ML19306G616
| ML19306G616 | |
| Person / Time | |
|---|---|
| Site: | McGuire  |
| Issue date: | 11/28/1984 |
| From: | James Keppler NRC Office of Inspection & Enforcement (IE Region III) |
| To: | Harold Denton, Heltemes C NRC Office for Analysis & Evaluation of Operational Data (AEOD), Office of Nuclear Reactor Regulation |
| References | |
| NRC-2022-00032, TAC 57211, TAC 57212 NUDOCS 8803010173 | |
| Download: ML19306G616 (3) | |
Text
-
J P &fc UNITED STATES oq[C
' NUCLEAR REGULATORY COMMisslON
' g g
y,.
REGION til j
ne noostvsty mean g (p coca cu.vu, stune s som KC
/ / 3g 1\\ft November 28, 1984 MEMORANDUM FOR: 41.R.Denton, Director, OfficeofNuc15EReactor~
C
'.. Regulation " '
C. J. Heltemes, Jr., Director, Office for the Analysis pM i
and Evaluation of Operational Data
/y y FROM:
James G. Keppler, Regional Administrator, Region III
SUBJECT:
FAILURES IN THE UPPER HEAD INJECTION SYSTEM
/i i
i On November 2, 1984, Region II reported a problem with the upper head
/
injection system at the McGuire Unit I facility (Reference PNO-II-84-81
/
_ ;--, ).
+ ni
..]
i A-1On: November.17 an individual wiio harasked to remliin anonymous visitsd-sy
' ' l' i office and exp'ressed his concern that NRC does,npt lippreciate fully "t_hfe. safe.ty~~ /~_
sigiiificance of this type of event. He showed me a,cemoran.Ifurn which he had
/
~~ j (prepared prior to the McGuire event and came to me liecause WbE11 eves his p
companypes not been responsive to his' concerns."' I.am passing m a copy of f
'thisJoemorandum to you, with the individual's permission, foy.phatever action
~"
(
you deem appropriate (Enclosure 2).
\\
j k
1 5
James G. Kepp er Regional Ariministrator I
Enciosures: As Statea (2) cc v/ enclosure:
J. F. O'Reilly, RII e
e.*
4 j
3 I
i 8803010173 841120 CF ADDCK 050J0369 CF t
a -
r
l
?'~,
Waf. i ?,
i October 3, 1984
SUBJECT:
UHI - Ultra Hich Risk Re:
AEOD/E404, Failures in the Upper Head Injection System,
~~
February 28, 1984
.. ' ~
, g_
~ AEOD dbEs not 'r:.cognizb the 56tiousness of the events described inYabove J P.,
n- -
report. For ExamplerAE0D does not recognize that;;plaiit operation with the-UHI~ accumulator water saturated with dissolved :nitp.p. gen increases thi chahc'e"~"
T c-of iustained core uncevery during some a:cidents. *
- ~ - "
~
~
" -The _AEDD report includes the operating experience wTth brok,eiTupture.
+
membranes:
~
l "Both Sequoyah and McGuire c:perienced broken rupture membranes.
i Sequoyah operated for atmost a year after the membnne ruptured thinking that the nitrogen concent n tion in the a1ter accumu2ator vas within the Technica1 Specification timit. Upon rec = amination l
cn! consultation with other.:tilities it u1s found that the test to l'
determine dissolved nitrogen concentration was inaccurate. The c=use for the ruptured membmne was undetermined. The membmne u1s repasred and the testing procedures vere modified. The me:brane at
.i McGuire u1e ruptured when a section in the Line a1s not property fi11ed and vented."
j However, AEOD completely overlooks the implications of this long term opera-l tion at power with broken rupture membranes and concludes:
1 i
?
"Most of the problems appear to be norm 11 difficulties assoc.icted with starting up a neu unit and ine:perience with the UHI ~syttem...
j Our analysis of the causes and consequences.of the reported. events involving UNI did not identify any safety concerns regarding.the i
opembility of UHJ."
7 3
j s
i a
According to the Sequoyah FSAR, the amount of water injected by UH1 during a large LOCA is 1500 cubic feet.
Injection ceases at about 500 psi, but this injected fluid is subsequently depressurized as the accident progresses. The amount of nitrogen then released from this water (if nitrogen saturated) is about 1800 cubic feet. The volume of the upper head is 800 cubic feet, the volume of the reactor vessel is 4000 cubic feet. According to Amendment 61 of the Sequoyah FSAR only 80 cubic feet of nitrogen would be expected:
"Considering an amount of nitrogen introduced from the UBI eystem has a volume of Less than 80 cubic feet, this amount of nitrogen vouTd represent tese than 10 percent of the upperhead volume, Lese than 2 percent of the reactor vesect votume, and tese than 1 percent of the ICS volume."
The ultimate nightmare with UHI would be failure of the isolation valves to close when necessary. This would lead to injection of the nitrogen from the. _
m-1800 cubic foot nitrogen accumulator. The storage pressure...of 1350. psi.mean_s../
i-the expanded volume would be in the range of 100m 00 to 200,000 cubic feet.:.:.
0
- + ~
-depending on the terminal pressure.
~~
m 2
~_
_ The-UH1 system was a last minute add-on to several; ice _ con (efspr plants fol-lowing pe imposition of Appendix X "conservat4ms" during the early 1970s.
The subsequent integrated s stem development and tire reduction 7o practice of m
UHI appears to have been relegated to the operating sites. apd' cbntinues while
'the plants operate at power. Considering the poor performance of the level detectors, the complexity and interdependency of the UH1 instrument valving and sampling arrangements, the history of long term operation at power with broken rupture membranes, the excessive dependence on field experience in developing operating procedures and training materials, the cramped working spaces, the likely distraction of plant management by NRC inspectors and other NRC inputs, the encouragement of complacency toward UHI by an AE00 report which describes risk-amplifying conditions as " normal dif ficulties", and the chance of other weaknesses which have.not yet been detected; the likelihood of coremelt, containment failure, and reie6se of fission product!nunder rela-tively non-humid conditions may be greater than our experts calculate.
~ In addition to AEOD, ORNL did not recognize any safety concerns with UH1 in its independent review of operating experience which was recently described in the draft report, Survey and Evaluation of System Interaction Events ay Sources.
Furthermore, these possibilities have not been identif ted in PRAs.
Moreover, the NRC safety test programs such as FLECHT, LOFT, Semiscale and the related analysis codes have not been aligned with these UH1 safety concerns.
Finally, the NRC staff in a status report to the ACRS during %ugus't,1976 wrote, "... the staff concludes that dissolved nitrogen wi'il have,' negligible effect on conditions in the reactor vessel during a LOCA."
~
N 1
i
.