ld be referred to NRR.
- He wasn'* particularly impressed with the response from IE:HQ. He feld that someone took a little time to look into it, found out it was only a 21/2 inch. valve with a blocking valve and let it go at that.
By the time he received the response (May 1978) he was too busy with other things to press the issue.
i
7 e
3 July 1979 Call from[ h [ M d(IE:HQ)
--. I called him earlier to ask him about his response to the"$Wilihg~)
,'@ He had said that he wanted to consult his files fiFit7^^-'
-- Now he says he'd rather send me what's in his file and discuss the matter after I've taken a look at it.
I
6 July 1979 rm nn~wnerm CalledBegerJWoodruff1(IE:HQ)
.-m.m-,
,.,w He decided the[Sternbeirg,memoiwasn't a problem because the licensee had assessed the problem in the SAR.
He didn't realize at the time that the indication was,on the solonoid and not the stem..1 Dov'1 wu. t w_
T o G e-Ecw; eo tam us2cc m9a ', teca (:.rn Not t,n N cc ro N M t T)co.
-- They considered sending a Current Event (sent by MPA) or a gircular (sent by IE). They would have described the incident, warned licensees to be prepared to take action and/or recommended hardware or procedure changes. They didn't send a Circular (he's fairly sure MPA didn't send a Current Event) because they felt the issue was already adequately covered.
F 9
m
,.n..
m---y,
I 26 June 1979 (R85(f3nnjs(B&W) returned my call to Jim Taylo (B&W). concerning the letter to Davis Besse concerning t rc 261978'Raricho,Seco*-
incident, ar.d thelDi6$F 83i _ R W -
~ ~ ~
The Rancho Seco transient occurred March 20, 1978 and TMI had a similar transient on April 23, 1978.
There was considerable correspondence with each utility on each transietn.
(He'll send copies to me.)
Sent letterss to all utilities, except Met Ed concerning the Rancho Seco transient. Decided not to send it to Met Ed because the material had already been discussed with them as a result of the analysis of the April 23, 1978 transient.
It was a conscious decision not to send it.
The specific fault (i.e., the dropped light bulb) at Rancho Seco was not applicable to TMI because TMI had a different " generation" of non-nuclear instrumentation A similar letter was not sent out as a result of the Davis Besse Sept. 24, 1977 incident.
The incident was considered to be plant unique because:
1.
The Secondary Feed Rupture Control System (SFRCS) which started the transient was unique to Davis Besse.
2.
The PORV cycling was due to a missing relay.
~
3.
The PORY at Davis Besse was made by a different manufacturer than at any other plant.
They didn't analyze the scenario to see if a similar scenario could have occurred at other B&W plants.
21 June 1979 Called.WQisl(NRR)
- He lead a team that went to DiilEBe5k'~ to investigate the 1
{ % @ }p3]SFJJ, incident.
' ~ ~ ~ ~ " ' ~
- He briefed Mattson and Seyfrit (IE) when he returned and it was decided that Seyfrit would follow up on the investigation.
He doesn't kntw-what eventually came of Seyfrit's work.
- Thadani has Mazetis background file from the trip.
- Nazetis attended a meeting on September 28, 1977 that may have been the briefinj that Israel remembers. Bill Little and Terry Harpster may ha'e been the ones who set up the meeting.
CalledM1MTNM(Section Chief, Nuclear Support 21 June 1979 Section, IE:RIII)
.T..Rarpstar-, works,.for.him._,Harpster conducted the investigation of the Bev6stasseeSeptesten34G197711ncident. The rosults are in Inspection
~
,Report dated November 22, 1977 (50-346/77-32). As far as Little knows, no one else in NRC investigated the incident. He believes that Creswell's concerns that later surfaced in his board notification may have started with this incident.
i I
L e
,s f
-g--
y m
a g m
22.. June 1979 Called $5 E Er1Ta
%rf,ll,9.77}inciderit)7 " R' (SRO aE6'aE85&ielduring the
' ~ ~ ~ ' ~ ~
- He didn't realize initially that the pressure decrease was due to a leak because 1) the pressurizer was full, 2) they were having problems with the feed system and he thought they'd overfeed (too fast) the steam generators.
- He realized they had a leak when they got a high containment pressure alarm. Then the fact that the quench tank rupture disc had blown caused him to realize that the PORV was stuck. They didn't have in-dication of the PORV and the tailpipe alarm was waiting to print out.
It prints out on the computer only. He was also confused because they had been operating with an elevated Tave for a test.
Didn't initially realize that the rupture disk blowing meant the PORV was stuck because it had blown before.
- He knew during the transient that he had reached Psat.
They had a Psat/Tsat table under the plexiglass on the operators desk. He also realized that they had voiding and that was causing an insurge. He realized this because it had happened to him before. During hot function testing (HFT) they were at very low Temp and Pressure. They started to spray the pressurizer.
The hot water went out the surge line, collected in the hot leg candy cane (no RCPs runneing) and drew a bubble which caused an insurge. B&W had a precaution in one of the HFT procedures warning about this during cold testing.
- About one year after the incident they got a letter from IE:RIII saying that they shouldn't secure HPI because it couldn't do any harm and it might do some good.
Derivam had secured HPI because he was concerned about going solid.
Looking back he can't understand why everyone's so concerned about going solid.
- He doesn't think that his training prepared him for identifying a small leak. Prz level was always the primary indication of a leak.
- He feels that this was the most difficult incident he'd ever seen.
It's the only "real incident" he'd ever been involved with.
- He believes it would have been alot easier if they'd had high speed recorders.
- The other operators were:
R0 - Ed Knight (he's now working at a plant under construction in Michigan (Consumer Power?))#
R0 - Roger Breubaker Operations Engineer - Bob Zeminski arrived within about 5 min.
Assistant Station Engineer - Terry Murray i
1
1 1
22 June 1979 Called [NBEdiirfj(NRR) i He has some of Mazetis material from the[..m _,,~,--~;.l He'll Oswis)Sessejincident send a copy to me.
i
- He considers the Davis Besse incident to be a precursor.
i Similarity:
1.
PORV stuck i
2.
Aux feed problem Differences : 1.
Low power history and power level 4
2.
HPI wasn't secured 3.
Aux feed wasn't delayed so long i
il f
f l
4 A
+
c -..-
22 June 1979
...c~~.
Called K nSeyfritl(IE:HQ) about the; Davis Besse Sept._...
4
.24, 1977L ncident.
i The inspector wrote a memo that was forwarded to Seyfrit who consulted with f1RR. NRR concluded that changes in tech specs were not required. He thinks the memo probably came as a result of the inspection / investigation.
-- He doesn't remember the Mazetis investigation or the meeting with Mazetis and flattson.
He doesn't know of any other investigations besides the one that would have been done by the Region.
-- It was common practice to have briefings of NRR concerning significant incidents.
If a briefing was held the LPM (Leon Engle) should have prepared a summary.
2 July 1979 Called 55Ee70EI55d($j{@i]e,
The rupture disk had blown during HFT (hot functional testing) while they were lifting the PORV as part of the test. He thinks they may have put too much energy in the tank and that's what caused it to fail. The PORV had lifted during subsequent operations w/o rupturing the disk. This was before they had an OL so they wouldn't have had to submit an LER.
The incident concerning the bubble in the hot leg occurred around Dec. of Jan. before fuel loading.
It was a fairly slow transient.
It was part of the initial drawing of the bubble and RCP NPSH test. He may have made a note in the shift foreman's log but he doesn't know if anyone followed upon it.
The precaution concerning bubbles in the hot leg concerns the minimum A P to allow for the ACheight. He thinks it's still in Plant Limits and Precautions section 110101.
-- He thinks the precaution about securing HPI came in about Sept 20,1978. Call Steve Quenoz to get a copy.
)
4 2 July 1979 Called N M 3 M ( @ {
<or191"e.-o Ottv w ev7 roar 40 yo s
-- He didn't have a copylbut he thinks the Station Superintendent has found a copy. Quenoz will send a copy to me or call me if he can't find it.
1 1
i
3 July 1979 Called [,Je._rry _l W _E IR (NRR)(u m, w,n m ~ m +~;
- q DA$.SidC95CW'4.P_i _a1%
u g,.
What were the "what ifs" in his trip report? higher power, another failure, etc.
-- He assumed someone would follow up on the issues.
f He was assigned to other duties after he made his report to Mattson and Seyfrit.
IE took over the investigation. He gave a copy of the handwritten trip report to Mattson. He may have also given a copy to Seyfrit.
The 9 cycles of the PORV was based on the plot of pressure vs.
time. They may ahve also used the computer printout.
-- He doesn't recall that the trip report says that they secured ECCS @ 4-1/2 min. (TECO says 53 min.). He may have been mistaken.
Numbers came from a group discussion with the plant personnel.
i i
)
3
-c-w n
e-J 5 July 1979 Ca,ledf51Mabout McDennott's review of QA associated with the Ml$f 34F 1977. i nc i den t a t (th...i$_5!$ss_i.s_!.
mu.
--- ~
He doesn't remember anything about it. Recommends contacting McDermott.
'l 4
i 1
1 i
i 4
i i
I I
t t
t 4
f 6
4
1 i
i j
5 July 1979' '
CalledMt(IE:RII.Il about the Sept. 30 Insnediate Action letter I
concerningth&$ep$~.ik'OssW5sise71ncident. He doesn't remember Cill"ToirPi Tainpfing for details.
Immediate action anything about it./n Section.0800 of the IE Manual.
letters are covere They are i
j considered to be mutually agreed upon requirements.
1 4
4 0
ll i
i I
3 i
b 1
l
-,e,--r-
,,,,-,,nw.-+
,-.,s,
.., -, - e prv
,r,
,..w,
--v m - ~,
e
,-n.
,e--
6 July 1979 Called d iEdpliM l(IE:RIII) 6 E E 8 d q1 % }$[IT M
-- He's fairly sure they secured HPI when they initially saw PZR level increasing. He'll check his records to see when it was secured and if it was restarted.
a
-- A few weeks ago he calculated that the inflow was 71300 GPM which he decided was greater than the HPI capacity for the pressure they were at.
Note: HPI capacity is 900 GPM @ 1500 psi but shutoff head is only about 1650 psi.
2 HPI pmpus come on
-- He thinks they started makeup after securing HPI.
I makeup pump =-
75 GPM 2 makeup pumps = 125 GPM
-- Makeup pumps are required to provide seal injection to the RCPs.
However, running RCPs don't leed seal injection if CCW is available. They have an internal flow path and a heat exchanger.
-- He thinks the corrective action was completed and evaluated at the site before returning to power. They were down about 22 days.
-- The supplemental to the LER was a summary of the evaluations prepared at Tampling's request after he had reviewed the required evaluations at the site.
i 1
1
l 10 July 1979 mmmm,
.~,,,.
.s CalledkTom.Tamplin[,i(IE:RIII) iMstiGECCC,$..,<.-__, D~ _12 eAiSRd He is out for about 2-1/2 weeks so I talked to Streeter.
I He is fairly certain there was no separate memo concerning securing HPI (mentioned by Dervan) but he'll check and call back.
There was an unresolved issue in the IE inspection reports (he'll give me the reference) because the operator secured HPI before he really knew what was happening.
i Resolution was a revision to the plant operating procedures telling the operator to have the HPI on until he knew what was happening.
It can't cause an overpressurization because the shut-off head is so low.
1 l
l i
.i i
g-,
r 9
y.,
i t
I l
i 10' July 1979 g e.r - -,
Call fromtStmeetpr,1
,,n e ca find no reference to a letter from IE:RIII toiDe~i >
v
'~
telling them not to secure HPI.
~.s-
-- There is some reference to this subject in i
IE Insp. Report 50-346/78-27 Detail 2 p.3 l
78-30 Detail 2 p.4 1
i l
j 1
l 4
j i
i
- )
1 1
1 i
j i
h a
J e
-+
,r4 g
+-y----
,--.,..,. - = -.. -,,
-e,.,,..,_-.-r_,, ~.,,.
-em_.
~
\\
11 July 1979 Call from [ M ]
-- The April 20, 1978 letter that Derivam recalls may be the forwarding ctio_n.repor,t,78-06 3 ch didn't have anything to letter for i hi t
do with thel ilt421977j2ncidents i
11 July 1979 remer---,~.
Called SobertiMcDemptt}(NRR)
-- He wrot'e another memo (October 20,,1977)-concerning the,QA,_ -
implications of the missing relay (DavislBesse' Sept.i24.?hl_977,1 The utility verified that the relay was there and the valve worked during initial testing.
He and the inspector speculated that the relay (it's a standard relay with thousands of applications in the plant) was pulled out by someone to use somewhere else to get a job done.
1 I
t I
i w
w
.-+-e
,-y-c-
r r
-r y
y
11 July 1979 n--,
Called (Leon,Engle1(NRR)
-- He was the LPM for[6EvTs7iEisO}fo110 wing thbepb3M~5f73 incident.
(It hadn't'bseh~t0Fned over to 00R yet.)
" ~
He talked with the utility immediately and had a meeting at the site a week later. They discussed the scenario and the concenrs. Major concerM1) were PORV failure, rupture disc failure causing steam generator damage,-and the feedwater transient.
-- The licensee speculated that the relay was cannabilized for another job. However, they couldn't find any specific evidence.
Based on discussions with the utility and with Harpster (IE:RIII) they (NRR) were satisfied that the problems would be taken care of.
-- On the Monday following the site visit they had a meeting with Seyfrit, and Seyfrit agreed that IE would retain the lead. He doesn't recall any briefings by IE.
Israel may have attended this meeting. Mazetis gave the briefing. A summary of the meeting and/or the site visit was not prepared.
-- As far as he knows, Harpster's report is the only report of the investigation of the incident. He reviewed the report in detail and agreed with the conclusions.
He also sent a copy to D0R because they (DPM) were about to turn over Davis Besse to DOR.
-- He didn't review the incident for generic implications.
He felt that IE would " tag it" as generic if necessary.
He thinks the operators secured HPI at about 7 minutes but he doesn't agree that the increasing pressurizer level was caused by voiding in the primary. He thinks the HPI caused the insurge.
He thinks (but he's not sure) that B&W said they planned to give a copy of the event report to other B&W utilities.
-- Davis Besse was " infamous" for the number of LERs they submitted.
Engle looked into it at the request of the ACRS and concluded that they were overzealous in reporting and they were having a lot of problems with the secondary system. Maybe if there had been fewer LER's they'd have received more analysis. Unlikely.
-- Engle was very impressed with the operators. They were very calm and appeared to have had the situation under control. The operators told him that they were concerned about the RCPs because of steam bubbles in the RCS but kept the pumps running to keep the bubbles from collecting.
12 July 1979 Called @d%Mabout the Evi55N$$[ikiM(!;
- He gave the information to Tampling who wrote the report.
- He does not remeser many of the details of the event.
- He considers the major differences between the Davis Besse incident and the TMI accident to be:
1.
They closed the PORV sooner 2.
They only lost flow from one aux feed pump 3.
They didn't secure RCPs
- He was very impressed with the quality of the operators.
- He says that the operators got a containment humidity alarm in the area near the quench tank and that's what brought their attention back to the water coming out of the quench tank.
From that they figured out that the PORV was stuck.
- He's skeptical tnat the operators realized during the transient that they had two phase flow in the pumps or boiling in the primary. He and the TEC0 engineers didn't fully realize it after reconstructing the event.
- Creswell or Tampling has his notes on the investigation.
- He doesn't know if any subsequent actions based on his inspection i
report.
- He doesn't recall ever seeing Mazetis' trip report.
l 1
..r
- n..,,..mne.-,,, nc 12 July 1979 Cal l e d4NHieJ Berivanl( Davi s. Sasse)1
- He secured HPI after about 41/2 minutes because pzr level was increasing. Since pzr level kept increasing after HPI was secured he thinks that the increasing level was due to voiding in the RCS and the HPI.
- The September 20, 1977 ltr from IE that he referred to earlier may have been Teco notes concerning an exit interview.
The notes would have been dated about September 20.
He's not sure when the interview would have actually been held. Quenoz would have the exit interview
- notes, i
1, 1
17 July 1979 Called 7k@j(IE:RIII)
- In Creswell's inspection report 50-346/78-27 he discusses his concern that HPI was secured and SFAS blo.cked before discovery of the cause of the problem during theiSeptember 24D1977fincident atlBad8esg*[ He notes that the lidensee~sgreid'!td' review the issue.
- Teco concluded that the HPI shouldn't have been secured.
They didn't submit a formal report but they changed the procedure to include a warning (EP 1202.06).
- They didn't consider informing other utilities or B&W because they assumed that other utilities would have trained their people not to secure safety systems until they identified the problem.
He thought that Teco was an inexperienced operator (ie licensee) that didn't know how to write procedures.
I pointed out that the procedure didn't say to shut off the pumps, but it didn't say not to.
If the operators in the control room (which he admitted were above average) made this error wouldn't it be reasonable to assume that other operators elsewhere might make the saae error and therefore the precaution should have been sent to everyone. Streeter didn't agree, he felt and still feels that other operators should have known better (ie been trained to know better).
This doesn't seem to be consistent logic but I couldn't get Streeter to agree that it was inconsistent.
18 July 1979 l
Called M 1 Q M 3[E is'5Tito get some documents.-
- --- He wasn't in so they referred me to Jack Lingenfelter.
He'lk sendJne_.a copy of their Reactor Trip Report file for the
{
b.Mffi2}97JJincident and EP 1202.06.
i i
He'll look through CP 1101.01 to see if there are any precautions about pressurizer level and bubbles in the primary. He'll send anything that looks interesting.
+
8 I
4
't U
4 i
1 4
1 1
1 a
i-l i
9
)
L Am--
L-
23 July 1979
,...s Call from! Jack)Ljngenfelter (Davis..Besse)'. He'll send a copy of EP 1202.06~~ kith precaution re HPI. He talked to Derivam and he
~
can't find anything in DP 1101.01 that looks relevant.