ML20198M050
| ML20198M050 | |
| Person / Time | |
|---|---|
| Issue date: | 10/07/1981 |
| From: | Gammil W Office of Nuclear Reactor Regulation |
| To: | Wagner P Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20198L994 | List: |
| References | |
| FOIA-97-455 NUDOCS 9801160177 | |
| Download: ML20198M050 (14) | |
Text
_
Lj.,4{'
e s
.I ,
f* (=
( 4
- c.
9 DISTRIBUTION:
8%entral F11e ETSB Reading File OCT 0 71981. ETSB Subject File 5.4
=l
'amill c aul l
IEtiORANDlPi FCP.: P. C. Wagner, Least Ennineer, ALARA Operating P.eactors Granch 4, DL FROM: W. P. Gamviill, Chief Ef fluent Treatnent Systens Branch, 051 SUNECT: CUP.RthT POSIT 10tl Ott P.ADIOLOGICAL EFFLUENT TECHNICN.
SPECIFICAT10lis (RETS) IllCLUDlWG EXPLOSIVE GAS CONTROLS
- This is in response to your tentenber 10 memo which asked for the latest flRR
- position on RETS prwisions for control of explosive mixtures of waste gases.
Dasically, the position is unchane,ed since our last discussions. That.15, the issue tnust be addressed in the RETS but properly justified deviations from the nodel RETS can be accepted. The explosive gas issue is further discussed in Enclosure 1. Enclosure !! summarizes the current staff position on the PETS.
Ot intn il cisned by:
W1111a.m P. Ca.;. sill Villiam P. Gammill, Chief Efnuent Treavent Systens Branch Division of Systems Integration
Enclosures:
3 1 Explosive Gas issue 11 Sunnary of the Current Staff'
+ Position on RETS g.
cc: R. Mattson ,7
- w. Krc9er /t6- d' utM y1\{Q ' ' '$
T. I ov ak i J. Stoir ; OCT ,a16 W0W F. Congel
- w.d' b
"* ,e #
F. Simpson (EG&G) O' ,.!.
\/,. /d J. Donohrw R. Ireland N.b'/7 9fI S.' Pandey (FRC)
'*. 4. y P. Bangart ,[' '
C. Willis (j
9901160177 990107 PDR FOIA CHAPNAN97-455 PDR ]
od ?N '
wna , .os1.ig.:.E.Tsa. .o.s I
..ss, .. . Os.l.,:RPy
,,,yEK,reE,,,
f ,, ,,,,
.--> .cH.!1.1..!sd.1,t .. .H.P,,, ,,,,,,1,,1,,, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,
em> . ...10../..
-.1.0../.0.5./.8..1.... . . ../. 8..1.... ...10../. ../.8.1.
A FFin al. R E COP O COPY 486//Z677)
- ;l _,
CONTROL OF EXPLOSIVE
- GAS MIXTURES IN WASTE GAS SYSTEMS t
Requirements ,
The model radiological effluent technical specifications (RETS) include requirements for control of explosiv: Jas mixtures in waste gas system (WGS) for both BWRs and PWRs. The RETS requirements basically are for the normal precautions which are included in industrial systems when there is a potential for hydrogen and oxygen mixtures. That is, the concentrations must be monitored and action must be taken to reduce the concentrations when they become too high.
3 Controversy Even though the explosive gas mixture control requirements have been approved in both the RETS and the standard review plan (SRP), some staff members and others continue to argue that these requirements should not be implemented.
The p.rtncipal arguments against implementation are the following:
- 1. This is not an " Appendix I" issue and so should not be included in the RETS.
- 3. The consequences of a waste gas explosion are too small to warrant attention in the Tech Specs.
- The first argument is inealid because the RETS are the Tech Specs on the rad-waste and ef fluent treatment systems and on environmental monitoring. The l
RETS have never been limited to " Appendix I" issues.
t
- The term " explosion" is used here to mean " detonation and/or deflagration".
l l, s 1
l
_ __ ~. _. _. . _ . _ _ . - _ . _ . _ .
> n
\
. 2 The second argument is a cause for concern. The unreliability of hydrogen
., f and oxygen monitors is recognized. In fact, this unreliablity is reflected in the requirement for dual and redundant monitors. The potential impact is in having to take and analyze gas samples every 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> and in having to repair the monitor quickly, within 2 weeks in some cases. While this impct is limited, the staf f is being flexible in its review in order to minimize the potential impact on operations of instrument unreliability.
The third argument, concerning the limited consequence of an of fgas explosion, is discussed in the following section.
Consecuences General Design Criteria (GDC) 60, 61 and 64 require the control of releases, the control of radioactivity and the monitoring of releases. The failure to prevent an of fgas explosion is a violation of all three GDC. On the other hand, an of fgas explosion has little chance of causing a release that would exceed the Part 100 dose criteria (but the same can be said of a core melt accident).
/urthennore, en of fgas explosion probably will not prevent shutdown, interfere with core cooling, or violate containment integrity. The possible ef fects on
! safety systems appear to be plant and accident specific. Actually, the possible and the probable consequences of an of fgas explos,fon have not yet been delineated.
From the viewpoint of of fges explosions, BWRs are very different from PWRs.
l An of fgas explosion in a BWR inherently has a high probablity of occurrence l
i 1
I
r ~
.! but has a reistively low consequence potential. For PWRs the situation is reve rsed. That is, the probability of a PWR of fgas explosion is inherently low but the potentia) consequences are relttively high.
In BWRs, hydrog'en and oxygen are continuously produced in stoichiometric ratio and in considerable quantity (about 70 SCFM/GWt). Without controls, only an ignitten source is needed to produce an explosion. In the '70s, of fgas explo-f sions bNwse endimic to BWRs. The NRC study (Ref.1) reports 29 incidents in 13 plants. BWR of fgas explosions are expected to be of little consequence because continuous release precludes buildup of either explosive gas or radio-active material. This was indeed the case in the explosions of the seventies.
Workers were injured but luckily not killed. A small building was blown up.
Rooms and pieces of equipment were damaged. Power production was reduced and even interrupted. Releases of radioactivity were slightly increased but of fsite doses were not significantly increased. The consequences were far from catastrophic but they were serious enough to justify corrective action.
The long succession of BWR of fgas explosions seems to have been ended by new requirements and correction of WGS augmentation problems. One augment, the recontiner, reduces the probability of an explosion but another augment, the charcoal adsorber beds, increases the possible consequences by collectng the
- radioactivity (see Figures 1 & 2).
In PWRs, on the other hand, the hydrogen is not the result of radiolysis but is deliberately added to the primary system to control oxygen. The hydrogen f
8 l
. - . . - - ._ ... ...- a is transferred to the WGS* through letdown and degassing. This hydrogen is in small quantities (perhaps 20 SCF/ day during normal operation) and is free of oxygen. An explosion is not possible unless oxygen enters the system and usuady oxygen is present only as a result of a failure such as a leak in a compressor. Thus, the probability of a PWR of fgas explosion is inherently l ow.
In PWRs, the waste gases are collected in tanks and these tanks present the potential for explosions more serious than those in BWR WGS. Without con-trols, a tank could butid up a substantial quantity of hydrogen and oxygen.
A 1000 ft 3tank at 100 psia could contain enough hydrogen and air to release the same amount of energy as 160 lbs of TNT. A WGS explosion would not actually be the equivalent of 160 lbs of TNT, but the possible magnitude of such an explosion h.s not yet been established.
. A PWR WGS explosion does appear to have the potential for causing severe damage, including the rupture of the other WGS tanks. The failure of all WGS tanks could result in the release of up to 67,000 Ci Xe-133 (for a 3.5 GWt plant with it failed fuel).
Thus, the possible consequences of WGS explosions seem serious enough to warrant Tech _ Spec controls.
I i
- 5ce Figures 3, 4 and 5.
1, l
5-The San Onof re Deflagration On July 17, 1981, there was a deflagration in a waste gas decay tank at San Onofr 1. The incident has generated international interest and may be reported to Congress as an " abnormal occurence".
The incident is of particular interest in the present context because it is-the only instance we know about of an WGS explosion at a PWR. There have been "close calls" aplenty, where concentrations reached dangerous levels, but in the other cases, careful releases to the atmosphere were ef fected before things blew up. These near-incidents span the period from the present (with Arkansas-One) back to at least 1970 (at Ginna) but generally they were not reported and are not documented.
The incident at San Onofre (as described in Ref. 2-4) started with a "TM1 Action item" requirement to improve the reliability of the irstrument air system.
The instrument air was tied into the nitrogen system. Unfortunately, the check valve between the two systems was improperly installed, so inadvertently air, rather than nitrogen, was being used to dilute waste gas. The San Onofre WGS normally operates in a hydrogen rich mode so the introduction of air quickly produced an unsafe situation. When installed, the WGS included an oxygen monitor to warn of such cunditions, but, in the absence of NRC require-ments, the monitor had fallen into disuse. The WGS included a recombiner designed to remove small quantities of oxygen. The high concentrations of oxygen caused the recombiner to over heat and this is believed to be the
4 ignition source. The deflagration that occurred caused relatively little I
damage and did not propagate to the seven other tanks then containing com-bustible gas mixtures.
The consequences were slight for several reasons that appear to be simple matters of good fortune.
- 1. No one was close enough to the tank to be injured by the blast.
- 2. The explosion did not propagate to the other tanks.
- 3. The release was small because the gas had undergone sufficient decay and was ready to be released; also fuel had been good.
- 4. There was no ef fect on operations because the plant was shut down at the time of the explosion.
The incident does illustrate several points that are of interest here.
- 1. A WGS explosion need not have dire consequences (Murphy's law does not always hold).
- 2. A WGS explosion at a PWR is not so highly improbable as has been suggested.
- 3. Safety features that are not required tend not to be maintained.
- 4. The model RETS requirements (if implemented) are appropriate for l
prevention of such incidents. -
Over all, the San Onofre deflagration suggests that the model RETS prwistons l
for the control of explosive gas mixtures are needed.
i l
.o . . i 0
4 1
Further Study ,, .
Our understanding of WGS explosions seemed inadequate, so a contractor (EG&G ,
Idaho) has been asked to study the problem and to try to provide definitive 4
h answers to our questions.
9-Spekt ffcally the contractor was asked to do the following:-
l Evaluate the problem of potential flammable or explosive gas mixtures in nuclear power plant waste gas systems. Collect, I
l summarize and evaluate existing data of flammable and explosive .
I concentrations of hydrogen, oxygen, nitrogen and other relevant gases. Evaluate the reliability and dependability of commer-4 cially available instruments for montioring hydrogen and oxygen concentrations. Evaluate'the risks (probability and consequences) 4 of gas explosions in reactor ef fluent treatment systems. Evaluate and recommend appropriate control measures. Document this evaluation in a form suitable for pubitcation as a NUREG report.
1 This offort is just starting but the conWactor already seems to have found a reliable hydrogen monitor that could relieve some problems.
Conclusions-It is concluded that WGS explosions can be controlled, that they should be controlled and that the baste-provisions of the model RETS constitute an appro-priate approach.
6-It is further conc 1vded that some detatis of the model RETS requirements may be unduly onerous, espects11y for operating re6ctors and appropriate modt fica-tions should be accepted where justified on a case by case basis.
References
- 1. Lo, R., et al., 'T2chnical Re Water Reactor Offges Systees" ports(Aprt
, Nr,CG-0442 on 1978).
Operating) Experience With B
- 2. Haynes, J. G. (SCE), Letter to Nuclear Regulator / Commission, July 31, 1981.
- 3. Ornstein, H. L. ' Ignition of Caseous Waste Decay Tad at San Onofre 1--
July 17,1981,", Memorandum for Carlyle Michelson, August 6,1981.
- 4. IE Information Notice No. 8127, ' Flammable Gas Mixtures in the Waste Gas Decay Tads in PWR Plants," September 3,1981.
i a
1 i
i l
1
- a 4
STACK RELEASE AL M IN W E ,
MAIN STEAM DELAYP1PE i
CONDEi4SER OFFGAS MEPA -
FtLTER STEAM JET AIR EJECTOR V 1I O DRAIN U DRAIN SUMP SUter i
j- ;
i
- . l
.i t a
i
- Ryerd 8 .* Schematic of a Typical BWR Offsms Syseesse j I
$f
?
~
f j
._ j . .
Ii i -
1 i
4 STACE I RELEASE ,
)k 1-MAIN STEAM DELAY PSPE .
CONDENSER I ~~ls a i (
I STEAM JFT PREHEATER & DDENSER g AIR EJECTOR HYDROGEM-OXYGEN 04ARCOAL AFTER i RECOMEINER ADSORSER FILTER '
MM i
~
Q DRAIN Q DRAIN '
, v t
4 ?;
't i i!
-t s
Figure 2. Schematic of a Typical BVvR ? , .cted Offsms System ,
i
! I
+
i 6
t !
I e
- 6 e e
c.cs ui s,- . ;
,~,,
g--
- .m .
. ,_3w -
a uC.O.tva w ,D. I L
. T :
- e- - k 1 =1 I
- n. , . u = .
- i
..cr c .. . = k , = "5 .
!i .
..sn .. - ,.- ; ---
ic= 1
-,n.-,=
3
= .
.c, as s
&n i I
, oJ i =, = f,
.., = =
A
) ) ) ) ) k l )
s
)
))
s s
)J s s
)J s s
)J ss l I4 F S
- H:H:' 1H:H :,bl i H :H: H:', i qH:2 EENI 3 L_
i a u <- L_L_37 i t : '
i i i g '
3 A
39 n
5 '
y 4 ' '
c...
.,1,[k, . , , [,' .k i #
, V.0, VO VO
.. , .. .,m.. .,.,...,....V rs se M cyca t
- n. _ ,. ,
- I- H: H: 6: H:
- c' <
- - : : ;
as d%
Y LYtt.
y
&I se ..De.5,5
. i Evt t. '
TYPICAL PWR GASEOUS WASTE PROCESSING SYSTEM ,
(STORAGE AND RELEASE) i
l 1i
( To WCT aos ret '
O r
g.
Db r
Orts E*1sferrt: ;,
W.eut teeimetiama - -
- a87E CAS 'T O,
Y.
notoneT- -- { touretssee =. N 1;. y,,,,,,,,
se.c aos avamenareas --_ mm ,
ser ser Pet 25mm u amELattiama n y StaC7 E rectanT sease Tamu- _
masTE Everesary
$ P
- h' (
T CA$amat1 mea (T - l9451EEa3
, DCT Eassamptt atT_ - -_ Compe tssee - g n ,ta., srar g.can: ; j -
l staf
.r .e.. -.
) k ENI, .
- -: g mamaras y P ' CaratTTuC CaratTTIC
- 1 _
===ren 4 .a ma . escs easa g
. d[ 1 1 .
I I !
7 Cosimet .
s f.
3 P F P T TCT lf l hai,0 =
?: r .. O
===rea V
= 4.- - , == 4.- st ==- *
.a.,t. , !
a gr :y; = escette ;;
Ad ' systes ,l A
- i m S-~D du S - , k >]
%A %. -,
= ,
_7 ' ' ,ON" J i
,( %a 2 i[d [ i .y_
4
. { - % (, 5
_ . . . .a, e=ecamas set '.=
i
}
1
-- /. -,
T- ( f , ,,
(f
, ,s i,
TYPICAL PWR GASEOUS WASTE PROCESSING SYSTEM
- (STORAGE AND RELEASE WITH NITROGEN RECYCLE) l
- 1. o. .- ,
an=TMs g a
y
- g: "*
C a -. -
%= == q p '> u. 4 >;
. , ,,,.,,o,.
- e. 9 "O."r' wetuutCearmetrama - 1 Camparetas nott spress 8genuter 1 y O=As ~j I I
i y
samt aos yvapenate.
am.
- e ,,, '
%1 I #
sriE sagamatvna Mruma
=
{
%y
==n sas
.e. =
paCamst l~
- l 14l %h l 3 _
- 3 M,sas i sas =
i I
=av + =av =ut meCrassas [ **** - **** ****
=== ;
c = aces .a== ra== ; , O$
se===ser - ra==
. o$ IE
[ sa sa j s. ,,,
j r .
j = ; ; 2,
- "*** {
i I
. . }< _ % "r 8, 1P 1II 4
D I
i w~ m,m., et' l, y _
p l Car =vre
= -,.
g 4
P
] h-OC v
EF aECvCtt I' w
}( I sas i ! 13 a r O O l' - atteesses.
- ! ra taCree
,wss..n.m Qeen A ~ l (****S5" g utnera== 34 1 -
h I
- Ic MCvns 4
4 IComentsse.!. -
. ; M.
5 i _
TYPICAL PWR GASEOUS WASTE PROCESSING SYSTEM (STORAGE AND RELEASE WITH N2 RECYCLE AND CVCS GAS STRIPPING)
I
__ _ _ _ . .. _-