ML20028C412
Text
.
.f g.
m DIRECTORATE OF REGLT ATORY OPERATIONS.
REVIEW OF EbLOSIONS IN BOILING WATER REACTOR 0FF-GAS SYSTEMS Report Prepared By:
R. J. McDermott, SRIS Facilities. Inspection Branch, DRO Report Reviewed By:
K. Seyfrit, Chief Technical Assistance Branch, DRO i
i b
i ENCLOSURE 1 8301100000 821207 PDR FOIA HIATT82-545 PDR
o e
's n)i@j) b.
, l',
i INDEX I.
INTRODUCTION II.
SUMMARY
AND CONCLUSIONS III.
EXPERIENCE WITil EXPLOSIONS IV.
DESCRIPTION OF OFF-GAS SYSTEMS AT FACILITIES EXPERIENCING EXPLOSIONS V.
OFF-GAS SYSTEM OPERATIONS IV.
EXPLOSION SENSING DEVICEi AND ISOLATION VALVES VII.
OVERPRESSURE RELIEF DEVICES IN OFF-GAS SYSTEMS VIII.
RADIOLOGICAL EFFECT OF EXPLOSIONS IN OFF-GAS SYSTEMS IX.
PROCEDURES OR PRACTICES FOR COPING WITH EXPLOSIONS DURING REACTOR OPERATION X.
MAINTENANCE PRACIICES AND PROCEDURES FOR OFF-GAS SYSTEMS ATTACHMENT 1.
DETAILS ON EXPLOSIONS THAT HAVE BEEN EXPERIENCED ATTAC10!ENT 2.
DETAILS ON OFF-GAS SYSTEMS AT FACILITIES EXPERIENCING EXPLOSIONS ATTACHMENT 3.
DETAILS ON EXPLOSION SENSING DEVICES AND ISOLATION LOGIC ATTAC1! MENT 4.
DETAILS ON OFF-GAS SYSTEM OVERPRESSURE RELIEF DEVICES s
s
.m._
I I.
INTRODUCTION At the tima of the initiation of this review, seven explosions were known to have occured in off-gas systems in boiling water reactor facilities. These explosions have resulted from the ignition of the hydrogen (in an oxygen rich atmosphere) that is normally present in the off-gas system.
The hydrogen and oxfgen are generated by radia-tion disassociation of primary coolant water as it passes through the reactor core. - These gases are carried in the steam flow to the main turbine.
These and other ncn-condensable gases are continuously removed from the main turbine condenser during plant operation and directed into the off-gas systa= by the steam jet air ejectors.
During plant operation, approxinately 60% of the total volume of non-condensables in the off-gas system is hydrogen - well above the 4%. flammable limit for hydrogen in the presence of oxygen.
Becaus e of the very slow race of recombination of hydrogen and oxygen, under ambient temperature conditions, explosive concentrations usually exist in off-gas systems during reactor shutdown periods unless the system has been purged.
In general, the explosions experienced to~
date have resulted in blowing of system rupture diaphragms (over-pfessure protection devices).
A broken rupture diaphragm effectively t
bypasses the normal 30-minute delay line in the off-gas system and results in a significant increase in the release race of gaseous radioactive wastes if the plant is in operation.
Although AEC inspections were conducted following each of these oc'currences to review the occurrences and to assure that no immediate health and safety problems existed, Regulatory Operations initiated this special study of six of these occurrences to review the collec-tive experience related to such explosions and to review selective aspects of off-gas system design and op,eration for boiling water
{
reactors. The study included a review of factors responsible for the explosions, radiological safety considerations, the design and per-formance of off-gas system isolation features, and operations, testing, and maintenance practices for off-gas systems. The study involved a review of related abnormal occurrence reports that were submitted to i
the AEC by licensees, applicable AEC inspection and inquiry reports, j
and information contained in Safety Analysis Reports on off-gas system design.
Visits were made to the reactor facilities that had experienced explosions (with the exception of those facilities l
that experienced explosions after the s'cudy had been essentially completed) to review system design specifications, system and compo-
{
nent arrangement, and off-gas system operating, tes ting, and main-tenance practices.
s l
/
g
. II.
SUMMARY
AND CONCLUSIONS A.
Causas of Explosions.
The explosions in off-gas systems at BWR facilities have resulted from a variety of causes.
Of the six explosions covered by this study, one resulted from a lightning strike at or near the plant, two resulted from f ailure to properly pu'rge the off-gas systa=
prior to or during maintenance, two resulted 'from equipment =al-functions, and one explosion occurred from an unknown cause.
The diverse causes that have been responsible for explosions to date, suggest that additional explosions can be expected to occur in the future.
The total experience with licensed BWR's discloses that explosions have occurred at a frequency of once per seven reactor-years of operation.
B.
System Design
A review of off-gas performance and plant design features,
disclosed the following:
1.
A single failure,of one of four off-gas system isolation 2
valves to close would result in failure of the off-gas system to isolate following an explosion.
(Valves are arranged in parallel with two valves normally open at current generation j
BWR's.)
2.
The automatic reset feature of the isolation initiating circuitry at the majority of current generation BWR's would allow an off-gas system isolation condition to automatically clear (and isolation valves to reopen) following an explosion.
3.
The design of the off-gas system isolation initiating logic does not provide protection against a single failure of an explosion sensing device.
4.
Design pressure of the off-gas system piping and overpressure protection devices provided for the off-gas system appear adequate in that no significant failures or damage have occurred to piping systems following explosions.
5.
Lightning protection for reactor facilities has not proven to be adequate based on the explosion experienced at Quad Cities (and the recent one at Vermont Yankee not covered in this review).
I l
- -. - -,,.,, - ~. - -...., - -,
.. =.
- ggg
- ,,
5 2"
.g-
=
C.
System Operation. Testina and baintenance A review of procedures -and practices relating to the operation, testing and maintenance of off-gas systems disclosed the ' oi, lowing:
f 1.
Operating or emergency procedures to cope with ' xplosions e
during plant operation had not been prepared at the ti=a of the explosions.
2.
Maintenance procedures recognizing the explosion hazards associated with the off-gas system had not been prepared at the time of the explosions.
3.
Routine purging of explosive gas concentrations has not been perfor=ed following plant shutdowns as directed by GE specifications.
Use of service air to purge, as pro-vided in the system design, has proven to be ineffective.
4.
Periodic calibration and functional testing of off-gas system. isolation features has not been performed at all facilities.
D.
Radiological Ef fects of Explosions The effects of the off-gas system explosions experienced to date, as well as the potential effects of such explosions, were reviewed with the following findings :
l.
Based on information contained in licensee reports to the AEC and AEC inspection findings, the explosions have not resulted in significant off-site doses or significant i
exposures to in-plant personnel.
3 2.
Conservative calculations performed as a part of this study indicate that-off-gas system explosions that result in blowing of rupture diaphragms in current generation SUR's during plant operation, could result in a =aximum increase in the stack release rate of radioactive noble gases by a
{
factor of approximately 7, and could also result in a maximum stack release rate of short-lived activation gases (principally N-16 and 0-19) of approximately 4 Ci/sec.
i i
These increases in release rates result from a reduction l
in the time available for radioactive decay of the off-gas before release from the facility (from a nominal 30 minutes to a conservac1vely estimated 40-second delay from the time the radioactive gases leave the reactor vessel).
I l
l l
+$
. P.
m t -
For BWR facilities that-have off-gas system rupture diaphragms that discharge diredtly to the outside atmos-phare, decay times would be further ieduced and are con-servatively estimated to be approximately 15 seccnds from the time the radioactive gases leave the reactor vessel.
- 3.
- Dose calculations performed by DL, using conservative assumptions, indicate that off-site doses on the order of a few rem could occur as a result of failure of the off-gas systet that results in a ground level release.,
The principal contributors to the dose were found to be the immersion dose from noble gases and particulates.
No credit was given in these calculations for.an elevated release through the plant stack.
An elevated release, the probable release path following explosions, would result in a significant reduction in the calculated off-site does.
III. EXPERIENCE WITH EXPLOSIONS To date, seven explosions in off-gas systems have occurred at BWR facilities as a result of ignition of the hydrogen-oxygen gas mixture present in the system. The hydrogen and oxygen are non-condensable gases that are formed by radiolysis of the primary coolant as it passes through the reactor core.
These non-condensablu gases are removed from the main condensor during plant operation by the steam jet air ejectors which discharge into the off-gas system.
i Hydrogen and oxygen (in the ratio 2:1) typically represent %80%
of the volume of the non-condensables in the off-gas system (well above the flammable limit of 4% for hydrogen in the presence of oxygen).
Because of the very slow rate of recombination of hydro-gen and oxygen, explosive gas concentrations also exist in the off-gas system during reactor shutdown periods, unless the system is purged.
The frequency of these explosions, as well as the diverse causes that have been responsible for the explosions to date, sug-gest that additional explosions will occur in the future.
Bas ed on the collective operating history of all AEC licensed BWR power-reactors, explosions have occurred at a rate of once per seven reactor-years of operation.
f The experience available from the explosions reviewed shows, in l
general, that:
1.
Causesforexpkosionswerevaried.
2.
Explosions have occurred both during plant operation and during plant shutdtwn periods.
1 i
l l
l 1
,e
-n,--
,--n---..
,-,,,..-,,,-w----.
-2
~"
- str.
- J."."
+::::::
3.
Immediate actions taken by Licensees relating to cont $nued operation of the reactor have varied between licensees.
4.
Corractive actions taken by licensees appear adequate.
5.
Several deficiencies relating to construction, design, maintenance, and opera' tion have been disclosed as a result of the explosions. provides a listing of the facilities that have experienced explosions and includes infor=ation on the-following aspects of each:
1.
Date of event.
2.
Reactor status prior to explosion.
3.
Indicated stack release rate prior to event.
~
4.
Indicated stack release rate (maximum) following ' event.
5.
Detection of explosion.
6.
Activities in progress immediately prior to event.
7.
Reactor status following event.
8.
Inspections and checks made to system following event.
9.
Damage to components and sys tem..
10.
Cause of explosion.
11.
Deficiencies observed.
12.
Corrective actions taken.
IV.
DESCRIPTION OF OFF-GAS SYSTEMS AT FACILITIES EXPERIE"CING EXPLCSIONS To aid in understanding similarities and differences in off-gas systems at the BWR's that had experienced explosions, a review was made of information in Safety Analysis Reports and discussions were I
held with licensee personnel. provides a summary description of the major components in the off-gas system and describes f
n 1,'
=L 43 7
I -
the normal flow path through the' system.
In addition, Attachment 2 provides information on the following as'pects of the off-gas system for each facility:
1.
Off-gas system piping design pressure and
- temperature.
2.
Design flow rate of' system.
3.
Discharge path and dilution air supply for off-gas.,
4.
Isolation features provided for explosions.
5.
Explosion sensing devices.
c 6.
Overpressure relief devices.
7.
' Manual isolation provisions.
~
8.
Design assumption of the concentration of combustible gases present in the system.
9.
Filter design.
10.
System designer.
A review of GE design specifications for the off-gas systebs at
~
Monticello, Dresden 2 and Quad Cities 2, indicated that the explosive 4
concentratfons of hydrogen were considered in the design of the off-gas system.
Typical design specifications stated that "the presence of an explosive mixture in the system requires that certain portions of the air ejector system piping, instrumentation and equipcent be designed to eliminate all sources of ignition, such as spark, heat, static electricity, etc., and that the system is to be designe,d to l
withstand both static pressure and shock wave forces should an i
explosion occur.
The method of determining these pressures may be found in NACA - TN-3935 "Hvdrogen Explosions in Exhaust Ducting."
In only one instance (Monticello), could the inspector obtain documentation or information at the site that provided the specific provisions that were selected by the system designer to eliminate sources of ignition in compliance with GE's specifications.
In this instance, it was noted that the design specification for the off-gas system radiation monitor required the use of sparkless motors, external heaters, and insulation of the heater from the radiation monitor sample line.
t x..x a. ss. n
_7-A review of NACA - TN-3935 disclosed that a pressure rise ratio of ~
21 could be expected in piping following fetonations of hydrogen-oxygen gas mixtures (oxidant - fuel mole-ratio o*f 0.82) at 1 accos-phare. In addition, data from NACA - TN-3935 disclosed that total pressures exerted on various designs of 90' elbows by detonations were expected to increase by a factor of approximately 60.
These
~
~
data also indicated slightly higher. pressures would be expected for oxidant - fuel mole ratios of N.5,,the gas mixture generally e'xperienced during reactor operation.
Specific details of piping system design were not reviewed during site visits to date.rmine if the'syste'm had been designed to withstand the dynamic effects of explosions.
Based on the explosion experience to date, it appearc that a design pressure for piping systems of approximately 300 psig is adequate, in that no significant damage to piping systems hat occurred.
Two relatively minor failures of system co=ponents occurred following the explosions at the Lacrosse SWR (See Attach-ment 1).
Stack lightning protection was reviewed at the Monticello facility.
At this plant, the stack lightning rods (and grounding wires) and '
the metal stack located within the concrete stack (through which l
the diluted off-gas passes and to which the off-gas system is directly connected) provide parellel electrical grcunding paths for arresting lightning strikes of the stack.
No special lightning
, shielding provisions were found to be present over the inner stack that is exposed at'the top of the plant stack.
As no special l
. grounding or electrical isolation provisions are provided for the off-gas system, it appears probable that lightning strikes of the stack at this facility would result in electrical charge =ovement through the parallel paths. This appears to be a potential source of ignition of the gases in the off-gas system.
j At. Quad Cities 2 and at Vermont Yankee where explosions were experienced following lightning strikes, the inner metal stack is L
not present within the concrete stack and the concrete stacks are not lined with metal.
No possibilities or mechanisms were identi-i fled in this review to explain these explosions.
Recent investi-gations perfor=ed by Commonwealth Edison Company at the Quad Cities plant indicated a possible explanation for the Quad Cities 2 explosion.
The possibility identified relates to a concern that a potential difference could be created between the filter media (in the filter at the end of the off-gas delay line) and the filter cartridge body by a lightning strike in the vicinity of the plant.
This could result in a spark and ignition of the off-gas.
This concern results from the fact that the metalic portion of the filter media is electrically isolated from the cartridge which is l
solidly grounded to the filter vessel and off-gas pipe.
j
N g5y
.. ;.3 i
V.
OFF-CAS SYSTEM OPERATION A review of operation of the off-gas system at the five facilities visited, indicated fairly standard practices.
On plant startup, the mechanical vacuum pump (no holdup, with a direct discharge to the plant stack) is used to establish an initial vacuum in the main condenser.
Af ter reactor steam pressure is raised to 100-200'psig, the SJAE is placed in service and the mechanical vacuum pump is.
removed from service prior to condensing any significant quanticles of primary steam in the main condensar.
Therefore, under nor=al operating startup modes, no combustible concentrations of hydrogen and oxygen shou.\\d exist in the mechanical vacuum pump and its dis-charge piping.
System shutdown is accomplished by maintaining the SJAE in service and maintaining vacuum until system steam pressure decreases below that required for operation of the SJAE (typically 100-200 psig).
The mechanical vacuum pump is then placed in service and vacuum is maintained until sealing steam is lost, at which cite the vacuum is broken on the main condenser.
None of the facilities
, visited perform routine purges of the off-gas system following plant shutdown to eliminate explosive gas concentrations that could remain in the systems.
This is contrary to system specifications for CI-supplied BWR's.
Purging of the system, by means of the installed purge connections using service air, at the Monticello facility has not been effective in removing explosive concentrations even af ter four-hour purges.
Actual plant operating data indicates that typical off-gas system flow rates and concentrations of combustibles in the off-gas system at 100% power are as follows:
1.
Monticello - 70 SCFM, 66 of which are hydrogen and oxygen (H-0).
2.
LACBWR - 12 SCFM, no measurements have been made at the site to determine H-0 concentration.
3.
QC 120 SCFM, 99 of which are H-0.
4.
D 110 SCFM,103 of which are H-0.
)
5.
D 28 SCFM, 24 of which are H-0.
sk.
t 3,.
3
. u combustible gas would also result in failure of the system to isolate. Although the initiating iggic at Monticello differs, 1.e., any combination of two of the four channels tripping result in isolation, it would also.be subject to a single failure-defeating isolation because the temperature and pres-sures required for trip actuation may not occur simultaneously.
Information on temperature sensor response time to step changes in system temperature, was not available at the sites.
At one facility, instrument " snubbers" were noted to be installed in the sensing lines for the pressure sensor.
Their effect on operation of the pressure sensors during explosions was unknown by licensee personnel.
System specifications for at. leas t, one facility r,equired redundant features to mini =1 e sperious iso-lation valve closure on electrical or air failure.
The require-ment precludes logic arrangement of one-out-of-two and indicates the optimal choice may be one-of-two-twice using four pressure switches for inputs to the logic.
3.
The parallel arrangement of isolation valves between the main condenser and the SJAE's at Qua'd Cities 2, Dresden 2 and Monticello, permits a single valve failure to defeat isolation of the off-gas system.
Operational problems (principally failures to open) have been experienced with these isolation valves at one facility.
4.
The isolation circuitry at Quad Cities 2 and Dresden 2 will reset automatically following an explosion.
It appears, there-fore, that the isolation feature only serves a purpose of tempo-rarily isolating the system to stop the flow of combustible gases into the sys tem.
It also appears possibic that multiple explosions could result with this automatic reset capability if a source of ignition remained present in the system, such as burning of absolute EEPA filters.
5.
The calibration or testing frequency of instrumentation that initiates isolation and the frequency of functional testing.of isolation valves had not been established as a routine item at all facilities. provides detailed information in the following areas for each site visited:
1.
Pressure sensors 2.
Temperature sensors
~
s.
?,
A e
%K?
x=
3.
Logic for. isolation
~
t 4.
Reset. features of isolation circuitry 5.
Calibration and functional test frequency of sensors VII.
OVERPRESSURE RELIFF DEVICES IN OFF-CAS SYSTEMS At each facility visited, a review was made of system overpressure protection features that could be expected to operate following an
. explosion.
This review disclosed that the facilities had provided' rupture diaphragms for overpressure protection for explosions with the exception of LACSUR.
In addition, most of the off-gas systems contained loop seals (for continous drainina.of water from the system).
These could also be expected to provide some overpressure.
protection.
It appears probable, based on experience to date and review of the system design, that the rupture diaphragms would fail following the explosiens as piping system pressure can be expected to increase by a factor in excess of 20 (%300 psig).
Explosions could also result in blowing through loop seals; however, the experience at OC-2 indicates that these could refill following the.
initial pressure pulse with the water formed in the system as a combustion product of the H2+O2 explosion reaction.
The types of devices installed, set pressure, location in system, and discharge path from each overpressure relief device are listed in Attachment 4.
V'III.
RADIOLOGICAL EFFECTS OF EXPLOSIOUS IN OFF-GAS SYSTEMS The potential effect (increase in the stack release rate of radio-active gases) of explosions occurring at full reactor power were esi-culated for the five facilities that had experienced explosions.
The results of these calculations are contained in Table 1.
The assunp-tions used in perforning these cciculations are given in Table 2.
The calculated stack release rate for-Ouad Cities 2 was compared with the measured stack release rate following the explosion at that faci-lity.
The stack release rate as measured by the stack monitor, increased from 5,000 to 107,000 uCi/see which was significantly lower than the calculated peak stack release rate value of 4,000,000 uC1/sec, however, several factors vere identified that could partially explain this large difference.
These factors are given in Table 3.
After applying appropriate correction factors to the measured value, it is estimated that the actual peak release rate was near 300,000 uci/
sec.
It is also estimated that the major fraction of this activity was due to short-lived activation gases (N-16 and 0-19).
Calculations s
.,,._,.,_,-x
_-.,,.-.w--
~~.y,,,,.y,.
,_mm.__..._.
=h
- s.
b.
performed by the AEC staff,1/ using the actual meteorology that existed at Quad Cities 2 at the time of the explosten, and an ele-vated release from the plant stack resulted in a calculated dose of less than 1 mr at the exclusion fence boundary.
Dose calculations were also perforced by the AEC staf f / that dis-1 closed a potential for off-site doses on the order of a few rem if ground level releases were assumed.
The principal contributor was calculated to be the immersion dose from noble gases and parti-culates.
The assumptions used in this calculation were as follows:
1.
100,000 uC1/see release rate through the plant stack existing prior to the rupture of the off-gas system piping.
Gas cQ=po-sition identical to NEDO-10734.
2.
30-minute delay time in the off-gas system hold-up piping prior to the explosion and system failure.
3.
Radioactive gases assumed to be released to the envirens 15 seconds after they leave the reactor vessel (assuming delay time due to transit between tho reactor vessel and point of failure in,the of f-gas sys tem).
4.
A ground level release of radioactive gases continuing for one hour af ter the explosion.
5.
Meteorology conditions - wind speed of 1.0m/sec and X/Q of 3
1.6 x 10-3 sec/m,
6.
Distance from failure point to exclusion fence boundary 1000 ft.
1/
Memorandum, H. R. Denton, Assistant Director for Site Safety, L to J. G. Keppler, Acting Assistant Director for Construction & Operation, dated August 29, 1973.
l
gEa g:;;;;b F
O l
TABLE 1 CALCULATED MAXDfUM PEAK RELEASE RATE OF ACTIVATION GASES FOLLOWING FAILURE OF OFF-GAS SYSTEM [uci/sec]
~
Facility Tsotope N-13 N-16 N-17 0-19 retal 2
0 5
LACBWR 5.6 x 10 2.3 x 10 2.3 3.0 x 10 2.6 x 10 0
1 5
6 Monticello 5.7 x 10 2.4 x 10 2.4 x 10 3.1 x 10 2.7 x 10 Quad Cities 2, 3
6 1
5 6
Dresden 2 8.5 x 10 3.5 x 10 3.5 x 10 4.6 x 10 4.0 x 10 7
2 5
7 Dresden 1 2.4 x 10 1.1 x 10 6.3 x 10 2.4 x 10 1.1 x 10 CALCULATED INCREASE IN NOBLE GAS RELEASE RATE The noble gas release rate from the stack was calculated to increase by l
a factor of 7 from the pre-explosion value based on data provided in t
NEDO-10734 and using a 40 second hold-up time for the gases in the plant (15 seconds from reactor vessel to point of failure of the off-gas system and 25 seconds holdup due to transport of 3ases to the top of the stack via the ventilation system).
t i
i t
i i
i i;
r t
i n,
- =n '
, " =
'i.5. ;
z"_:.....
TABLE 2 ASSm!PTIONS AND REFERENCES USED FOR TABLE 1 CALCULATIONS ACTIVATION GASES 1.
Release rate in uC1/sec from the reactor vessel - NEDO-10734, dated February 1973, data.
Activation gas data was scaled down by multiplying all data by the ratio of [ plant licensed power level + 4200] since NEDO-10734 applies to a 4200 Mit reactor.
2.
Delay time from. reactor vessel to steam jet air ejectors of 13.5 seconds - NEDO-10734*.
3.
Additional time delay before gases exit from point of failure in off-gas system of 1.5 seconds (esti=ated).
(Assumed point of failure for all facilities was the rupture diaphragm.)
4.
Ventilation system delay time of 25 seconds before gases exit the top of the stack - this estimate was based on a review of the ventilation system for LAC 3WR, Monticello, Quad Cities 2 t
and D.resden 2 (and averaging the delay times).
In the case of Dresden 1, this additional delay time was not used in the cal-culation because the rupture diaphrages relieve directly to the environs.
NOBLE GASES
~
1.
Delay time of 30 minutes in the off-gas hold-up piping prior to the explosion.
1
- This estimate is considered to be conservative for Dresden 1 (dual cycle BWR) as NED0-10734 is only valid for a direct cycle BWR.
Kahn,
Bureau of Radiological Health, HEW estimates 60 seconds for Dresden 1.
Ref. BRH/ DER 70-1.
f F
+
e r-
6 s
A J",
, TABLE 3 COMPARISON OF CALCULATED AND MEASURED STACK RELEASE RATE AT QUAD C'ITIES 1.
MEASURED PEAR STACK RELEASE RATE 107,000 uCi/sec 2.
CALCULATED PEAK STACK RELEASE RATE 4,000,000 uC1/sec 3.
Possible causes for difference between measured and calculated values:
s.
Capability of stack radiation monitor to correctly measure release rate - There is an inherent delay time associated with the stack
~
monitor sample line that permits additional 'deday time for short-lived isotopes.
This delay time is estimated to be 42 seconds, based on 200 ft, sample line length, 1 in. diameter sample line and 1.7 cfm flow race.
This delay time would result in the stack monitor measuring less activity than what is actually being discharged from the plant stack if short-lived isoto' pes are present.
In addition, the stack monitor would not respond accuracoly to high energy radiation such as given up by 0-19 and N-16 due to the reduction in total intrinisic efficiency of the NaI detector at these higher energies,. Applying correction factors for sample line decay time and reduction in detector efficiency (33%) to the calculated stack release rate of 4.0 x 100 uC1/sec of activation gases yields a calculated peak stack monitor response of 149,000 uC1/sec. Adding the calculated stack release of 35,000 uCi/sec (a f actor of 7 increase) of noble gases yields a total calculated stack monitor response of 184,000 uCi/sec versus 107,000 uCi/see actually measured.
b.
Incorrect assumption used in ventilation system hold-up time -
if increased hold-up time of the gases within the plant is assumed to be greater than the 25 seconds used in the calcu-lations, the calculated stack release rate for activation gases is substantially reduced.
Some additional decay time could be assumed for the gases to pass through the steam jet air ejector room before they enter the ventilation system.
The assumption of total times of 1 minute and 2 minutes from the time the gases exit the rupture diaphragm until released from the top of the plant stack, reduce the calculated value f rom
- Nuclear Instruments and Their Uses.
Volume 1, A. H. Snell.
a
=
g
-u-i 5
4 4.0 x 106 uC1/see to 3.0 x 1 and 4.6 x 10 uCi/sec, respec-tively.
Applying correction'f actors from a. above and adding the calculated increase due to noble ~ gases,, yields calculated peak stack monitor responses of 108,000 uCi/sec (1 minute delay) and 55,000 uCi/sec (2 minute delay) versus 107,000 uCi/sec actually measured.
c.
Incorrect assumptions of delay time in the off-sas system hold-up piping - The calculation for the increase in noble gases was made using an assumed 30-minute delay in the hold-up. piping.
Extrapolation of NED0-10734 data to a 70-minute hold-up tice (the actual delay time at the time of the explosien). indicates the release rate of noble gases would increase by a factor of 9 if the hold-up time in plant was reduced (by a blown rupture.
diaphragm) from 70 minutes to 1 minuta.
This factor would only have minor effects on the calculations and would not appreciably change the calculated values.
d.
Differences between the isotopic composition of off-gas assumed in NEDO-10734 and actually present at QC-2 at the time of the explosion - NED0-10734 data that were used in the calculation of the release race of noble gases are based on the. GE "19 71 diffusion" mixture exis ting in the off-gas sys tem.
The compos 1-tion of off-gas at Quad Cities 2, measured a few days prior to the explosion, was essentially a 100% " equilibrium" mixture.
Comments received from Quad Cities personnel indicate an expected-increase in the release rate by a factor of 2-3 for noble gases following blowout of the rupture diaphragm when an " equilibrium" mixture is present.
e l
1 l
l
h' g
. gf
')
. ~ '
II.
PROCEDURES OR PRACTICES FOR COPING WITH EXPLOSIONS DURING REACTOR OPERATION Discussions during site visits disclose the following; A.
LACBWR - No procedure was available and no established pra'c-tice was in effect to cope with explosions in the off-gas system.
Discussions with station personnel indicated that the plant would probably be shut down following an explosion; however, this had not been done for the two explostons experi-enced to date.
B.
Monticello - No procedure available and no established practice; however, discussions with station personnel indicated the following actions would probably occur or be tahen:
1.
Automatic isolation of off-gas system would occur fol-loving an explosion.
2.
Operator would reduce power to %40% with recirculation pump speeds.
~
3.
Operator would manually scram the reactor before vacuum on main condenser was lost.
C.
Quad Cities 2 - A procedure was developed following the explo-sion expereinced at the facility.
The procedure s'pecifies continued operation of the reactor and manual isolation of the i
air ejector with the blown rupture diaphragm, within 30 minutes from the time of the explosion.
The procedure also requires valving in the standby EEPA filter.
D.
Dresden 2 - Procedure is availab'le.
Procedure requires a reactor scram if an explosion occurs.
E.
Dresden 1 - Procedure is available.
Procedure requires a l
reactor scram if diaphragm ruptures, t
9 M
....... e
5:59. -
iSE::.
- Er "su t
X.
3fAINTENANCE PRACTICES AND PROCEDURES FOR OFF-GAS SYSTEMS Discussion during site visits disclosed the following:
A.
LACBWR - No special procedures were in effect to control maintenance; however, special precautions were being taken to purgs tanks of explosive gases prior to maintenance.
No pro-cedures were in effect that require sampling for explosive mixtures prior to maintenance activities.
The off-gas system is not purged routinely on shutdown.
B.
Monticello - No general procedure was in effect; however, any modification to the off-gns system is accomplished by a special' procedure that is reviewed by the site Opqyations,Conmitree.
System modifications are preceeded by a purge utili:ing the mechanical vacuum pump lined up to discharge to the off-gas line, followed by a 1/2 hour purge with the steam packing exhaust blower and a 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> purge (50 SCFM) with service air fed into the system ahead of the air ejector inter-condenser.
Work Request Authorization forms are used to control routine system maintenance and these typically require special pre-cautions including the use of mechanical cutters or drilling through piping and sampling of gas mixture prior to maintenance on the system.
The off-gas system is not purged routinely on reactor shutdowns.
C.
Quad Cities 2 - No general procedure was in effect.
Practice is to sample the gas concentration if system is to be. opened.
The off-gas system is not purged routinely on reactor shut-downs.
D.
Dresden 1 and 2 - No general procedure was in effect. Any work on the off-gas system is accomplished by a specific work pro-cedure. Maintenance is preceeded by the system purge and samp-ling of the gas mixture.
9 l
r, y. - ---., -
v
(=5=
e
~h 1
8.
3 ATTACHMENI 1 DETKILS ON EXPLOSIONS THAT HAVE BEEN EXPERIENCED A.
Lacrosse Boiling Water Reactor 1.
Date - August 19, 1969 2.
Reactor status prior to event - 100" power, steady s tate.
3.
Indicated stack release rate prior to event - below detectable limit.
4.
Indicated stack release rate (maximum) following event - below detectable limit.
5.
Detection of event - Explosion heard throughout turbine building and control room.
An off-gas low flow alarm was received and off-gas low flow indication was observed in the control room.
6.
Activities in progress i= mediately prior to event - Weekly test l
of off-gas waste compressor begun 20 minutes prior to event..
Compressor was lined up to take suction from turbine and contain-ment building which should not have contained an explosive gas mixture.
Off-gas from main condensor lined up for normal operation, discharging directly to stack af ter passing through 150 f t. 3 delay tank.
7.
Reactor status following event - Reactor continued operation.
Waste compressor stopped.
8.
Inspection and checks made to system - Visual inspection of I
off-gas piping system.
Inspection revealed delay tank %50*F varmer (due to exothermic reaction of recombination of hydrogen and oxygen) than piping to and from the tank 5-10 =inutes af ter the explosion indicating an explosion had occurred in the off-gas system.
9.
Damage to components and system - Grouting broken around delay tank mountings indicated tank movement.
Off-gas f' low meter damaged.
l 5
Q 10.
Cause of Explosion - Not determined, however it was postu-lated that operation of the compressor may have pulled so=a gas from the off-gas system through *a closed valve that was later tested and found to be leaking sligh~tly.
Ignition apparently took place in the was te compressor with the flame wave passing through the leaking isolation valve into the '
off-gas system resulting in the explosion.
11.
Deficiencies observed - Leaking isolation valve between off-gas system and compressor suction.
12.
Corrective actions:
Instructions for operation of the waste conpressor were a.
changed to allow operation only when hydrogen-oxygen recombiner was in operation or the reactor was shut down.
b.
Foreign matter present in the system that loged in the off-gas flow instrument lines following the explosion was removed.
3.
Monticello 1.
Date - November 14, 1971.
2.
Reactor status prior to event - Shutdown prior to event for 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />.
3.
Indicated stack release rate prior to event - Not available.
4.
Indicated stack release rate (maximum) following event -
10,000 uCi/sec (due to condenser vacuum pump flow rather than the explos' ion).
5.
Detection of event - Audible, by maintenance personnel working on system.
6.
Activ1' ties in progress i=cediately prior to event - Mainte-nance personnel welding with an acetylene torch on off-gas system piping.
Off-gas line had not been purged of hydrogen and oxygen gas mixture.
7.
Reactor status following event - Reactor remained shut down.
8.
Inspections and checks made to system:
System instrumentation checked for proper operation.
a.
t
--,-,-,.--r,.
' q;..
b.
Hydrostatic test of tube side of air ejector inter and after condensors.
c.'
DOP tests of absolute filter at and of off-gas line.
d.
Leak test of off-gas delay piping.
e.
UT testing of selected welds in system piping.
~!
f.
Visual inspeccion of piping, hangers, clamps and support beams, including inspection of off-gas system radiation monitoring system.
i g.
Functional test of of f-gas system loop seal isolation
{
valve.
5 9.
Damage to components and systems:
I a.
Both system rupture diaphragms broken (130 psi pressure
[
rating).
i b.
One pressure switch used for initiating isolation of the off-gas system, form the main condensor (following an explosion) was damaged by the pressure pulse of explosion.
i 10.
Cause of Explosion - Failure. to purge system of explosive con-centration of hydrogen and oxygen prior to maintenanca.
11.
Deficiencias observed.
i a.
Pressure sensors used for initiating system isolation were rated for 30 psig service.
Over range pressure protection had not been provided for instruments.
b.
Isolation logic found to reset automatically af ter explosion and allow isolation valves between the main condensor and the air ejectors to reopen following the blowing of the rupture diaphragms, i
t c.
Work Request Authorization form (Work Request) had not been r
prepared for maintenance on system.
d.
No purging of off-gas system prior to maintenance on system.
i I
I t
I i
f i
$$il 1.
12.
Corrective Actions:
a.
Pressure sensors in isolation logic replaced with switches containing over range pressure protection (1250 psig).
i.
Rupture diaphragms replaced.
c.
Isolation logic modified to seal-in and require operator action to reset an isolation condition.
d.
Proc 3 dure developed to purge off-gas system prior to maintenance.
C.
Dresden 1 1.
Date - August 12, 1971.
2.
Reactor status prior to event - 100" power.
3.
Indicated stack release rate prior to event - 40,000 uCi/sec.
4.
Indicated stack release rate (maximum) following evert - Not available.
Rupture diaphragms which blew out relieve to out-side of the turbine building elevation 12 f t.
Release rate at this point was estimated by licensee to be N200,000 uC1/sec.
5.
Detection of event - Control room alarm receivad of CO2 injec-tion into off-gas system.
CO2 injection ecurs on indication of explosion, i.e., system pressure >6 psig.
6.
Activities in progress immediately prior to event - No un-usual activities in progress.
7.
Reactor status following event - Rapid controlled shutdown of reactor.
8.
Inspections and checks made to system:
a.
Visual inspection of off-gas piping.
b.
Visual inspection of filters at end of off-gas delay
- piping, c.
Functional test of instrumentation that initiztes CO2 injection into system.
m.,
m-
~
,-m
5@.Q~
fuy-t.
~,
d.
Leak test of alternate filter bank to detect possible damage.
9.
Damage to components and systems:
a.
Both rupture diaphragms on discharge side of air ejector af ter-condensors blown.
b.
In-service off-gas system filter badly damaged.
Debris from filter blown out into the stack.
10.
Cause of explosion - Postulated to have occurred from static electrical discharge from ungrounded metal rods contained in the off-gas system filter.
11.
Deficiencies observed - None.
12.
Corrective actions:
a.
New rupture diaphragms installed.
b.
New filters installed.
c.
Removal of filter debris from plant stack.
d.
Licensee originally. planned to provide special grounding for metal rods in off-gas system filters.
Plans have since
.been abandoned.
D.
Lacrosse Boiling Water Water Reactor i
1.
Date - June 27, 1972.
2.
Reactor status prior to event - Operating at 9 7* power.
3.
Indicated stack release rate prior to event a 3,000 uCi/sec 4.
Indicated stack release rate (maximum) following' event - Not available, but no increase expected as off-gas was being routed to storage tanks through the recombiner and compressor when explosion occurrred.
5.
Detection of event - Audible indications of explosion through-I out turbine building and control room.
r l
I f
4..
.na eg n
6.
Activities in progress immediately prior to event - Dilution steam to the SJAE at the inlet of the catalytic recombiner was inadvertently turned off N1 minute prior.co explosion.
7.
Reactor status following event - Reactor continued operation at 97% power.
Off-gas flow redirected to discharge to the stack.
8.
Inspections and checks made to sys tem:
a.
Visual inspection of off-gas piping system.
b.
Calibration and functional testing of sys tem instrumen-tation.
9.
Damage to components and system:
a.
Differential pressure gauge across air dryers located upstream from the delay tank found to be ruptured.
This had resulted in leakage of system gas to the ventilation tunnel and subsequently to the plant stack.
Ins trument line was mannually isolated a few minutes af ter explosion, b.
Grouting around supports of delay tank and associated holddown bolts found to be loosened.
c.
Heat transfer tubing in refrigeration dryer was subse-quently found (January 6, 1973) to have been ruptured by the explosion...
10.
Cause of explosion - Inadvertent removal of dilution steam by operator due to f aulty steam flow instrument.
Ignition was postulated to be caused by the electrical preheater coils (for the recombiner) or by the catalytic action in the recombiner.
11.
Deficiencies observed:
a.
Pressure rating of the differential pressure instrument that had been added since original construction across system dryers, was substantially below system piping de-sign pressure of 300 psig.
~
[,
($[')
- 24 b.
Condensate had evaporated from the sensing lines for.the recombiner steam flow instrument.
This resulted in an incorrect (high)~ indication of
- dilution steam flow to-the recombiner.
Incorrect indication res'ulted in complete shutoff of. dilution steam by an operator.
In addition,
- the dilution steam flow instrument was found to be out of calibration.
12.
Corrective actions:
a.
Differential pressura instruments across.sys tem dryers were replaced with instruments of a higher pressure racing.
b.
Procedure for isolating a standby dryer was modified to require both the inlet and outlet isolation valves to be closed to preclude shock damage to the standby dryer from explos' ions.
c.
Dryer replaced with unit rated for 600 psig (originally rated for 300 psig).
-d.
Condensate pots installed on dilution steam flow instru-ment sensing lines to maintain lines full.
Sensing lines also shortened.
a.
Instrumentation and recorders used to monitor the operation of the recombiner woro put on a weekly surveillance schedule to verify proper operation.
f.
A requirement was added that the recombiner-compressor system be functionally tested at monthly intervals for operator training.
E.
Quad Cities j[
1.
Date - tiarch 6, 1973.
2.
Reactor status prior to event - Reactor operating at 100" power.
3.
Indicated stack release race prior to event from Unit 2 -
5,000 uCi/sec.
4 Indicated stack release rate (maximum) following event from Unit 2 - 107,000 uCi/sec.
('
$h E
6
%e 5.
Detection of event - Alarms received from turbine building area radiation monitors.
6.
Activities in progress immediately prior to event - No unusual activities in plant.
Electrical storm being experienced in the vicinity of plant.
7.
Reactor status following event - Reactor continued operation at N100% power.
Air ejector with bloun rupture diaphragm isolated manually in N35 minutes.
8.
Inspections and checks made to system:
a.
. Visual inspection of accessable portions of system piping.
b.
DOP tested filters at end of delay line.
c.
Conducted radiation surveys outside plant to insure off-Cas delay piping was not leaking.
d.
Inspected filters.
e.
Functionally tested isolation ins truments.
9.
Damage to system components:
a.
Both off-gas filters burned and blown through.
b.
Rupture diaphragm broken downstream of 1 air ejector af ter condensor.
10.
Cause of explosion - Postulated to have resultad from lightning strike at or near the plant.
11.
Deficiencies observed:
a.
Filter " catchers" had not been installed during plant cons t ruction.
The " catchers" should have been installed to trap pieces of filters (not designed for explosions) in off-gas system to provent release of radioactivity contained on filters.
b.
Both filters in off-gas system were in service at tica of explosion.
System should have been operated with 1 filter in service, and 1 in standby, s
r
. e.
,=
g...
- 2o -
c.
Off-gas system either did not isolate, or the isolation condition reset automatically..
12.
Corrective actions:
a.
Blown rupture diaphrag=s replaced.
b.
Filter " catchers" ins talled.
c.
Damaged filters replaced and DOP tested satisfactorily.
F.
Dresden 2 1.
Date - March 27, 1973.
2.
Reactor status prior to event - Shut down for maintenance on March 24, 1973.
3.
Indicated stack release rate prior to event - 0 4.
Indicated stack release rate (maximum) following event - Not available.
5.
Detection of event - Audible.
6.
Activities in progress immediately prior to event - Purging of off-gas system in progress.
Purge involved use of temporary ductwork and pulling air backwards through an of f-gas filter -
(opened to atmosphere) and through the gland seal exhauster.
Welding activities in progress in steam jet air ejector room in the vicinity of the temporary ductwork.
The temporary ductwork collapsed when the gland seal exhauster was started and allowed off-gas to enter this area resulting in the explosion.
7.
Reactor status following event - Reactor remained shut down.
8.
Inspections and checks made to system:
a.
Visual inspection of system piping and gland steam exhauster.
b.
Visual inspection of of f-gas filters and housings,
c.
Recalibration of off-gas system flow instrumentation and j
inspection of off-gas flow measuring orifice.
ifE.
'S
=,::
9.
Injuries to personnel and damage to components and system:
s.
Two maintenance personnel outside the-SJAE room received minor injuries.
b.
Off-gas system filters damaged and blown free of filter housing that was opened to atmosphere.
10.
Cause of explosion - Ignition resulted from welding activity outside the steam jet air ejector room when a temporary purge ducting collapsed ar.d permitted combustible gases to enter the area.
11.
Deficiencies observed - Purging procedure did not provide for controls to exclude other activities, such as welding, in the vicinity of purge route.
- 12. Corrective actions:
,a.
Off-gas filters replaced and DOP tested.
b.
Procedures modified to require roped-off exclusion areas in locations that have a potential for combustible gas accumulations.
(5; dih; a
AITACHMENT 2 DETAILS ON OFF-CAS SYSTEMS AT FACILITIES EXPERIENCING EXPLOSIONS A.
LACBWR - The off-gas system consists of one set of steam jet air ejectors (SJAE's) and condensors that take suction from the = sin condensor through manual isolation valves.
Non-condensable of f-gas then passes sequentially through 1 of 2 parallel, refrigera-ted dryers, a 150 cubic foot, 10-minute delay tank, charcoal filters, and is then selectively routed (from the control room) to either the stack plenum, where it is diluted with ventilation air, or to the recombiner systen.
The recombiner system consists of a 5JAE, supplied by the heating steam system, that diluces and drives off-gas through an electrical preheater, a catalytic recombiner and a recombiner condensor.
Following the exit from the recombiner con-densor, the gas can be compressed and stored in either of two 1600 cubic foot storage tanks or discharged to the stack plenum and released af ter dilution with ventilation air.
A mechanical vacuum pump is provided to establish a vacuum in the main condensor on start-ups.
The vacuum pump takes suction through manual isolation valves and discharges through a reparate line to the stack plenum.
Infor-mation obtained on the off-gas system includes:
1.
Design pressure and temperature:
n.
Off-gas piping 300 psig, 250*F b.
Delay tank 150 psig, 650*F c.
Gas storage tanks 350 psig, 650*F 2.
Design flow rate - 15 SCFM.
3.
Discharge path and dilution air supply - Off-gas normally directed to stack plenum where it is diluted with ventilation air from two 35,000 cfm fans.
4.
Isolation features provided for explosions - No automatic isola-tion features provided.
5.
Explosion sensing devices - No special instrumentation provided.
6.
Overpressure relief devices - Relief valves located at the air ejector af ter condensor, compressor dicharge, recombiner condensor, and compressor af ter cooler.
7.
Manual isolation provisions - Manual isolation valves in the vicinity of the SJAE's.
y
,1 flll@i-c
+
= 8.
Concentration of combustibles, assumed in system design - 70%
of volume is hydrogen and oxygen in stoichiometric quantities.
9.
Filters - Information not available on filters.
10.
System designer - Sargent and Lundy to Allis Chalmers speci-fications.
B.
Montien11o - The off-gas system consists of two sets of SJAE's and condensors that take suction from the main condensor through four air-operated, parallel isolation valves.
- on-condensable off-gas then passes sequentially through a 30-cinute holdup line that is buried underground outside the plant, 1 of 2 parallel absolute filters, an automatic closing isolation valve (actuated by high radiation in off-Eas systen), and is then discharged into the t ase of the stack where it is diluted with ventilation air.
A mechanical vacuum pump is provided to establish a vacuum in the main condensor on startup.
The vacuum pump takes suction through the same air-operated isolation valves used for the SJAE's and normally discharges directly to the stack. The discharge of the mechanical vacuum pump can also be routed through the 30-minute delay -line and filters before release to the stack.
Information obtained on the off-gas system includes:
1.
Design pressure of system piping - 350 psig.
2.
Design flow rate - 166 CFM @ 130*F, 1 atomsphere.
3.
Discharge path and dilution air supply - Of f-gas discharged to stack. dilution chamber where it is diluted with ventilation air from the turbine building exhaust fans.
Isolation features provided for explos' ions - Systam automatically 4
isolates air ejection suction from main condensor following an explosion.
5.
Explosion sensing devices 2 temperature and 2 pressure instru-ments.
Alar =s also provided.
6.
Overpressure relief devices - Rupture diaphrsges provided at the discharge of each air ejector af ter condensor and at the dis ~harge c
of the mechanical vacuum pump.
5%
s"* -
t,
- p 7.
Manual ' isolation provisions - Individual air ejectors and con-densors can be manually isolated during plant operation by means of valves, that have operating handle's that, pass through the shielding wall of the SJAE rooms.
A single blown rupture diaphragm cannot be isolated during operation as no isolation valves are provided in the off-gas piping between the diaphragms.
Sys' tem can also be isolated from the control room.
8.
Concentrations of combustible gases assumed in system design -
60% of volume is hydrogen and oxygen in stoichiometric quantities.
9.
Filters - Not designed to survive explosions - failed filter catchers installed downstream of each HEPA filter. Filter catchers designed for 350 psi differential pressure from either side.
10.
System designer - Bechtel to GE specifications.
C.
Quad Cities 2 and Dresden 2 - The off-gas system is similar to the off-gas system at Monticello in terms of major equipment and flow path, however, there is no means provided at Quad Cities 2 or Dresden 2 to direct the discharge of the mechanical vacuum pump into the inlet of the 30-minute delay line.
Information obtained on the off-gas system includes:
1.
Design pressure of piping - 300 psig.
2.
Design flow rate - 200 CFM @ 130*F, 1 atmosphere.
3.
Discharge path and dildtion air supply - Off-gas discharged to -
stack dilution chamber when it is diluted with ventilation air from turbine building exhaust fans and 2 rsdwaste building exhaust fans.
i 4.
Isolation features provided for explosions - System automatically isolates air ejector suction from main condensor following explosions.
5.
Explosion sensing devices - 2 temperature and 2 pressure instruments.
Alarms also provided.
6.
Overpressure relief devices - Rupture diaphragms located at the discharge of each air ejector after condensor.
7.
Manual isolation provisions - Individual air ejectors and associated rupture diaphragms can be isolated during plant operation by means of manual isolation valves that have operating handles that pass i
l h
through the shielding wall of the SJAE rooms.
System can also be isolated from the control room.
8.
Concentrations of combustible gases assumed in system design -
70% of volume assumed to be hydrogen and oxygen in stoichiocotric quantities.
9.
Filters - Not designed to survive explosion - Failed filter
" catchers" provided at QC-2.
Information not available on filters or filter catchers at D-2.
10.
System designer - Sargent and Lundy to GE specifications.
D.
Dresden 1 - The off-gas system consists of two sets of SJAI's and condensors that take suction from the main condonsor through motor operated valves.
Non-condensables from the main condensor then pass sequentially through the holdup delay piping, parallel absolute filters, an automatic isolation valva (actuated by high radiation in the off-gas system) and are then discharged to the plcat stack where ic is diluted with ventilation air.
A mechanical vacuum pump is provided to establish a vacuum in the main condensor on startup.
The vacuum pump takes suction through an air operated valve and discharges into the delay line used by the SJAE.
Infor-mation obtained on the off-gas system incudes:
1.
Design pressure of piping - 450 psig, (50*F).
2.
Design flow rate - 72.5 SCFM.
3.
Discharge path and dilution air supply - Of f-gas discharge to D-1 stack where it is diluted with exhaust ventilation air.
4.
Isolation features provided for explosions - No automatic isolation features provided.
CO2 system is provided to inject on system overpressure that would result from an explosion.
5.
Explosion sensing devices - Single pressure switch.
Alarm also provided.
6.
Overpressure relief devices - Rupture diaphragms at discharge of SJAE af ter condensors, 7.
Manual isolation provisions - Motor operated valves provided to isolate an air ejector and associated condensors.
Manual valves are provided that could isolate a single blown ruptura diaphragm.
s v-Mii.;.%.
8.
Concentrations of combustible gases assumed in system design -
Not available.
9.
Filters - No information available on capability to withstand an explosion.
Based on the experience of the August 12, 1972 explosion, it appears there are no failed filter catchers and filters are not designed to withstand explosions.
10.
System designer - Not available.
6
c G
i{!%
ATTACHNENT 3 DETAILS ON EXPLOSION SENSING DEVICES AND ISOLATION LOGIC A.
LACBWR - No sensing levices or isolation provisions provided.
B.
Quad Cities 2 1.
Pressure sensor - Two provided.
s.
Type and Manuf acturer - Bellows type, Barton Model 2SSA.
b.
Range 60 psig.
c.
Actuation setpoint - 10 psig, d.
Pressure sensing point in system - Downstream of air ejector after condensors.
e.
Standard working pressure of instrumant - Estimated to be in excess of 500 psig, f.
Alarms - 10 psig.
2.
Temperature sensors - Two provided.
a.
Type and Manuf acturer - Liquid-filled, s tainless-steel capillary tube-United Electric controls,1200 Series,
Model 6BS.
b.
Range 250*F.
c.
Actuation setpoint - 140*F.
d.
Sensor location in system - Downstream of air-ejector after condensors.
e.
Standard working pressure of instrument - Not availab le,
f.
Alarms - 140*F.
3.
Logic for isolation - Both temperature channels or, both pressure channels.
4.
Reset feature of isolation circuitry - Isolation will reset auto-matically when trip channels reset.
n,,
--,.n._
-, - ~ - - -
..D
- T'.
m
= =.
34 -
5.
-Calibration and functional tast frequency of pressure and teep-erature sensors - None established at present time.
Licensee representatives stated, however, that instruments would probably be checked on a frequency of 1/ year.
l C.
Dresden 2 1.
Pressure sensors - Two provided.
a.
Type and Manufacturer - U. S. Gauge Ametek Model 887A i
b.
Range 200 psig.
c."
Actuation setpoint - 10 psig.
I d.
Pressure sensing point in system - Downstream of air ejector after condensors.
e.
Standard working pressure of instrument - Not available.
f..
Alarms - 10 psig.
I 2.
Temperature sensors - Two provided.
I Type and Manufacturer - SS capillary tubes - U. S. Gauge a.
Ametek Model 3056.
i b.
Range - 32*F - 220*F.
c.
Actuation setpoint - 140*F.
I d.
Sensor location in system - Downs.tream of air ejector after f
condensors.
I i
e.
Standard working pressure of instrument - Not availab le.
f.
Alarms - 140*F.
3.
Logic for isolation - Two temperature or two pressure channels.
4.
Reset feature of isolation circuitry - Isolation resets auto-matica11y when channel trips clear.
A change is planned to provide seal-in feature requiring operator reset.
l t
5.
Calibration and functional test frequency of pressure and tempera-j ture sensors - Once per refueling cycle.
j t
I t
I
e O
4==.
c.:n..
g-D.
Monticello
~
1.
Pressure sensors - Two provided.
a.
Type and Manufacturer - Diaphragm type - Ashcrof t.
b.
Range - 1, 0-15 psig.
1, 0-30 psig.
c.
Actuation setpoint - 5 psig.
d.
Pressure sensing point in system - Downstream of air ejector af ter condensors,
e.
Standard working pressure of instrument - 2500 psig.
1250 psig ins trument over range protection, f.
Alarms - 5 psig.
2.
Temperature sensors - Two provided.
a.
Type and Manuf acturer - Not available.
b.
Range - Not available.
c.
Acuation setpoint - 130*F.
d.
Sensor location in system - Downs tream of air ejector af ter condensors.
e.
Standard working pressure of instrument - Not available, f.
Alarms - 130*F.
3.
Logic for isolation - Any combination of two of the four tempera-ture and pressure channels.
4.
Reset feature of isolation circuitry - Requires manual reset following isolation.
5.
Calibration and functional test frequency of pressure and tempera-ture sensors - Not availab le.
l l
P
a, Ei..n E...
, E.
Dresden 1 1.
Pressure sensor - One pressure switch provided to actuate at N10 psig.
Instrument alarms and initiates injection of CO., into off-gas system at the beginning of the holdup pipe.
Once accu-ated, the CO2 injection system must be reset by operator action.
4 4
4
?.
Gb O/S r
ATTACHMENT 4 DETAILS ON OFF-GAS SYSTEM OVERPRESSURE RELIEF DEVICES A.
LACBWR Set Location 4
Type Device Pressure in System Relieves To Relief valve N.A.*
Air compressor Air compressor suction dis charge Relief valve N.A.
A2.r compressor Stack, via off-gas after cooler discharge line Relief valve N.A.
Gas storage Stack tanks Relief valve N.A.
Of f-gas holdup Stack via off-gas line tank Relief valve N.A.
Air-ejector Stack, via discharge ling after condensor from mechanical vacuum pump
- N. A. - Information not available B.
Monticello Set Location Type Device Pressure in System Relieves To Rupture 130 psid Off-gas discharge SJAE room, then to the diaphragms (2) line from air-stack via ventilation ejector af ter system condensors Loop seals 1 psid Collect mois ture Turbine building from various points equipment su=p in off-gas system C.
Dresden 1 Set Location in Type Device P ressure System Relieves To Rupture 28 psid.
Off-gas discharge Outside turbine diaphragms (2) line from SJAE building at elevation after condensors 12' r
~.
.--n
Qf)
($
D.
Quad Cities 2 and Dresden 2
~
Set Location Type Device Pressure System Relieves To Rupture 35 psid Off-gas discharge SJAE room, then to diaphragms (2) line from SJAE the stack via af ter condensors ventilation systen r
Loop seal 1 psid Collects moisture Main condensor from various points in off-gas system Loop seal 1 psid Collects moisture Radwaste equipment from filters drain sump Relief 28 psid SJAE after Main condensor valves (2) condensors L
1 a
1
__.