ML20028C441
Text
s.,
-. :_= i.. :.
~
. 4..n j. s
..-,. i
.-3.,
JMi 3 1378
- - l-Y b..
4#.,
-*, ;_ ;~ ~J.,:. ;, f y.,
e t.
I'.r.;iD'?.'.!!DGi FO.~::
R. J. !!ctcen, Dir2ctor. Divisien of Syste:ns Safety,IRI.
p,
FP0:!:
V. Stello, Jr., Director, Division of Operating Reactors, !!3R
- 4..
- ...n S!?nCT
COR I*:PUT TO COPUIESTM:: DI.': GELL !.ETTER r. ESP 0::SE O!!
~ '
0FfrAS D:ILOSIO::S i
,.o
-c
~
Per l/ro:r rrm'nt.*.. c" closed is the inferration reTnsted by Congressnan
. D. Dingell in b:s lett:r to Cht.ir.rn 1:endric d.r.ted Cecertcr !E,1377; t
~
~-
Cri;..at >!p4 b;- 1. SI:"r.
V. Stello, Jr.. Direc tor
~
Division of C a ratinc P.: actors Of fice of l'ucicca' ?.cEctor R ? quia. tion (~
D.ci r.:ure:
!sa ;tcted cc:
K. Seyfrit (T.Shediosky,RegionI B. Grier J. Cotton, lieg' ion i li. Denten R. Vollcr:r J. Colliar, t!. Butler E. Adanst,n Sectica i'./EEB D.Eisenhut DISTRIBUTIOil Central files EEB Rdg VStello JGuibert RCudlin 8301100158 8212'07
'~
HIATT82-545 PDR n
m
\\
- (/ /
\\
EEBROR Ef.E/j5N@pm. ' _...
EErfdca
..AD/dTID0R D/ DO L. c.2...;
o,r es
- RLo;jpfhi
L'BarrStIt.
.B hi$es h- ^DElienhut_.
VSts.l.lo'
~a-c*g.733. g to 2F6 12L7.2/27..
12/72 /77_
12/ N /77.. i 12/T-:-177 ld/.. /.E _
a*u +-
Hendrie to Dingell Letter input on Offgas Explosion Your letter also requested a written report on the recent explosion at a nu-clear facility in Connecticut.
This explosion occurred at the Unit 1 reactor at the Millstone Nuclear Power Station at Waterford, Connecticut.
Damage was caused by the explosion of hydrogen in the base of the effluent release stack following an earlier internal ignition within the offgas system.
The explo-sion blew the door off the bottom of the stack.
An individual near the base of the stack was knocked down and subsequently treated at a local hospital for lacerations of his leg and possible concussion and later released.
Physi-cal damage caused by the explosion was confined ;c the instruments and fixtures of the stack basement area. The NRC staff has calculated the amount of radio-activity released and determined that the radiation dose consequences were in-significant.
(The calculated maximum radiation that an individual offsite received was less than one percent of the annual natural background radiation.
Radiation doses to onsite personnel, including the injured man, were well within Federal radiation standards as given in 10 CFR Part 20 of the Comission's regu-lations.) Millstone has taken precautions to prevent the recurrence of such an explosion by valving off the loop seals and reviewing procedural controls.
Further investigation of this incident by the NRC is being conducted.
A sumary report describing the events that led to the explosion on December 13 1977 is enclosed.
There have been about 29 other off-gas system explosions since 1971 but in only three other cases gas escaped the system and caused damage as at Millstone.
Included in the attached report is a brief summary of those explosions and a comparison of the consequences of those explosions with the
1 2-explosion at Millstone. These explosions are unique to boiling water re-actors manufactured by the General Electric Company. "The other three major U. S. light water reactor vendors manufacture pressurized water reactors which do not transfer potentially explosive gases to the turbine system.
4 0
4 0
h e
0
0 6
e DESCRIPTIONOFTHEMILLST0liEHYDROGENEXPLOSION we An explosion of hydrogen gas from the offgas system occurred at Millstone Nuclear Station Unit 1 on December 13, 1977. The explosion resulted from ignition of hydrogen and oxygen that had collected in the basement of the stack.
The following describes the causes of the explosion and its relation to other similar explosions that had occurred since 1971.
Millstone Unit 1 is a Boiling Water Reactor (BWR) manufactured by the General Electric Company. After steam generated in the reactor vessel is transferred to and leaves the turbine generator, it enters a vacuum con-denser where it is condensed to water to be returned to the reactor.
Figure 1 is a schematic which illustrates the flow path of the steam.
Noncondensible gases also enter the condenser along with the exhaust steam from the turbines.
The noncondensible gases are composed of the following:
1.
hydrogen and oxygen formed by the radiolytic decomposition of water in the reactor when irradiated in the reactor core, i.e., 2H 0 + 2H2+0I 2
2 2.
air which leaks into the condenser because the condenser is at a vacuum relative to the atmosphere; 3.
radioactive gases such as N-16 formed by the neutron activation of re-c.2__
actor coolant and dissolved gases; 4.
trace amounts of radioactive fission products leaked out of small de-fects in th'e fuel cladding.
2-The noncondensible gases are drawn from.the condenser by the steam jet air ejector (SJAE) to keep the condenser at a vacuum.
The noncondensibles which pass through the air ejector are referred to as "offgas" and the piping sys-tem which receives the offgas is called the offgas system.
At Millstone Unit 1, the offgas flows through the large sections of pipe at a slight positive pessure created by the air ejector.
The purpose of this pipe is to delay the release of offgas for about 30 minutes to allow the decay of short-lived radioactive gases.
After passing through the delay line the offgas is filtered and released from a tall stack.
The offgas is a highly explosive gas mixture because of its high hydrogen and oxygen content. Whenever explosive gas mixtures are present it is good en-gineering practice to assume that some event may initiate an explosion of the gases.
Consequently, BWR offgas systems are designed to withstand the effects of an offgas explosion internal to the offgas system.
Offgas explosions Lave occurred when the hydrogen and oxygen mixtures were ignited by a variety of causes, such as lightning, presence of foreign mterials which may initiate an explosion (catalyst) and personnel errors such as welding or heating offgas piping above ignition temperatures.
A list of occurrences and probable causcs is given in Table 1.
At Millstone it is thought that the initial explosion within the offgas systen u.
was caused by a welder who, when welding an adjacent pipe (not part of the offgas system) inadvertently made contact with an instrument line which con-tained explosive offgas.
The welder's are heated the instrument line suffi-ciently to ignite the offgas within the pipe.
The ignition quickly spread
[
~
' down the instrument line to the delay lihe causing a rapid burning or explo-Thisoccurredatapproximately9:30a.m.onDecemberi3, sion of the offgas.
1077. 'This did not cause any apparent external damage; however, it did re-move water from loop seals in lines which drain water condensed from the moist air ejector effluent.
drain pipe which maintains a water seal A loop seal is a vertical loop in a in the pipe to prevent the explosive offgas from entering the drain system.
Figure 2 is a sketch of a typical loop seal.
Without water in the loop seals, explosive offgas entered the drain sumps and spread into the reactor building basement and the basement of the stack.
An effort was made to refill the loop seals with water to keep the offgas inside the offgas system. The loop seal in the reactor building was quickly re-established.
However, the attempt to re-seal the loop seal in the stack basement was not successful and a slow reactor shutdown was initiated. The explosive offgas continued to accumulate in the stack basement while the reactor power was being reduced. At approximately 1:00 p.m. on December 13, 1977, this buildup of explosive offgas in the base of the stack also ignited causing an explosion.
The explosion blew the door off the bottom of the stack and caused a man near the base of the stack to be knocked down.
The man was treated at a local hospital for lacerations of his leg and possible concussion and later released.
Physical damage caused by the explosion was confined to the instruments and fixtures of the stack base-u ment area.
The NRC staff has calculated the amount of radioactivity released and determined that the ratiation dose consequences were insignificant. The i
calculated maximum radiation that an individual offsite could have received was less than one percent of the annual natural background radiation.
Radiation
0
. doses to onsite personnel, including the injure,d man, were estimated by the licensee to be well within the permissible Federal radiation standards,10 CFR Part 20.
Millstone has taken precautions to prevent the recurrence of such~an explosion by valving off the loop seals and reviewing procedural con-trols.
Further investigation of this incident is being conducted by the NRC.
There have been 29 known offgas explosions since 1971; all except four (in-cluding the Millstone explosion) were confined within the offgas system and caused no damage other th 9 to internal seals and filters.
Table 1 lists the 29 known offgas explosions since 1971 together with the probable causes and the damage caused by those four explosions external to the system.
Generally, the explosions only occur inside of the offgas system, however, it is possible (as at Millstone) to also have an explosion external to the offgas system if the explosive offgas is allowed to accumulate in areas outside of the system.
Offgas escaped from the offgas system at Cooper once because the loop seals were blown out by a system pressure transient caused by ice buildup in the stack and once because of improper valve lineups.
Offgas escaped from the system at Brunswick because of a pressure transient caused by condensation in the final offgas filters.
S-__
l
TABLE 1 BWR OFFGAS HYDR 0 GEN EXPLOSIONS
~
Ignition or Damage External Plant Date to System Probable Causes Dresden 1 8/1971 None Static electricity charge Monticello 11/1971 None Improper grounding of filters Lacrosse 6/1972 Nnne Catalytic action in recombiner Quad Cities 3/1973 None Lightning Dresden 2 4/1973 None Welding Spark Vermont Yankee 6/1973 None Lightning Vermont Yankee 8/1973 None Lightnino Vermont Yankee 9/3/1973 None Lightning Vermont Yankee 9/9/1973 None Lightning Dresden 3 2/1974 None Filter improperly grounded Quad Cities 1 3/1974 None Inadvertent grinding on offgas pipe Monticello 5/1974 None Catalyst migration from re-combiner Monticello 6/1974 None Catalyst migration from re-combiner Monticello 7/1974 None Catalyst migration from re-combiner Quad Cities 1 10/1974 None Improper grounding of filter Dresden 3 11/1974 None Spark from recombiner valve operation Dresden 2 5/1975 None Recombiner malfunction Cooper 11/1975 (1)
Improper valving caused of fgas to flow to sump area. Off-gas ignited by spark from a
- monitor, Cooper 1/1976 (2)
Ice plug at stack a--"
Brunswick 2 1/1976 (3)
Condensation blockage at filter Quad Cities 2 3/1976 None Improper grounding of filter Cooper 4/1976 None Static charge at filter Dresden 3 4/1976 None Static electricity at filter Dresden 2 6/1976 None Recomt ler malfunction Oyster Creek 8/1977 None Catalyst migration from re-combiner
TABLE-1 (Cont'd)
BWR OFFGAS HYDR 0 GEN EXPLOSIONS Ignition or Damage External Probable Causes Plant Date to System Dresden 1 9/1977 None Improper grounding of filter Browns Ferry 7/19/77 None Recombiner malfunction Vermont. Yankee 12/10/77 None Offgas accumulated in'decom-missioned pipe section Millstone 1 12/13/77 (4)
Initial detonation caused by welding (1)
Explosion at sump area.
No physical damage.
One person received burns.
(2)
Demolished small auxiliary equipment building containing the offgas dilution system.
(3)
Damage in room at base of stack, broken glasses, relay covers, lighting and blown access hatch.
(4)
Damage in stack basement, instruments, and blew out access door.
O e
a
STACK RELEASE A
l ',
MAIN
~
STEAM TURBINE
~ - ~ ~ >"- E LECTRICITY
- [
GENERATOR OFFG AS SYSTEM REACTOR D
FILTEP.
CON ENSER l
A j
ISOLATION STEAM JET DELAY VALVE AIR EJECTOR-PIPE CONDENSATE REACTOR CONTAINMENT
t g
4 9
1
OFFGAS c.
~
h.f D
=.
\\
~ _ -
1:
~ _.. -
WATER DRAIN SUMP Figure 2 Typical Loop Seal.
P b
6 e
e-