ML19275A954
| ML19275A954 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 10/31/1979 |
| From: | Lawroski H PRESIDENT'S COMMISSION ON THE ACCIDENT AT THREE MILE |
| To: | |
| Shared Package | |
| ML19254E707 | List:
|
| References | |
| NUDOCS 7910300403 | |
| Download: ML19275A954 (75) | |
Text
,
TECHNICAL STAFF ANALYSIS REPORT ON TRANSPORT OF RADIOACTIVITY FROM THE TMI-2 CORE TO THE ENVIRONS TO PRESIDENT'S COMMISSION ON THE ACCIDENT AT THREE MILE ISLAND ADVANCE COPY NOT FOR PUBLIC RELEASE BEFORE AMs, WEDNESDAY, OCTOBER 31, 1979 i
79/o3coyo3 1182 361
TRANSPORT OF RADIOACTIVITY FROM THE TMI-2 CORE TO THE ENVIRONS BY HARRY LWROSKI, CONSULTANT TECHNICAL ASSESSMENT TASK FORCE OCTOBER 1979 WASHINGTON, D.C.
4 1182 062
This document is solely the work of the Commission staff and does not necessarily represent the views of the President's Commission or any member of the Commission.
This pre-publication copy is a final document and will be subject only to minor editorial changes in its published form.
1182 063
TABLE OF CONTENTS Page AC KNOWLEDGEMENTS EXECUTIVE
SUMMARY
ES-1 l-2 1.0 PURPOSE AND
SUMMARY
1-2 1.1 Purpose 1-2 1.2 Summary
2.0 DESCRIPTION
OF ACCIDENT-
3.0 DESCRIPTION
OF CORE BEHAVIOR 3-1 4.0 PATHWAYS FROM THE REACTOR PRIMARY SYSTEM 4-1 4.1 Letdown /Make-Up System 4-1 4-4 4.2 Reactor Building Sump to the Auxiliary Building Sump 4.3 Reactor Coolant Drain Tank to Reactor Coolant Bleed 4-6 Holdup Tanks 4.4 Reactor Coolant Drain Tank Vent to Vent Gas Header 4-8 in the Auxillery Building 4.5 Reactor Coolant Drain Tank Vent to the Reactor Coolant 4-8 Bleed Holdup Tanks 4.6 Reactor Coolant Pump Seals to Seal Return Coolers in the 4-11 Auxiliary Building 4-11 4.7 Letdown Coolers Cooling Water 4-13 4.8 Leakage Coolers Cooling Water 5.0 INTERCONNECTION OF LIQUID AND GAS SYSTEMS 5-1 6.0 RADIATION MONITORING IN THE PLANT 6-1 7.0 POTENTIAL PRESSURE TRANSIENTS IN EACH SYSTEM 7-I 7.1 Letdown /Make-Up System 7-1 7.2 Reactor Building Sump to the Auxiliary Building Sump 7-2 l182 064
TABLE OF CONTENTS (cont.)
Page 7.3 Reactor Coolant Drain Tank to the Reactor Coolant Bleed Holdup Tanks 7-3 7.4 Reactor Coolant Drain Tank Vent to Vent Gcs Header in the Auxillary Building 7-4 7.5 Reactor Coolant Drain Tank Vent to Reactor Coolant Bleed Holdup Tanks 7-6 7.6 Reactor Coolant Pump Seals to the Seal leturn Coolers in the Auxiliary Building 7-6 7.7 Letdown Coolers Cooling Water 7-7 7.8 Leakage Coolers Cooling Water 7-7 7.9 Interconnection Pressure Transients 7-7 8.0 RESPONSE OF RADIATION MONITORS DURING THE ACCIDENT 8-l 9.0 DISCUSSION AND RATING OF POTENTIAL PATHWAYS 9-1 10.0 MOST PROBABLE PATMWAY 10-1 APPENDIX A - DISCUSSION OF METROPOLITAN EDISON / GENERAL PUBLIC UTILITIES CORPORATION'S PRELIMINARY REPORT ON SOURCES AND PATHWAYS OF THI-2 RELEASES OF RADIOACTIVE MATERIAL, July 16, 1979, REVISION 0 A-1 APPENDIX B - DISCUSSION OF THE NUCLEAR REGULATORY COMMISSION'S EVALUATION OF RADIDACTIVE RELEASE PATHWAYS B-1 APPENDIX C - COMPARISON OF RATING OF PATHWAYS OF RADIOACTIVITY TO THE ENVIRONS C-1 REFERENCES R-1 1182 065'
TABLES Page 1
List of Radiation Monitors and Locations - Process Monitors 6-2 11 List of Radiation Monitors and Locations - Area Monitors 6-3 111 List of Radiation Monitors and Locations - Airborne Monitors 6-4 IV Early Radiation Monitoring Responses, March 28, 1979.
8-2 V
Other Radiation Monitoring Responses, March 28, 1979 8-5 1l82 056
T -
48.
FIGURES Page 4-2 1.
Letdown / Makeup System 2.
Reactor Building Sump to Auxiliary Building Sump System 4-5 3
Reacter Coolant Drain Tank to Reactor Coolant Bleed Holdup Tanks 4-7 4.
Reactor Coolant Drain Tank Vent to Vent Gas Header in Auxiliary 4-9 Building 5
Reactor Coolant Drain Tank Vent to Reactor Coolant Bleed Holdup 4-10 Tanks 6.
Reactor Coolant Pump Seals to Seal Return Coolers in Auxiliary 4-12 Building 7.
Vent Gas / Relief Valve Header System 5-2 8.
Liquid / Gas Systems Interconnections 5-3 1182 067
EXECUTIVE
SUMMARY
The major radioactive releases from the Three Mile Island (TMI)- 2 accident were airborne noble gas fission products, xenon and krypton, as well as a small fraction of the radioactive iodine isotopes. These isotopes, in addition
.o other fissior. products, were dissolved in the reactor primary coolant water. It is believed that the major pathway of radioactivity release from the primary system to the auxiliary building was through the reactor coolant letdown /make-up system.
TL pressurization of the reactor coolant drain tank by the blowdown from the pilot operated relief valve (PORV) pushed water from the drain tank through t,e reactar buildir.g vent header into the auxiliary building vent header. It is believed that there was sufficient pressure imposed on the auxiliary building vent header to damace sor.;e component, probably the water traps incorporated for drainage of tne system. The damaged components were future leakage points.
The dissolved gases were released from coolant during depressurization of the letdown fluid. The relief v3.ves just downstream of the block orifice valve (pressure reduction of 2135 pounds per squart inch guage(psig)to about 20 psig) and of the make-up tank discharged.into.the reactor coolant bleed hold-up tanks. This in turn pressurized the reactor coolant bleed boldup tanks and the pressure relief valve on these tanks lifted, venting the gases to the relief valve vent header.
The relief valve vent header has a dircret and unencumbered pathway to the station vent. The make-up tank also was vented to the vent header in the auxiliary building. The radiation monitaring systems showed a direct correla-tion of these ventings with releases pro 4.bly through damged components in the vent header system. The ventilation systems of both the auxiliary building and the fuel handling building transported the released gases out the station vent.
The following observations should be noted:
ES-1 1182 068
1.
The discharges of pressure rel.ef systems that communicate with the primary coolant were not routed to the reactor containment system. Examples are the discharge of the relief valves of the reactor coolant bleed holdup tanks and of the waste gas compressors.
2.
The reactor containment building is not isolated on radiation signals.
This isolation probably would not have precluded the airborne radioactive releases at TMI-2.
3.
It appears that inadequate attention was given to design to assure matching component capabilities.
For example, it is believed that the water traps WCG-U8A and WDG-U9A had lower pressure capabilities than the pressure relief valve WDG-R-3 of the vent header system 4
The concrete in the auxiliary beilding and the fuel handling building was not sealed prior to startup.
5.
Readily accessible up-to-date, readable drawings and specifications were nct available on site.
55-2 1182 069
1.0 PURPOSE AND SLWiARY 1.1 Purpose The purpose of this report was to prepare an evaluation as to how and why the radioactive of the TMI-2 accident got from the reactor to the auxiliary Du11 ding and fuel handling building and out of the station.
1.2 Summary The report presented here is an evaluation of how the radioactivity (a) left the reactor core, (b) was transmitted to the auxiliary buildin', and (c) was exhausted g
to the atmosphere.
The reactor accident was basically caused by not providing adequate cooling to the reactor core. The fuel cladding was extensively damaged and released all of the gaseous fission products, from the clad to fuel gap, to the primary coolant system. A part of those gaseous fission products are now in the reactor contain-fraction of those gases escaped to the auxiliary building.
ment building, and a
Those gases that were transported to the auxiliary building were either released to the atmosphere or routed to the waste decay tank. Part of the gases in the waste decay tanks were transferred back to the reactor containment building There are eight pathways for the rel?ases of airborne fission products discussed.
Other pathways were reviewed and covered in referenced documents. These other path-ways were not included here because they were not worthy of detailed study and were considered less significant. Of the eight considered here, only the following Tive had significant potential:
(1) Reactor Ccolant Letdown /Make-Up System - This was the major pathway and the source of gas pressure buildup in the auxiliary building.
(2) Reactor Coolant Drain Tank Vent to the Vent Header in de Auxiliary Building - This pathway was believed to be the cause of damage to the vent header, thereby setting up significant leak paths for gases vented to the vent header.
The remaining three are considered substantially less significant.
(3) Reactor Coolant Drain Tank to the Reactor Coolant Bleed Tanks in the Auxiliary Building - This added water to the inventory in the auxiliary building.
(4) Reactor Coolant Drain Tank Vent to the Reactor Coolant Drain Tanks - After the rupture disk blew in the drain tank, there was little pressure differential to push activity from the reactor containment building.
Il82 070 1-2
(5) Reactor Building Sump to the Auxiliary Building Sump - This pathway contributed to excess water in the auxiliary building but had very little radioactivity because the water was released prior to significant core damage.
Simplified schematics of the above pathways were prepared and included in this report. To give a better understanding, drawings showing the interconnections of liquid and gas systems and the overall pathways systems were prepared and included.
A description of the radiation monitoring in the plant is provided. The eavluation of the response of the radiation c.onitoring showed that a very small release occurred early in the accident. After approximately 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> to 30 minutes after the core was partially uncovered, larger rele'.ses of radioactivity were evidenced. These releases continued for some time with the majority of the activity released in the first several days.
The most probable release pathway was through the reactor coolant le.down/
make-up system. These releases were effected primarily by degassing of the primary coolant water in the letdown /make-up system and either going directly to the station vent from the pressure relief valves of the reactor coolant bleed holdup tanks or by venting gases from the make-up tank to the vent header in the auxiliary building. The reactor coolant bleed holdup tanks were the recipient of pressure relief valves from the block orifice valve outlet and from the make-up tank.
Based on the evaluation developed in this report the following observation is made: The discharges of pressure relief systems that comunicate with the primary coolant were not routed to the reactor cot.taincettt syste::1.
I.m ples are the discharge of the relief valves of the reactor coolant bleed holdup sanks and the vaste gas comprecsors.
1182 071 1-3
2.0 DESCRIPTION
OF ACCIDENT Ref.
The Nuclear Regulatory Commisslan (NRC) (NUREG-0600), Metropolitan Edison (Met Ed) j General Public Utilicies (GPU) (Preliminary Annotated Sequence ot Events, March 25, 1979; July 16, 1979, Rev.1) and tt.e Nuclear Safety Analysis Center of the Flectric 2
Power Research Ins:itute (EPRI) (NSAC-1) have published extsnsive descriptions of tl.e accidant. The discussion hcre is to highlight events that might have consequences related to the release of radioactivity..
The plant was opcrating at 97 percent power at 4:00:37a.m. on March 28,1979. Reactor 3
primary coolant system pressure was2,155 psig. Reactor coolant make up pump B (MU-P-lB) was in service supplying make-up and reactor coolant pump seal injec-tion flow. Normal reactor coolant system letdown flow was approximately 70 gallons per minute.
Flow Recorder FR 7100 indicated that every 2 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> 4
liquid was being pumped from the reactor coolant drain tank to the reactor coolant bleed holdup tanks, 6ndicating that the cooling system was also in operation. The reactor primary coolant had about 0.337 pCi/cc radioactivity.
5 There were problems in transferring resins in the standby demineralizer of the condensate polishing system. The fuel handling building supply exhaust fans were in service. The auxiliary building exhaust fans were in service.
The status of the auxiliary building supply fans is not known.
The fallowing is an abbre viated sequence of events as reported in reference documents.
ITEM EVE:4T (Tine af ter Feed Water Pump Trips) 0 Feed water pumps trip; main turbine and main generator (4:00:37) trip 3 see Electromatic relief valve opens (2,255 psig) 4 see Pressure started to incresse in reactor coolant drain tank 8 see Reactor trips on high pressore (2,346 psig) 1182 072 2-1
m T
EVENT (TimegyterFeed Water Pump Trips)
Ref.
12 sec Signal that electromatic valve should have close'i.
13 sec.
Make-up pump IA (HU-P-I A) was started, and a higli pressure valve was opened. (Make-Up Pump IB was still operating')
30 sec The reactor coolant low-pressure trip setpoint was reached (reactor pressure at lS40 psig).
39 sec Make-Up Pump 1A tripped.
41 sec Make-Up Pump IA was restarted.
60 sec Reactor coolant drain tank pressure at 12 osig and increasing.
I sain 26 dec Reactor coolant drain tank temperature at 85.50F and increasing.
2 min I sec High pressure injection system automatically started at reactor primary coolant pressure nominal setpoint of 1,600 ps i g.
Make-Up Pump 18 tr ips automatically, and Make-Up Pump IC starts. Make-Up Pump 1A is still running.
3 min 13 see Reactor coolant drain tank relief valve li f ted at 120-122 psig.
2 min 13 see Manual bypass of high pressure injection system controls.
3 c.i r 26 sec Reactor coolant drain tank high-temperature alarm. (As stated in the introduction, the cooling system was in operation, but evidently could not keep uo with the heat input from the escaping water of the primary system through the e lectromati c rel ie f va lve).
k T'o 38 s =c High pressure injection throttled.
4 min 53 sec Letdown flow increased to rate greater than 160 gallons per 6
minute.
(This is accanplished by remotelv opening a bypass valve, MU-US) around the block ori fice valve MU-1-E in the reacter coolant letdown /make up system.)
i i82 073 2-2
ITEM EVENT (Tire af ter Feed Ref.
Water Pump Trips) 6 min 54 sec Let-down cooler l A outlet temperature alarms high at 139 F.
6 min 58 sec Reduced letdown flow to 71.4 gallons per minute ay closing bypass valve MU-U-5 7 min Reactor containment building purge air radiation gas monitors HP-R-225 and HP-R-226 indicate small increases 7
in radioact ivi ty.
7 min 29 see Reactor buildi~ng sump pump A (WDL-P-2A) starts.
8 min 18 sec Emergency feedwater block valves were opened.
10 min Reactor building sump pump B (WDL-P-2B) started.
10 min 48 sec Reactor building sump high-level alarm (4.65 ft) 14 min 48 sec Reactor coolant drain tank rupture disc WDL-U26 failed at 8
192 psig.
(This is an 18-inch vent line.)
15 min Radiation monitor on hydrogen purge HP-R-229 alarmed on 9
iodine channel.
16 min Reactor building air sample radiation nonitor HP-R-227 10 alarmed on gas, particulate and iodine channels 9
19 min Station vent monitor HP-R-219 alarmed on the gas erannel.
22 min Radiation monitors HP-R-221A (particulate), HP-R-221B 9
(particulate), HP-R-225 (gas), HP-R-226 (gas), HP-R-227 7,10 (gas), and IC-R-1092 showed increased radiation levels.
11 28 r.;n Radiation monitor HP-R-222 (gas). The evaluation by NRC 7
and Met Ed included responses by several other monitors showing the same pattern of increases.
3E min Reactor building sump pumps A and B were stopped.
60 min Letdown cooler A radiation monitor IC-R-1092 increased.
11 I hr 13 min Reactor coolant pumps IB and 23 were stopped.
1 hr 23 min Letdown cooler A radiation monitor IC-R-1092 increased 11 again.
I br 41 min Reactor coolant pumps IA and IB were stopped.
i182 074
'~'
~.
ITEM EVENT (Time after Feed Water Pump Trips)
I hr 51 min Reactor coolant loops A and B hotleg temperatures increasing.
10 2 hr 22 min Reactor containment building air sample monitor HP-R-227 gas channel starts increasing again.
2 hr 22 min Electromatic relief block valve RC-V2 was closed for the first time.
2 hr 22 min Reactor coolant system pressure begins to rise (680 to 2130 psig over the next 41 minutes).
2 hr 25 min HP-R-227 (see item 37) particula'te channel started to 10 increase again.
2 hr 31 min Radiation monitors start to rise in the reactor containment 7,9 building followed by radiation monitors rising in the 10,Il auxiliary building about 10 minutes later.
(See Section 8.0 12,13 of this report for specif c radiation monitors and times.)
2 hr 54 min Started reactor coolant pump 28.
2 hr 56 min site emergency declared.
3 hr 12 min Electrometic relief block valve RC-V2 was opened and was closed at 3 hr 17 min.
3 hr 19 min High pressure injection was started manually.
3 hr 23 min General emergency declared.
3 hr 29 min Fuel handling building air exhaust fans flow was zero.
NOTE: During the next 2) hours, the exhaust fans were 14,15 turned on and off several times with run times of 30 to 60 minutes.
3 hr 30 min Electromatic relief block valve RC-V2 was closed.
3 hr 51 min Electromatic relief block valve RC-V2 was opened.
3 hr 56 min Reactor containment building isolated by high pressure signal (approximately 4 psig). Each isolation valve must be re-set by operator action to put any system back into service which penetrates reactor containment.
From the time at which the reactor containment building was initially iso-lated at 7:56 a.m.,
it appears that the only pathways for radioactivity to leave the 2-4 1182 075
reactor containment building were through the letdown /makeuo system including the reactor coolant pump seals.
There were radiation releases by depressurization of the letdown water of such dissolv 1 gases as hydrogen, krypton, and xenon. Some fraction of the radioactive iodine also became airborne from the water. The evolution of dissolved gases pressurized the letdown /make-up system, including the make-up tank and the reactor coolant bleed holdup tanks. The released gases were also transmitted to the wasta gas decay tanks.
1132 076
-s
30 DESCRIPTION OF CORE BEHAVIOR The following discussion i: primarily f rom the standpoint of core behavior and its.:lationship to the release of radioactivity, particularly fission gases and volatile fission products. This discussion is also qualitative and is based on what one would expect sen a hot core became uncovered in'a steam atmosphere.
Early in the accident, stress conditions of the fuci pins appeared to cause a release of a small amount of radioactivity that was picked up by radiation 7
monitors in the plant (see Section 8.0).
This initial rclease for all intents and purposes was insignificant from a health standpoint.
Little further damage was imposed upon the core until the lack of liquid phase cooling water became irrportant at about 6:00 a.m.to 6:15 a.m. on March 28, 1979.
At this time,with the reactor coolant pumps off and the primary coolant inventory insuf ficient, the core started to become uncovered.
At near 5:45 a.m., the out-of-core nucl ear i nstrumentat ion indicated increased 16 flux level that can be rationalized as decreased coolant in the core and downcomer annulus. The temperatures in the hot legs of both Loops A cnd 8 of the reactor 3
primary coolant system were increasing at about 5:50 a.m. and went greater than 620 F.
Up to this time, the core cladding would have normally stayed within 0
several hundred degrees - Farenh:it of the water-steam mixture. However, when the to remove heat by vaporization, cladding temperatures liquid phase is not present rise to improve the lost heat-transfer condition.
When the cladding temperature of Zi rcaloy reaches about i,400 to 1,500 F, f a i l ure 17 or cladding breach becomes a reality. The breaching of the cladding releases fission gases not trapped in the fuel matrix. The fission gases would be krypton, 18 xenons, an d a t the >e tempera tures, iodine also would be a gas.
Since decay was still signi ficant, the core fuel would be expected to increase in temperature with the lack of suf ficient cooling.
That was indeed the case at TMI-2.
At about 2,200 F, reaction of Zircaloy with water molecules 17 (Zr0 ) and hydrogen (H )-
becomes rapid, forming zirconium oxide 2
2 1182 077 3.,
The zirt..nium oxide stays as a solid. However, the H is distributed between 2
the liquid still present and the gas phase. At the reactor primary system presures and temperatures, the equilibrium solubility of H2 is relatively high.
18 At i,543 psi g and 600 F, the mass concentration ratio of hydrogen between gases and water is 67.5 pg H /g gas /pg H /g liquid.
2 2
The estimates of chemical reaction of the Zircaloy of the core witn the water 17 is 40 to 50 percent.
The volume of hydrogen that was generated far exceeded the volume of fission gas.
For all practical purposes, the hydrogen became the carrier gas for the fission product gases to escape out of the primary system water.
Both the fission gas and hydrogen have significant solubility in water. When the letdown water, which had these dissolved gases present under pressure, was cooled and depressurized through the block ori fice valve, the dissolved gases (primarily hylrogen)were released and pressurized the auxiliary building portion of the reactor coolant letdown /make-up system.
These increased pressures were the primary driving force for the fission gas releases. The releases were ef fected through li f ted relief valves, damaged components or planned vent-6 ing of specific tanks in the auxiliary building.
1182 078 3-2
4.0 PATHWAYS FROM THE REACTOR PRIMARY SYSTEM TO THE AUXILI ARY BUILDING 4.1 Le:down/Make-up System The letdown /make-up system is the major control system for water chemistry, water conditioning, and maintenance of water inventory in the reactor coolant primary system. Control of the letdown /make up system ca'n be either manual or automatic. During reactor operations or standby, the letdown /make-up system can be and is used to furnish conditioned make up water for leaking seals or valves in the reactor primary cool' ant system. A schematic of the letdown /make-up 6
system is shown in Figure 1.
During normal operation, water is removed from the reactor primary coolant system f rorr the 28-inch diameter line between the Loop A steam generator RC-H-1 A 19 and Reactor Coolant Pump RC-P-1A. The 21/2 inch - diameter stainless steel pipe.
(Schedule 160) transfers the water nominally at 5500F to che letdown coolers, MU-C-1A and MU-C-1B, where the primary water is cooled to approximately 1200F. The 20 letdown coolers are located in the reactor containment building at a centerline elevation of 286'3".
Normal letdown flow is 45 to 70 gallons per minute with 19 a maximum capacity of 140 gallons per minute.
0 The cooled primary water at approximately 2,135 psig and 120 F exits the reactor containment building through penetration R 541 to the block orifice 6,29 valve. The block orifice valve is a pressure-reducing valve used to reduce le:down water f rom 2,135 psig to approximately 20 psig.
With both reduced temperature and reduced pressure, the dissolved gas content of the primary cool-ant is normally sufficiently low to preclude significant degassing or two phase flow in the letdown /make up system.
If the letdown water has significant.
dissolved gases, the pressure-reducing characteristics of the block orifice may not result in sufficiently low pressures in the letdown /make up system in the auxiliary building. To prevent overpressuring filters, demineralizers, tanks and other components in the low pressure portion of the letdown /make up system, a pressure relief valve, MU-R-3 (setpoint - 130 psig) was installed just down-5 stream of the block orifice valve and ahead of the lower pressure components.
1182 079
9 ti ll i-i:
y,1 i
2 i
- iliin
-l t
t I
s
- lll!!! y j
3 l3, 5
! 's !$'l u-Ic 8
ilini It t' O
e
't h
[. di L
3
, ;.n fr a e-h.,
si u.,
- e u e}i, e i p
- i..-) M.,. Q.
n!_'9!.11'=' !/,
ilt 1i Il L
'0'l'i 6
i 38tI 8
i e
.J
.hli re-4 id, 1
ji 3
ist a'. I t
1 I a, I
u
'I I
lilli !?
i 4i=Iiet
.1; ap"k ilu v
I I
i 11R2 080
This pressure relief valve also protects the downstream components in the event of blockage or inadvertent closure of inline valves. With lower-than-design flows, the block orifice valve is less effective in reducing line pressure as the flow decreases.
It is important to note that if the letdown /make-up system is in operation with the valves open upstream of the block orifice valve and a block or obstruc-tion is effected downstream, this pressure relief valve MU-R3 relieves pressure 6
by venting to the reactor coolant bleed holdup tanks. The same relief is provided if the overpressure is caused by dissolved gases.
A remotely operated diaphragm bypass valve is installed parallel to the block orifice valve and is used to maintain adequate fluws at reduced reactor coolant pressures.
The letdown water is filtered and passed through the make-up and purification 6
demineralizers MU-K-1A and MU-K-18.
The demineralizer can be operated singly or in parallel. There are two pressure relief valves, MU-RSA and MU-R5B (setpoint -
5 150 psig), downstream of the demineralizers that discharge to the waste disposal drain. The waste disposal drain directs liquids to the auxiliary building sump.
6 A three-way valve, MU-V8, located Just downstream of the demineralizer, is 6
used to route the letdown stream to the make-up filters, to the reactor coolant bleed holdup tanks.or to the deborating demineralizers. The reactor coolant bleed holdup tanks (three of them: WDC-T-1A, WDL-T-1B, and WDL-T-IC) function as surge capacity for the letdown /make-up system. Chemical injections, return irom the reactor coolant bleed holdup tanks, deborated water from the deborating demineral-Izer, borated water, and demineralized service water inlets are located between the three-way valve MU-V8 and the make-up filters, MU-F-2A and MU-F-2B.
The chemical injections include lithium hydroxide and hydrazine.
The make up tank (4,500 gal) receives filtered water for return to the reactor primary coolant system and to reactor coolant pump seals.
I182 081 4-3
Operation of the letdown /make up system usually maintains 2,800 to 3,000 gallons of water in the make up tank. The tank also provides a reasonable time and volume for gas-liquid separation and is vented by a remotely operated valve, MU-Ul3, to the vent header located in the auxiliary building.
A relief valve, MU-R1 (setpoint - 80 psig), is incorporated in the outlet 5,6 line from the make up tank and vents to the reactor coolant bleed hold-up tanks.
The exit line from the make-up tank is piped to the high pressure injection system pumps MU-P-1A, MU-P-1B, and MU-P-lC.
Normally, make-up flow is effected by pump MU-P-1B through a pressurizer level control valve to the high pressure outlet line of reactor coolant pump RC-P-18.
These high-pressure injection pumps raise the make up fluid pressure from the nominal 15 - 20 psig pressure of the letdown /make-up system in the auxiliary building to the 2,155 psig of the reactor primary coolant system.
4.2 Reactor Building Sump to the Auxiliary Building Sump 2I The reactor containment building is equipped wi th a pump system to remove uncontained liquid from the reactor building Sump to the auxiliary building.
Normal operation of this system would be to pump fluids f rom the reacter buliding sump by use of the reactor building s ump pumps WDL-P-2A t.:d WDL-P-28 through Fil ters WDL-F-8A and WDL-F-8B to the niiscellaneous was te hoidup tank WDL-T-2.
The system can also be valved to pump the fluids through Filters WOL-F-3A and WDL-F-39, to the auxiliary building sump tank. Consensus of operating personnel and level readings of the miscellaneous waste holdup tank before and after the event indicate the valve line-up was to the auxiliary building sump tank. The schematic of this flow path is shown in Figure 2.
The normal operation of reactor containment building sumo pumps is an automatic start on high-level and stop on low-level in the sump. Uncontained water accumulates in the sump which is the lowest point (bottom elevation 276'6") in the reactor building. Flow of liquid it. through a 4-inch line via reactor containment building penetration, R-547, to a tee in the auxiliary build-ing where it is routed either to the miscellaneous tank in a 4-inch diameter line.
4-4 1182 082)
49 Il:Iil
'1 :, 1 l.
eli :.
1pli bd e-3 si il r i,t a
h h,
ici
.1 I.
h i.
eD fY k
s's eb I*
O su s
It
-l <eb 0
a
-i I
l
!,i A 5
.s i
y g la l
a-
- ]
g W
Is ~.^
I M
ll:
ai
~
2 i
lI e
li
{l 8t i
i i
e i e
e e i
g, i
a:
s/
- s i
s l
p-
- i e e s
i e
i I,
Ii i's l
3*
lie 87l 1
r, i
g, d
I q
a 48
.I ji bI si t
I
?'
.s
.w_m.
=J
!I f;
1 l
1l82 083
or to the auxiliary building sump tank via a 2-inch-diameter line. The pumps can be also remotely switched on or off manually from an auxiliary building control panel.
At the time of the accident, the rupture disc, WDL-U-224, on the line from 22 the sump tank pumps (WDL-P-4A and WDL-P-48) back to the tank nad previously burst. This presents an open 2-inch line from the auxiliary building sump tank to the sump.
The auxillary building sump tank is equipped with a pressure relief valve, WDL-R200 (setpoint - 20 psig) that discharger to the relief valve vent 23 header.
It should be noted that the re!Ief valve vent header discharges directly 24 to the station vent.
4.3 Reactor Coolant Drain Tank to Reactor Coolant Bleed Hold -Up Tanks The reactor coolant drain tank receives the fluids released from the
.v 25 pressurizer relief valves through a 14-inch-diameter line. The tank is located in a cubicle at the bottom level of the reactor containment building at a center-line level of 289 feet outside the secondary shield wall. A schematic of the 20,26 connections a-d p"mpa of the reactor coolant drain tank is shnwn in Figure 3.
The reactor coolant drain tank is a 7,400 gallor stainless steel tank.
The 8
tank has two systems that are discussed in Sectionr. 4.4 and 4.5 25 A relief valve, WDL-R1 (setpoint - 120 psig), protects the reactor coolaat 5
drain tank from overpressure and discharges to the reactor containcent building drain system that empties into the reactor containment building sump. As a eackup, an 18-inch-diameter rupture diaphragm is installed, and the ourlet of this vent line is outside the ethicle. An examination of the p. ant drawings shows the ou 'et from the rupture diaphragm line is 7 feet above the top cf the tank.
20 The reactor coolant drain tank has a leakage cooling system including two parallel leakage transfer pumps WDL-P-9A and WDL
'-9B, with two parallel leakage coole rs, WDL-C-1 A and WDL-C-18.
This circuit is used to cool one drain tank fluids A 4-inch-diameter line tees off the reactor coolant drain tank
'~'
lib 2 084
h..
11 M:.. i i
2 3
h
,s e
i ALh i ililli I
i 4:_
ig s
2 hI t?
ilibi TP '!
u
.l 9
ililii i
!I'~
x,!
a e>
11
-1 I
t
=.
I5 Gn1 rb,in
-e
.,8 i
e* l, av l,
i u
er!
fr*p s
Cl g
ll l
1 2-6 i
p+-g g! liJ@.4 t
l I
Qil i
!v I!
E
!'i Il I
1!
1 l-l! -11 8l !;
ii
. 3.,
I 1182 085 1
~ ~.
cooling system and goes to the reactor coolant drain header. The reactor coolant 27 drain header connects to the reactor coolant bleed holdup tanks. To r emove fluids from the reactor coolant drain tank, normally a valve, WOL-Vill 8, is opened 28 remotely from the control room to permit excess water to flow to the bleed holdup tanks. A level-control signal automatically closes Valve WOL-Vill 8 at a pre-set level to prevent removing too much fluid from the drain tank and thereby running the leakage transfer pump dry. Removal of excess water due to the leaking 4
p;essurizer safety valves had been performed periodically prior to the accident.
Each of the reactor coolant bleed holdup tan'.s has two relief valves that discharge to the relief valve vent header. Each relief valve has a setpoint of 5
20 psig. The relief valve vent header discharges to the station vent stack directly.
A flow orifice is also located in the fluid line from the leakage cooling system to the reactor coolant drain header. This orifice has a flow recorder, 25 FR-7100, located in the centrol room of THi-2, 4.4 Reactor Coolant Drain Tank Vent to Vent Gas Header in the Auxiliary Building The reactor coolant drain tank has a two vent system. One vent goes to the 25 vent header in thd auxiliary building. The other vent goes to the reactor coolant bleed holdup tanks.
The vent to the vent header is a 1-inch-diameter line. In the event of a substan-tial flow of fluids into the reactor coolant draln tank, it is unlikely that this line would accommodate venting of gases from the tank to avoid pressurizing the tank.
Since the line f rom pressurizer relief valves is a 14-inch-diameter line, ic is rather obvious the vent system is designed to accommodate intermittent full pressurizer relief valve openings or only small relief valve leaks.
A schematic of the vent system is shown in Figure 4.
4.5 Reactor Coolant Drain Tank Vent to the Reactor Coolant Bleed Hold'-vo Tanks
'S The second vent system f rom the reactor coolant drain tank connects the tank to the reactor coolant bleed hold-up tanks. The schematic of this system is shown in Figure 5 4-3 1182 086
4 0
io li' la ibik
'J
$e ki Od Y
I 5
II
?
I s
If I
l I
i i
I 5
I.
v.
tl l
i
-r i
=
ll
'dqli--
,i i
r'i j! s a ll Il N
[?
!il i
i I
a-i h
I e
0 1182 087
0 li;* 1 I
11!!!
s il is i
ilii d s
sf I
J m
9 7
(
!k 5
W--
fi-
$1 5
ililli l
I 3
w I
.i !bi 2
,g-
.i
- b iH I m__
1 1
I I
i I
I I
i
- I i
j!:
p la e
i I
s L
(j 'i !
I i
i.
i' h
l
.?
il82 088
This vent has an orifice, WDL-U23, 0.285 : inch -diameter in line. Between the orifice and the bleed hold-up tanks,the vent line size is 2 inches in diameter.
Apparently the second vent line is incorr *ated to assist in relieving pressure at higher inlet fluid flows into the drain tank.
Even the two vent l i nes a re unlikely to accommodate full opening of a pressurizer relief valve.
A check valve, WDL-V1098, in the line permits increased backflow from the reactor coolant bleed holdup tanks to the reactor coolant drain tank.
4.6 Reactor Coolant Pumo Seals to Seal Return Coolers in the Auxiliary Building 6
The reactor coolant pumps have mechanical seals that are maintained at operating temperature by the primary coolant. The internai pressure of the seals is matched by the pump seal fluid on the outside of the mechanical seal.
This seal fluid is obtained from the letdown /make-up system. The return fluid is cooled to maintain low tenperatures on the shaft and external seal of the pump that, in turn, extends the running time of the pump seals. Also, i there
.ls leakage through the mechanical seal, the primary coolant leakage is caught by the seal fluid which minimizes contamination to the reactor containment build-ing.
Reactor coolant pump seal fluids are manifolded and exit from the reactor containment buliding to the seal return coolers MU-C-2A and MU-C-2B The system is protected by a relief valve, MU-R2 (setpoint - 150 psig) that c.scharges to the inlet line to the make-up tank.
The seal fluid return comes from the mixture in the make-up tank. Total seal flow to each reactor coolant pump is 8 to 10 gallons per minute.
A schematic of reactor coolant pump seal flow systems is shown in Figure 6.
47 Letdown Coolers Cooling Water The letdown flow exits from the primary coolant system and is cooled in the reactor containment building by a pair of letdown coolers. The letdown coolers 1182 089
- ~ "
's b.:
l.!
i 1! \\.
a lijd.0 s
a
=
3 8
I a
s 2
s I
iIi I.:
4.
9 e,
,1 i
W 1 If 7 I i
1 ;l.f.*
i,ii p
I l
I!
l
!t t
- =
4'd i
+
b I!
!ks I
{c I ci c;
al t
Idf D
!N!
.i li;*
l I
=;d; 111 i
i I
- =
11
- si i
- ll;
- 11; i
i I
i i
i i
I 1
i 1182 090 t
are located in the reactor containment building at the 286 fc.'t 3 inches level out-26 side of the secondary shielding. The heat-transfer surface in the helical-shaped 19 heat exchanger is made of 30 parallel tubes, 57 feet long. 3/4-inch-diameter 16 BWG stainless steel seamless tubing. The design pressure for the tube side is 2,500 psig, and design temperature is 6000F. The heat is remowed by cooling water from the intermediate, closed cooling water system. The major portion of the closed recirculation system of the letdown coolers is located in the auxiliary building. The intermediate closed cooling water system is a recirculating system that would retain any radioactivity leaked to it and would also increase in inventory if a leak occurred from the letdown /make-up system.
For radioactivity to be removed from the reactor containment building by the letdown coolers cooling water would require a leak from the prirrary side of the letdown cooler tubing or of the tube sheet.
4.8 Leakage Coolers Cooling Water The leakage coolers are used to cool the fluids in the reactor coolant drain tank.
These coolers are located in the reactor containment building at the 29 285 feet 6 inches and 290 feet 8 1/2 inches levels outside the secondary shield-to ing. The cooling water for leakage coolers is part of the decay heat closed cooling water system. The major portion of the decay heat closed cooling water is in the auxiliary building. Both the tubc sides and the shell sides of the leakage coolers operate at approximately 5 to 10 psig. Therefore, there is no strong driving force to transmit radioactivity to the auxiliary building via the leakage coolers cooling water.
I182 091 4-13
5.0 INTERCONNECTION OF LIQUID AND GAS SYSTEMS The liquid and gas systems are intercoupled, ar.d communications can be estab-lished from one tank to another tank through an inte rmediate tarin.
Further, it is conceivable that the relief of pressure in one tank or system may, in turn, pressurize the recipient tank so that its relief valve setting may be exceeded.
24 In most systems there are check valves that function to allow flow in one direction only. Howcver, most check valves are normally not suf ficiently reliable to prevent some backflow,particularly if the seacs are metal. Relief velve vent systems do not normally have check valves since these systems are designed to be ultimate free paths of pressure relief.
During the accident, the major releases were g as eous. To assist in under-standing the interconnections of the tanks, compressors, and other components, a summary drawing of the vent relief / waste gas ties is shown in Figure 7 Also, the overall connections of the most important sys 6 ems considered in this report are shown in Figure 8.
In evaluating the systems at TMI-2, the reactor coolant bleed holdup tanks appear to be the center of the potential release pathways.
The reactor coolant drain tanks' have liquid connections to the reactor coolant 27 bleed holdup tanks as well as gas vent ng connections. The letdown /meke up i
system is extensively tied into reacto.
slant bleed hold-up tanks. The pressure 6
relief valve MU-Fd just downstream from the block orifice valve discharges to the reactor coolant bleed hold-up tanks. The pressure relief valve, MU-R1, on the outlet stream of the make-up tank discharges to the reactor coolant bleed hold-up tanks. The flow f rom the letdown /make-up system can oe diverted via a three-way valve, MU-U8, to the bleed hold-up tanks. The wast transfer pump' WDL-P-5B and WDL-P-5A draw from the bleed hold-up tanks and discharge into the line feeding into the letdown /make-up system just upstream of the make-up fi l te rs MU-F-2A and MU-F-28.
1182 092
$-1
l;: I 1 l!
- l,l i
9 g
b 8I St I
.if I
.!s,.
i 4f t...* % a,
l 8
,8
'4 C )'f., di,F q
Mil,!!
t g2,
> P i 1i-i i;
4 i
- iliisi is,. :1 I
,gt_
,)
Tj
!.I,TP,A.,,
4 n-
.li,,,
c.*
4.
n
- i
.i mJ 3 a
,.,,.i:
.E i
i T
- i;iii; r
mi r i
' 11 y '
111
-#-l.
irt
- i' r-n l
C,lll11 li
'I
.,P l
it i
i
=e eld'
_.. g' y =
w O
...gr t 3
i
)J l[
l 1
ij i'
>l a
i
.m>
I;
' 15.1 f
I 1
I N
E i
e i
~;
$r!!
i l
i~
1 i
11821 093
I vl i
'I..
l1;l, ies : s lj s ;.,i j '.
' :. 3:
L a-N l ki, !
D 18 i
, at e ' *i si 7
f' l
e t
i I
,,f, I
4 i,
i a
i I
j i.
A T:a
, j.,
8 rt i
5 me' i d:,t,
.Ji.) i r J; t
it a.,.
- i I
c._
4 i l}illi s y. !4O - i, ', -
i l
i, %,,X;.,, ".
4, a,,
. i, j
, pr.
,j a
j,'ili
'I I.
- 3 r _I'1 f l*-
['
l' l
IL- +
i 'I i bir,
!.' M I j:. li s
O.
3 y
I
,ilill
_rn. j l
a
!_ i i.) ;
i l
i ",.< 5 ' ;
I il 1
i C'eS-i N, i !
i d
T'NA._;s j li ri i n nl I$
si -q >!
i
'ijllll{
~
4 -b,'% @ i
!u L
I ' I
's_.__si h}
il i
.I l
~
l i!i YITev.-- !.l !
d
'S
!!'ii i i
- 84 ge t:_
--f
-- C
._3~_8 F,
i dl,j 'l ':t
~
t
...g I l.~__J Di l
8s=.> ;i
.I
!..p t.,
i 8
i I ii ii i sin, u..g i a i
i 1
l
.j
_c---
i gi !!,!
lid
- il9,
.l I;^_
ml1:
lW 1
i a.i ii,
i) 2.,
L's.8' m *i u, p
~
--T, l I
W
!L 15il'
- i. l'}[ { i I,U 6
~- i i
A l
j eM' l8, Ti'.
2 Mi l'U !, j
,t! j,,
i lb,,..!l }' l M;j[I. a"! 7< 3 i
}
l
,h l
t_
J..'...,.l_!l! m.. b iii
'i i
k,ui,,!!..7 :P i
m n..5d,:
11!
,d,a y2 t i,i s
..v.
,c.
1 l.
Ip 1'; -, y
.' FW[
l i
i r!:h:-,!,*t
, i i
g
,jl,J,.;_ ol
,1
- 1 !
l i
r-1 i
i i ~, jj l
. E..
'il p! q s! ! n :j p}
.t i si 4
1 g [ =-,4 ii l
>1 i i li Dll p 4 52 y gl
- f 1
i.
?lji l.t.,
[" il
.I i : :
ie A'" *' &Q 'l e.'s_ y, j. tg
!".' l m
i, i l s u
yl.
l3 i,,'r
'l I
i ii 1:l Ir [
i i~~
T-
-"^~
4 11-ii r
'I i.gi i'
a a
l, M i- " "- T h ?!
!; ;!n > ;
i i i
- i. p T
- i. :
w g II.,..* 'i 1:-
i i
b,l:ly i i. i.,
. " p..
- , g_., f; : ld ;
I'
t' I
i.
,e,.
I ii l
i a C' i d. i:
l
,[-
- illi, 090.......!
J.
j f
- 1..
g s
l awk z.
-=
I~i :
--' rn?!
3.1nii
,12 i
a-e,1 s -
I I
li
!I f
i 11 1
1182 094
The reactor coolant bleed hold-up tanks are piped to the vent haader in the 27 auxiliary building, permitting ready venting of fluid from the tanks. A line cor.nects the auxillary building sump tank to the vent gas header.
Another extensive set of interconnections is the relief valve vent system. The 24 relief valve vent system discharges directly to THI-2 st'ation vent. This is important because if a relief valve has lifted, it may not always reseat to be le ak-f ree.
This could very well have happened during or, in the case of the auxiliary building sump tank, before the accident.
It is of interest to note that the relief valves from the waste gas compressors are piped to the auxiliary 24 building sump tank that had a blown rupture disc at the time of the accident.
The vent system is equipped with traps at the low points to drain water that might get into the vent system from lifting of pressure relief valvas. There are are at least two of these traps, WDG-U8A and WDG-U9A, in the auxiliary buildings.
The interconnections of the liquid system with the gas systems make it difficult to diagnose the release points of the radioactive gas from the auxiliary building.
1182 095 5-4
6.0 RADIATION MONITORING IN THE PLANT There a re 49 radiation monitors in the THI-2 plant. Of these, 21 5
are area monitors utilizing GM tubes for gamaa level detection; 12 a re a i r-borne radioactivity monitors that mea are particulate, iodine, and gaseous 4 are beta scintil ation monitors for gaseous r'adioactivity; and 12 content:
are gamma scintillation counters for detecting suspended or dissolved radio-active isotopes in liquid streams.
For purposes of this report, the only monitors that are useful are those for which data were retained on strip chart re co rt'e r s. Listings of these recorded monitors are shmn in Tables I, Il, and Ill.
The response of each monitor is recorded by a printed number stamped on the recorder chart. Also included in Tables I, 1I, and iII are the monitor locations as welI as the drawing fren which the locations were determined.
For more details, refer to the Met Ed/GPU " Preliminary Report on Sources and Pathways of TMI-2 Releases of Radior.ctive Material," Appendix G.
1182 096 6-1
TABLE I LIST OF RADIATION HONITORS AND LOCATIONS PROCESS M0tilTORS Burns & Roe Recorder Recorder Channel Recorder Drawing Location Plant Monitoring Point Stripchart Number Number Designator Number on Drawing Elevation Primary Coolant Letdown HL HP-UR-3264 7
1 HU-R-720 2066 AE/A63.5 305' Primary Coolant Letdown LO HP-UR-3264 7
2 MU-R-720 2066 AE/A63 5 305' Intermediate Coolant Letdown HP-UR-3264 7
3 IC-R-1091 2060 Ril 282'6" Cooler B Intermediate Coolant Letdown HP-UR-3264 7
4 IC-R-1092 2060 Rll 282'6" Cooler A Intermediate Coolant Letdown llP-UR-3264 7
5 IC-R-1093 2066 AB/A62.8 305' as 4
Cooler Outlet Plant Effluent Unit 11 HP-UR-3264 7
6 WDL-R-1311 Decay Heat Closed A Loop HP-UR-3264 7
7 DC-R-3399 2065 AK/A67 280'6" Decay Heat Closed B Loop HP-UR-3264 7
8 DC-R-3400 2065 AK/A67 280'6" Nucl. Serv. Closed Cooling HP-UR-3264 7
9 DC-R-3401 2066 AE/A61 305' Spent Fuel Cooling HP-UR-3264 10 SF-R-3402 2066 AN/A64.7 305' A
Yrs)
C)
'L)
M
TABLE 11 LIST OF RADIATION MONITORS AND LOCATIONS AREA MONITORS Burns & Roe Recorder Recorder Chai:nel Recorder Drawing Location Plant Area Monitored Stripchart Number Numbee Des i g.v a to r Number on Drawing Elevation Control Room HP-UR-1901 1
I HP-R-201 2381 CB/C48 341' Cable Room HP-UR-1901 1
2 HP-R-202 2380 CC/C48 305' Emerg. Cooling Booster Pump HP-UR-1901 1
3 HP-R-204 2065 AB/A61 280' R.C. Evap Control Panel Area HP-UR-1901 1
4 HP-R-205 2065 AG/A63 280' Make-up Tank Area HP-UR-1901 1
5 HP-R-206 2066 AG/A64 305' Intermediate Cooling Pump Area HP-UR-1901 1
6 HP-R-207 2066 AA/A63 305' Fuel Handling Bridge North HP-UR-1901 1
7 HP-R-208 2064 355' j[ Fuel Handling Bridge South HP-UR-1901 1
8 HP-R-209 2064 355' R.B. Personnel Access Hatch HP-UR-1901 1
9 HP-R-210 2064 310' R.B. Equipment Hatch HP-UR-1901 1
10 HP-R-211 2064 314' incore Instrument Panel Area HP-UR-1901 1
11 HP-R-212 2062 371' Reactor Building Dome HP-UR-1901 1
12 HP-R-213 2064 372' Fuel Handling Bridge HP-UR-1902 2
1 HP-R-215 2068 AM/A66.5 347-Waste Disposal Storage Area HP-UR-1902 2
2 HP-R-218 2066 AR/A67 305' Aux Bldg Sump Tank Filter Room HP-UR-1902 2
3 HP-R-231 2065 AQ/A62 280' Aux Bldg Access Corridor HP-UR-1902 2
4 HP-R-232 2066 AR/A61 305' Aux Bldg Access Corridor HP-UR-1902 2
5 HP-R-233 2066 AN/A63 305' Control & Service Bldg HP-UR-1902 2
6 HP-R-234 2380 CE/C50A 280' RB Purge Unit Area HP-UR-1902 2
7 HP-R-3236 2067 AE/A63 328'
(([
Aux Bldg Exh. Unit Area HP-UR-1902 2
8 HP-R-3238 2067 AJ/A63 328' 03 Fuel Handling Exh. Unit Area HP-UR-1902 2
9 HP-R-3240 2067 AL/A63 328' rs)
CJ W
CO
TABLE Ill LIST OF RADIATION MONITORS AND LOCATIONS AIRBORNE MONITORS Burns & Roe Recorder Recorder Channel Recorder Drawing Location Plant Monitored Honitoring Point Stripchart Number Number Designator Number on Drawing Elevation Variable Station Vent HP-UR-1907 4
i HP-R-219 2067 AF/A65 328' Particulate Station Vent HP-UR-1907 4
2 HP-R-219 2067 AF/A65 328 lodine Station Vent HP-UR-1907 4
3 HP-R-219 2067
'AF/A65 328 Gas Control Room Intake HP-UR-1907 4
4 HP-R-220 2381 CD/C48 351'6" Particulate Control Room Intake HP-UR-1907 4
5 HP-R-220 2381 CD/C48 351'6" lodine Control Room Intake HP-UR-1907 4
6 HP-R-220 2381 CD/C48 351'6" Gas Fuel Handling Bldg Exh.
HP-UR-1907 4
7 HP-R-221A 2067 AT/A63 328' Particulate Upstream of Filter S-Fuel Handling Bldg Exh.
HP-UR-1907 4
8 HP-R-221A 2067 AT/A63 328' lodine Upstream of Filter Fuel Handling Bldg Exh.
HP-UR-1907 4
9 HP-R-221A 2067 AT/A63 328' Gas Upstream of Filter Fuel Handling Bldg Exh.
HP-UR-1907 4
10 HP-R-221B 2067 AT/A63 328' Particulate Downstream of Filter Fuel Handling Bldg Exh.
HP-UR-1907 4
11 HP-R-221B 2067 AT/A63 328' lodine Downstream of Filter Fuel Handling Bldg Exh.
HP-UR-1907 4
12 HP-R-221B 2067 AT/A63 328' Gas Downstream of Filter Hydrogen Purge HP-UR-1907 4
13 HP-R-229 2067 AF/A65 328' Particulate Hydrogen Purge HP-UR-1907 4
14 HP-R-229 2067 AF/A65 328' lodine Hydrogen Purge HP-UR-1907 h
15 HP-R-229 2067 AF/A65 328' Gas w
rs)
CD
LIST OF RADIATION MONITORS AND LOCATIONS AIRBORNE MONITORS (continued)
Burns & Roe Recorder Recorder Channel Recorder Drawing Location Plant Moni:ored Honitoring Point Stripchart Number Number Designator Number on Drawing Elevation Variable RB Purge Air Exh. Duct A HP-UR-2900 5
1 HP-R-225 2067 AB/A64 328' Particulate RB Purge Air Exh. Duct A HP-UR-2900 5
2 HP-R-225 2067 AB/A64 328' lodine RB Purge Air Exh. Duct A HP-UR-2900 5
3 H'*- R-2 2 5 2067 AB/A64 328' Gas RB Purge Air Exh. Duct B HP-UR-2900 5
4 HP-R-226 2067 AB/A64.5 328' Particulate RB Purge Air Exh. Duct B HP-UR-2900 5
5 HP-R-226 2067 AB/A64.5 328' lodine RB Purge Air Exh. Duct B HP-UR-2900 5
6 HP-R-226 2067 AB/A64.5 328' Gas Aux Bldg Purge Air Exh.
HP-UR-2900 5
7 HP-R-222 2067 AT/A63 328' Particulate Upstream of Filter
}[ Aux Bldg Purge Air Exh.
HP-UR-2900 5
8 HP-R-222 2067 AT/A63 328' lodine Upstream of Filter Aux Bldg Purge Air Exh.
HP-UR-2900 5
9 HP-R-222 2067 AT/A63 328' Gas Upstream of Filter Aux Bldg Purge Air Exh.
HP-UR-2900 5
10 HP-R-228 2067 AT/A63 328' Particulate Downstream of Filter, Aux Bldg Purge Air Exh.
HP-UR-2900 5
11 HP-R-228 2067 AT/A63 328' lodine Downstream of Filter Aux Bldg Purge Air Exh.
HP-UR-2900 5
12 HP-R-228 2067 AT/A63 328' Gas Downstream of Filter Reactor Bldg Air Sample HP-UR-3236 6
1 HP-R-227 2066 AB/A63 328' Particulate Reactor Bldg Air San.ple HP-UR-3236 6
2 HP-R-227 2066 AB/A63 328' lodine Reactor Bldg Air Sample HP-UR-3236 6
3 HP-R-227 2066 AB/A63 328' Gas Waste Gas Discharge Duct HP-UR-3236 6
4 WGD-R-1480 2067 AB/A62.5 328' Gas WDG-T-IA Waste Gas Decay HP-UR-3236 6
5 WGD-R-1485 2066 AG/A62.5 305' Gas Q
Taik Discharge I'3 WDG-T-1B Waste Gas Decay HP-UR-3236 6
6 WGD-R-1486 2066 AD/A62.5 305' Gas Tank Discharge C:3 Condenser Vacuum Pump HP-UR-3236 6
7 VA-R-748 2052 TG/T42 281'6" Gas CZ)
Discharge
7.0 POTENTIAL PRESSURE TRANSIENTS IN EACH SYSTEM 7.1 Letdown /Make-up System The Letdown /Make-up System is explained in Section 4.1,.
This system normally is the main. flow stream of. reactor primary coolant from the reactor containment b ui l di ng.
Pressure in the reactor coolant system is normally imposed and sustained by 17 the pressurizer through the use of heaters. There is a slight partial pressure of hydrogen maintained to assure recombination of hydrogen and oxygen produced by radiolysis.
Since there is a minimum of dissolved gases in the primary coolant, the letdown /nake-up system is designed to handle a modest amount of gases by venting into the waste gas vent header from the make-up tank.
The block ori fice valve MU-1-FE in the letdown flow line is located in the 6
auxiliary building. The valve is designed to reduce the pressure in the system 29 from a system piessure of 2,135 psig at the outlet of the letdown cooler to about The relief valve MU-R-3 protects the downstream components from over-20 psig.
p re s s u re.
During the accident, three circumstances could, and probably did, cause the 30 31 relle.f valve MU-R-3 to lif t:(a) When the bypass valve MU-V5 around the block ori fice was oper.ed to increase letdown flow, there is a potential of a mismatch 32 in flow-related pressure drop through the make-up and puri fication demineralizer filters, the make-up and purification demineralizers and the associated piping before the make-up tank; (b) I f there is a flow blockage downstream of the block orifice valve when the letdown system is functioning, the pressure up to the point of blockage would rise until the relief valve li fted; and (c) When there are sufficient dissolved gases in the primary system, the reduction of pressure through the block orifice valve can cause substantial degassing with subsequent
'~'
1182 101
two phase flow that changes the pressure drop through the filters and deminerall-zers. There is a pressure indicator, HU-PI-1579 Just downstream of the block ori fice valve. However, since there Is no recorder, no trace is available for analysis.
A review of the operations during the accident indicates that all three 30,31,32 ci rcumstances were probably experienced. The relief valve is set at 130 psig 33,34,35 and discharges into the. reactor coolant bleed hold-up tanks. A mixture of 36,37,38 liquid and gases would be the discharge phases from the relief valve af ter 6:30 a.m.
on March 28. This is the approximate time when substantial amounts of hydrogen were generated in the primary system by the zircaloy-water reaction.
72 Reactor Building Sump to the Auxiliary Building Sump Fica of primary water from the reactor building sump to the auxiliary build-1,2,5 Ing sump occurred for approximately 30 minutes during the initial 38 mintues of the accident when the reactor building sump pumps were operating. The Met Ed/GPU analysis concluded that siphon flow probably did not occur with the pumps shut' down. This writer reviewed the plant drawings and agrees with the analysis conclusion.
Calculation by Met Ed/GPU indicates a 5
significant amount of water was pumped by the reactor building sump pump (8,400 gallons). An assumption was made that the water flowed into the auxiliary build-Ing sump tank and then into the sump via the burst rupture disc WDL-U224.
In examining the overall flow system, some of the water could very well have gone up the vent line into the vent header, depending on the respective flow resistance and the vent header pressure during the pumping period. The injection of water into the vent gas header would have deleterious effects on the operation of the vent gas header system.
There is some confusion as to the status of the controls for the reactor building sump pumps. They were turned off at about 6:38 a.m., and an auxiliary building oper tor recalled they were not operating at about 7:00 a.m.; however, a shift supervisor reported he found the reactor building sump pump controls in 39 the automatic pos i tion at about l,200 - 1,300 and turned them of f.
7-2 1182 102
If the pumps did run for the period from 7.:00 a.m. until 7:56 a.m. when the reactor containment building was isolated by a pressure signal, another 7,800 gallons may have,been pumped into the auxiliary building sump. The inlet of the reactor building sump pumps is at the bottom of the sump and would probably have con-tinued to pump "first out" primary water that was low in radioactivi ty.
What-ever water was pumped out by the reactor building sump pumps would have been degassed to a large extent by steam stripping and depress'urization as the water exiting from the electromatic relief valve through the reactor coolant drain tank.
In reviewing the overall behavior of circumstances of pumping water from the reactor building sump to the auxiliary building sump via the auxiliary building sump tank, it is not obvious that this caused any pressure transients, but it did contribute to the inventory of initial primary system water in the auxiliary building.
7.3 Reactor Coolant Drain Tank to the Reactor Coolant Bleed Holdup Tanks A review of operations before the accident Indicated that the reactor cool.
28 ant drain tank received leaktge from the pressurizer relief valves (probably the safety relief valves based on temperatures). The chart from Flow Recorder 4
FR 7100 indicated that periodically leakage water from the reactor coolant drain tank was being transferred to the reactor coolant bleed hold-up tank through the reactor coolant drain header. To perform this transfer, one of the leakage trans fer pumps, WDL-P-9A or WDL-P-98, must be operating. Since there were periodic t ran s fe rs prior to the accident, it is believed that the selected pump was operating during at least part of the accident.
The transfer of liquid from the reactor coolant drain tank tc the reactor coolant bleed hold-up tanks requires the opening of valve WDL-Vill 8 by use of 28 remote controls. Level indicator and controller WDL-1206 will automatically close WDL-Vill 8 if the level in the tank goes below 72 inches.
Probably the best indication of flow to the reactor coolant bleed hold-up tanks is the Flow Recorder FR 7100. The flow range on the chart is 150 gallons per minute maximum.
1182 103
,.3
I There should have been no flow af ter the reactor containment building was isolated at 7:56 a.m.
This statement is inconsistent with the indication on the chart from the recorder. The anomalies are that there were extensive flow 4
Indications from about 9:08 a.m. until 2:30 p.m.,i.e.,the start of what one might expect from the udtput of this recora was 9:08 a.m. instead of 4:00 a.m., and the duration was 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> 22 minutes instead of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> 56 minutes.
If one assumes the chart speed is not correct and the stamped date is in error by approximately 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, the traces on the chart are quite plausible. There were four periods of flow for about 50 percent of the lapse time which corresponds quite well with the open electromatic relief block valve plus a period with the block valve being closed and then reopened. The chart indicated intermittent flows greater than 150 gallons per minute. By normalizing the flow periods to the total lapse time of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> 56 minutes, a total transfer of 9.450 gallons could have occurred.
This would not have exceeded the capacity of the reactor coolant bleed hold-up tanks. The liquids also would have been essentially degassed during the flow from the pressurizer to the reactor coolant drain tank.
7.4 Reactor Coolant Drain Tank Vent to Vent Gas Header in the Auxiliary Building The reactor coolant drain tank vent is connected to the reactor building vent header by approximately 70 feet of 1-inch pipe. The 4-inch reactor building vent 40,41 header is coupled to the auxiliary building vent gas system. During normal opera-tion, venting would be modest due to relatively slow changing levels in the reactor coolant drain tank.
When the electromatic relief valve opened at the time of the accident, the 42 reactor coolant drain tank was pressurized first up to 120 psig when the pressure relief valve opened and then on up to 190 psig when the rupture diaphragm burst at about 15 minutes into the accident. The tank was subjected to pressures between 120 psig and 190 psig for some 12 minutes.
During this 12-minute period, a mixture of mostly steam and water was forced into the reactor vent header and then into the auxiliary building vent system.
Some of the water was removed by the trap WDG-U10A in the reactor containment 24 1182 104
,.u
building. The steam was condensed in the piping. A reasonable assumption would be that most of the fluid that went out the vent was water.
If one assumed a pressure differential of 100 psi across the one-inch pipe in 12 minutes, over 1,000 gallons of water may have been ejected from the reactor coolant drain tank into the auxiliary building vent header. This appears to be an excessive amount of water.
It also could be assumed that the released fluids through the vent line were a mixture of steam and water that would reduce the condensed volume of transported liquids. To give a perspective,1,000 feet of 4-inch-diamete; 43 Schedule 40 pipe has an internal volume of 660 gallons.
From the above, it can be seen that the vent headers in both the reactor building and the auxiliary could have received a substantial amount of water if the valves in the vent system from the reactor coolant drain tank were lined up normally.
The water would deteriorate the capabilities of the auxiliary building 44 vent header.
However, none of this water would have had extensive radioactivity since this event occurred prior to the uncovering of the core.
If the auxiliary building vent header had any leaks at the inception of the accident, this pathway would take even minor amounts of radioactive gas and liquids directly over to the auxiliary building. This might explain some of the early airborne radiation monitor responses (see Section 8.0, Table 11).
The outlet of the 18-inch-diameter vent pipe housing the rupture disc is 7 feet above the reactor coolant drain tank.
If one assumes a liquid head of water equivalent to about 3 psi, the internal pressure for liquid or gases af*
the failure of the ruptura diaphragm would be 4.5 to 5 psig.
Depending,on the pressure in the auxiliary building vent header, flow may or may not have occurred af ter the rupture diaphragm failure.
It was reported that the vent valves WDL-U126 and WDL-U127 were found ep -
after the accident and subsequently closed on June 5, 1979 b2
}bJ 7-5
7.5 Reactor Coolant Drain Tank Vent to Reactor Coolant Bleed Holdup Tanks The reactor coolant drain tank communicates with the reactor coolant bleed holdup tanks through a vent line also. This vent line is automatically closed by valve WDL-1095 when the pressure sensor WOL-1203 signal exceeds 10 psig.
During 46 the accident, the valve should have closed within 2 minutes and remained closed until the rupture diaphragm burst at 15 minutes.
Af ter the rupture diaphragm burst, the driving pressure would have been about 4.5 to 5 psi until isolation occurred at 7:56 a.m. After isolation was effected, the driving pressure was purely academic since no flow could result.
It is questionable that during the initial part of the accident -- the first 2 minutes that the vent line could have been filled with liquid to set up a potential siphon. A siphon is also unlikely since the reactor coolant bleed holdup tanks are connected to the auxiliary building vent header and therefore do not have free liquid surfaces exposed to the atmosphere. Further, it is likely that pressure built up as liquid was routed to the reactor coolant bleed holdup during the course of the accident.
7.6 Reactor Coolant Pumo Seals to the Seal Return Coolers in the Auxiliary Building The reactor coolant pump seal water to the seal return coolers in the auxiliary building are fundamentally part of the reactor coolant letdown /make up systen.
The pumo seal water is fu nished by the output of make up pumps, normally Pump MU-P-18.
The seal water may become contaminated by the primary coolant if the primary coolant pump recnan' cal seals are leaking.
The effluent pump seal water is cooled and returned to the reactor coolant letdown /make up system just upstream of the make up tank.
The reactor primary coolant pumps were subjected to rather severe conditions 3
during the accident. These severe conditions which included vibration and pumping of two-phase fluids could have damaged the mechanical seals and allowed primary coolant to mix with the seal fluid. The seal water return could therefore have transmitted radioactivity to the make up tank.
8 06 7-6
The behavior of the reactor coolant pump seal water being returned to the letdown /make-up system appeared normal,without incident, throughout the accident.
7.7 Letdown Coolers Cooling Water A review of the records of the intermediate cooling water system did not 3
reveal any abnormalities in the operation utilizing the leticwn coolers cooling water.
7.8 Leakage Coolers Cooling Water A review of the records of the operation of the leakage coolers did not 3
reveal any abnormalities in the function of the cooling water.
7.9 Interconnection Pressure Transients The relief valve MU-R-3 from the letdown /make-up system downstream of the block orifice valve MU-1-FE discharges into the reactor coolant bleed hold-up tanks.
It is believed that the discharges from the relief valve contained gas and liquid and subsequently increased the pressure in the bleed hold-up tanks.
The release of dissolved gases can sometimes take appreciable time.
For this reason, the gas phase and the liquid phase discharging from the block orifice valve may not be in equilibrium. The equilibrium may not be obta'ned until the phases reach a tank where there is sufficient residence time. The make up tank serves this purpose. A review of the records shows that the make up tank became 35 pressurized and lifted the relief valve MU-R1 downstream of the tank. This relief valve discharges into the reactor coolant bleed hold-up tanks, also increasing the pressure in these tarks.
The combined pressure increases could and did lift the relief valves of 35,36,37 the reactor coolant bleed hold-up tanks. The relief valves f rom the reactor coolant bleed hold-up tanksdischarge to the relief valve vent t cader that in 24 turn discharges to the station vent.
1182 107 7-7
6.0 RESPONSE OF RADI ATION N0HITORS DURING THE ACCIDENT The radiation monitoring system at the THI-2 plant is designed primarily for operating conditions with most setpoints in the low range, i.e.,
at 20 milli-5 rems per hour or lower. Areas where it is reasonably certain that high radiation levels may be encountered during fuel handling operations, the range of the equip-ment is higher. The locations of the monitors with recorders are presented in Section 6.0.
The strip charts from the radiation monitor recorders give the most significant information with respect to where and when radioactivity was detected in the pla. There a re, however, certain limitations. Each monitor response is identified by a number stamped on the recorder chart.
In many cases, the stamped numbers were not inking properly at the time of the accident.
This makes precise reading of the monitoring data very difficult, and, in some ins tances, it is not possible to differentiate between different data points. The diagnostic readings are further complicated by imprecise timing marks on the strip charts. Howeve r,
within a strip chart, relative times are qui te identifiable.
Comparison between charts inserts probable errors and care must be exe.cised in interpretation.
In s umma ry, it is the best time information available and should be used fully realizing the limitation.
A compilation of the early radiation monitoring responses is shown in Table IV.
7,9,10,11 As in past experience in nearly all reactors, when a scram occurs there is a slight increase in radioactivity distributed within the plant primary system, due to two reasons:
(a) There is normally a mismatch between heat production in the core and coolant flow, and (b) There is a cay.d cooling of the fuel pins that creates some stress in the cladding.
The mismatch in overheat production and coolant flow can dislodge small radio-active particles that have accumulated, thereby creating a " crud" shower. The crud shower is normally confined within the reactor coolant system so that no gaseous evolution occurs.
8" 1182 108
TABLE IV EARLY RADI ATION HONITORING RESPONSES, MARCH 28, 1979 Recorder i ns t rumer.t N um.be r Instrument Description Starting Time of increase 2900 (6)*
HP-R-226-G R.B. Pui Air Exhaust, Duct A 4:07 4:22 6:43 3-2900 (3)
HP-R-225-G R.B. Purge Ai r E xhaus t, Duct B 4:07 4:22 1907 (14)
HP-R-229-1 Hydrogen Purge 4:15**See note below 3236 '(3)
H P-R-2 2 7-G R.B. Ai r Ample 4:16 6:22 3236 (1)
HP-R-227-P R.B. Ai r Sample 4:16 6:25 3236 (2)
HP-R-227-1 R.B. Air Sample 4: 16 6:31 A
1907 (3)
HP-R-219-G Station Vent 4:19**
1907 (7) itP-R-22lA-P F. fl. B. Exhaust Ups ream of Filters 4:22**
1907 (10)
HP-R-221B-P F. ll. B. E xha us t Downstream of Filters 4:22**
3264 (4)
IC-R-1092 Intermediate Cooling of Letdown Coolar A 4:22 5:00 5:23 6:37 2900 (9)
HP-R-222-G A.B. Purge Ai r Exhaust Ups tream of Filters 4:28
- Instrument code n umbe r.
- These recorders
- data a re di f ficul t to separate due to illegible traces. They appear to' essentially respond together and rose slightly at the 4: 15 to 4:20 time period and then rose again at 6:40 to 6:50.
M N
The rapid cooling of the fuel pins can create stress, and if there is marginal cladding in any of the 39,000 pins in the reactor core, a defect could cccur, and radioactivity might be released to the coolant. These fission products would be noble gases and volatile fission products.
It is believed that the initial responses to the radiation monitors reflect the rupturing of some fuel pins in the reactor core at or sho'rtly af ter reactor power was shut down. The fact that monitors in the auxiliary building indicated radioactivity shows that the gas systems were not leak-tight at the inception of the accident.
It should be realized that it is almost a Herculean task to ever get a reactor plant with the numerous pumps and valves to be a completely leak-tight system.
In the case of THi-2, the movement of primary water from the reactor contain-ment building sump to the auxiliary building sump via the failed rupture disc U-224 of the auxiliary sump tank during the ini tial 38 minutes probably prolonged the initial minor release. This probably was the source of the first response of the monitors in the auxiliary building. Obviously the responses in the reactor con-tainment building were f rom venting of the reactor coolant drain tank by way of the failed rupture disc.
The responses of IC-R-1092 in which there were increased boilt at 5:00 a.m. and 5:23 a.m. were probably caused by more direct exposure to the liquid in the reactor building sump which is just below the location of this monitor.
The major releases of radioactivity started Just af ter two hours into the accident when the core became uncovered.
it appeared that the first instrument to indicate additional and increasing radioactivi ty was the reactor containment building air sampler HP-R227 at approximately 6:22 a.m.
This would indicate there was a delay of 10 to 20 minutes from the time that it was thought the first uncovering of the core until fission gases got outside the reactor primary system.
Of the area moni tors, the reactor containment building monitors, HP-R-213, -214, 12
-209, and -210 first gave increased responses starting at about 6:31 a.m.
In the 1182 110 3-3
auxiliary building, HP-R-207 showed a modest increase at about 6:41 a.m. and a snarp increase at 7:19 a.m. This first response of HP-K-207 at 0:41 a.m. may nave been from the reactor containment building since this monitor is right against the con-ta'.iment building. These and other responses are shown in Table V.
Coupling the response of HP-R-207 at 6:41 a.m.with the make-up taak area monitor 12 HP-R-206 at 6:43 a.m. and intermeciate coolant letoown cooler outlet monitor IC-K-1093 at 6:43 a.m.,it appears tnat the initial radioactivity reached the auxiliar_y building about 10 minutes af ter at least some dispersion in the reactor containment building. This time delay is quite logical, considering that in the case of let -
down flow, the radioactivi ty had to get f rom the core.ove r to the l et down l i ne outlet in the A loop of the reactor coolant system.
Examination of the strip chart f rom HP-UR-1902 during the period early March 30 at 1:35 a.m. and 3:33 a.m. snowea a' difference in the response of radiation monitoi; HP-R-208, -2 32. -32 36, and -3240. When the make up tank was vented to the. vent 13 header, there appeared to be a delay of 4 to 10 minutes before the moni tors showed an increase in radioactivi ty.
Monitors HP-R-3240 and -3236 responded concurrently but to a di f ferent extent. Both of these monitorc are on the 328 foot level of the auxiliary building. Monitor HP-R-232 in the access corridor near the radwaste panel area on the 305 foot level responded about 2 to 5 minutes after the above monitors indicated increased radiation. HP-R-232 response usually was greater than HP-P-3236 but less than HP-R-3240.
HP-R-218 was at the 305 foot level in the fuel handling building waste disposal area and showed increased radioation about 7 to 15 minutes af ter HP-R-232.
The HP-R-218 response did not peak as much as the others and probably showed j us t the overall fuel building / auxiliary building background change.
The write-up by Met Ed/GPU indicates that HP-R-207 (305 foot level) and HP-R-204 (280 foot level) both at tne other enct of the auxiliary builoing from HP-R-232 responded aimilarly to HP-R-218.
In diagnosing these rerponses, it appears that the major portion of the released gas went out of the vent header and into the ventilation system that was monitored by HP-R-3240 and HP-R-3236. However, there also appears to be a path opening some-where near HP-R-232 that took longer to reach HP-R-232 than for the gas to get inte-the vent ilation system and out the station vent. This would indicate quite strongly that there is more than one escape point of the gas from the vent header or associated systems.
\\
TABLE V OTHER RADI ATION N0h lTORING RESPONSES, MARCH 28, 1979 instrument Starting Time Recorder Number instrument Description of increase 1901 (11)
HP-R-213 I nen re inst Panel Area 6:31 a.m.
1901 (12)
HP-R-214 R.B. Dome 6:32 1901 (7)
HP-R-209 F.H. Bridge, N 6:33 1901 (8)
HP-R-210 F.H. Bridge, S 6:34 2900 (10)
HP-R-228-P A.B. Purge Ai r Exh Downstream of Filters 6:40 19b1 (6)
HP-R-207 Intermediate Cooling Pump Area 6:41 7:19 2900 (4)
HP-R-226-P R.B. Purge Ai r Exh Duct B 6:42 2900 (1)
HP-R-225-P R.B. Purge Ai r Exh Duct A 6:42 3236 (7)
VA-R-748 Condenser Vacuum Pump Discharge 6:42 1901 (5)
HP-R-206 Make-Up Tank Area 6:43 7;4 (5)
IC-R-1093 Intermed Coolant Letdown Cooler Outlet 6:43 2900 (12)
HP-R-228-G A.B. Purge Ai r Exh Downstream of Filters 6:45 1992 (2)
HP-R-218 '
Waste Disposal Storage Area 6:46 1902 (7)
HP-R-32 36 R.B. Purge Unit Area 6:48 3264 (9)
NS-R-3901 Nuct Service Clo:ed cooling 6:53 3236 (4)
WDG-R-1480 Waste Gas Discharge Duct 6:54 1902 (1)
HP-R-215 F. H. Bridge 6:55 7:45 9:37 1902 (9)
HP-R-3250 F.H. Exh Unit Area 6:58 7:22 3264 (1)
MU-R-720 Hi Primary Coolant Letdown 7:04 3264 (2)
MU-R- 720 Lo Primary Coolant Letdown 7:04 3236 (6)
WDG-R-1486 Waste Gas Decay Tank B Discharge 7: 15 1902 (4)
HP-R-2 32 A.B. Access Corridor 7:21 3266 (6)
WDL-R- 1311 Plant Ef fluent Unit 2 7:26 3264 (10)
SF-R-3402 Spent Fuel Cooling 7:29 3264 (8)
D C-R- 3399 Decay Heat Closed A Loop 7:45 3264 (9)
DC-R-3400 Decay Heat Closed B Loop 7:45 3236 (5)
WDG-R-1485 Waste Gas Decay Tank A Discharge 7:52 1902 (6)
HP-R-234 Control f, Service Bldg Access Corridor 7:55 9:46 8-5 ll82 112
On April I there was an ef fort to vent the waste gas decay tanks to the reactor containment ballding. The line was to be connected from radiation monitor WDG-R-1486 on the outlet of waste gas decay tank WDG-T-1B through a flame arrester and into the reactor containment building through penetration R-571C.
After four attempts in which various leaks were found and repaired or isolated, a tight system was obtained. The rudiments of this effort are explained in
~
Appendix C of the Met Ed/GPU Preliminary Report on Sources and Pathways of TMI-2 Relecses of Radioactive Material and SOP Z2. Radiation monitor WDG-R-1486 1s on 46 level 105 feet of the auxiliary building.
During in.cial attempts to utilize the venting system, leaks were found at WDG-R-1486 and WDG-R-1485 The gas to be vented was probably the most concentrated radioactive gas in the system exclusive of the make up tank and the primary coolant system.
A review of the HP-UR-1902 strip chart of April 1 at 4:31 a.m. and 6:30 a.m.,when waste gas valve WDG-V-30B was opened showed that HP-R-3240 and HP-R-3236 responded within several minutes (2 - 3). HP-R-232 responded only a minute or two af ter HP-R-3240 and HP-R-3236.
HP-R-218 was responding at about another 5
minutes af terward.
These responses are Just about what would be exp5cted for an open release in the auxi liary building. The ventilation system immediately starts exhausting the radioactive gas and hence the relatively quick response of HP-R-3240 and HP-R-3236.
HP-R-232 which is some 100 to 150 feet away would probably see the cloud due to 13,47 diffusion through the auxiliary building. HP-R-218 is even more isolated by being over against the opposite side of the fuel handling building and should probably see it well af ter HP-R-232.
Since these were known to be releases directly to the atmosphere in the auxiliary building, the response times with respect to the venting of the make-up tank to the radwaste vent header showed that the released radioactivi ty from the make-up tank had to follow some torturous path before exi ting the sys tem.
During early March 29, there were indications on the HP-UR-1907 strip chart that the ventilation was secured from about I hour 5 minutes (1:05 to 2:10 a.m.). The
))Of 8-6
radiation monitors started rising immediately. The various monitor readings went up by the following factors:
HP-R-3236 - 2 HP-R-218 - 20 HP-R-3240 - Off scale & then a HP-R-234 - 60 factor of 5 HP-R-215 - 400 HP-R-232 - 8 Af ter the ventilation was restarted at 2:10 a.m., March 29, the monitor readings started to decrease. This indicated that radiation levels within the auxiliary building were being kept lower by continued exhausting of auxiliary building air.
g.7
9.0 DISCUSSION AND RATING OF POTENTIAL PATHWAYS The release of airborne radioactivity from the TMI-2 was not a single episode but a series of events that were unavoidable given the design of the plant. The analyses of the fission products that escaped from the reactor core indicated that they had been processed through water prior to being released. The major isotopes were xenons, krypton, and some iodine. The releases were somewhat complex in that several events occurred before the releases were made.
The releases for the most part, were involuntary, but several were specifically made in attempts to gain better control of the plant.
The pathways primarily considered in this report are those that had high potential of significant leakage of radioactivity. The major characteristic of these potential pathways was the proximity to the reactor primary coolant.chich was under pressure.
It should be understood that had the circumstances and results of the accident been different, a dif ferent evaluation would have been performed to address that situation.
It is believed that events during the first hour of the accident set up some of the conditions contributing to the uncontrolled release of radioactivity. The destruction of the core completed the setup.
The pressurization of the reactor coolant drain tank is believed to have caused water to be forced into the auxiliary building vent header, probably at some elevated pressure. This pressure is believed to have damaged the internals of the liquid traps of the vent header, thereby setting up one of the leakage pathways.
The water could have damaged such other components as valves and the operation of the waste gas compressors. The likely release points of the radioactivity escapes, evolved with venting the make-up tank to the vent header, were the damaged traps or other components.
The relief valves of the waste gas compressor vent into the auxiliary building sump tank.
If there had been major releases early in the accident (less than 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />) due to lif ting of these relief valves, it would be expected that 1182 1!5
radiation monitor 232, which is the closest monitor, would have responded. This monitor did not exhibit the expected response which would have been expected had the release been from the auxiliary building sumo tank.
The reactor coolant bleed hold-up tanks are the recipients of the relief valves of the make up tank and of downstream of the block ori fice valve. The routing of letdown flow, via the three-way valve MU-V8, into the reactor coolant bleed hold-uptanks on March 29, also increased the pressure in these tanks. The lifting of relief valves from the reactor coolant bleed hold-up tanks discharges fluid into the relief valve vent header that is connected, unencumbered, to the station vent.
The attempts and final _ success of venting the waste decay tanks to the reactor containment building were traceable and present essentially no uncertainties as to pathways and time of release.
The analyses of the waste gas decay tanks indicate high hydrogen content which srb-stantiates the thesis that the radioactive gases were carried along by the degassing process.
There were some very low radioactivity-content water releases as explained by the NRC.
Since the release was so small, little is said about liquid releases in this report.
A rating of pathways can be made by examining data from the operations and releases. This provides a reasonable ranking of pathways. In the order of importance, the following appears consistent with the data:
1.
Reactor Coolant Letdown /Make-up System 2.
Reactor Coolant Drain Tank Vent to the Vent Header in the Auxiliary Building 3
Reactor Coolant Drain Tank to Reactor Coolant Bleed Hold-up Tanks 4.
Reactor Coolant Drain Tank Vent to the Reactor Coolant Bleed Hold-up Tanks 5
Reactor Building Sump to Auxiliary Building Sump The other pathways appear to be negligible with respect to the above list.
9-2 i182 116
10.0 M05T PROBABLE PATHWAY The most probable pathway is the letdown /make up system. The venting of the reactor coolant bleed hold-uptanks and make up tank to the vent header and the pressure relief valve lif tings of the reactor coolant bleed hold-up tanks are considered to be the major pathways of uncontrolled radioa'ctive releases. Al!
other pathways basically terminate when isolation of the reactor containment building occurs.
During the first 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> 56 minutes, some release may have been through the reactor coolant drain tank vent to the vent gas header in the auxiliary build-Ing.
However, the transfer of water to the vent with the pressurizing effects on the vent header is believed to be the most important aspect of this pathway.
In summary, the fission products were r eleased to the reactor primary coolant system and exi ted the primary system through the letdown /make-up system into the auxiliary building. The fission products were released through the damaged vent header. to the auxiliary building and the fuel handling building and from the lifting of the pressure relief valve or valves of the reactor coolant bleed hold-uptanks that vent directly to the environs through the station vent.
From the analyses of the charcoal from the auxiliary building exhaust and the fuel handling building exhaust, a major portion of the airborne activity release in there buildings was exhausted through the fuel handling building exhaust.
1182 117 10-1
0 9
9 APPEND!CES 1182 118
e APPENDlX A DISCUSSION OF METROPOLITAN EDISON / GENERAL PUBLIC UTILITIES CORPORATION'S PRELIMINARY REPORT ON SOURCES AND PATHWAYS OF TMl-2 RELEASES OF RADIOACTIVE MATERIALr JULY 16, 1979, REVISION 0 The Metropolitan Edison / General Public Utilities Corporation's Preliminary Report discusses pathways for the transport of radioactive material from the reactor building into the other plant building and finally to the environment, isolation of the reactor containment building due to pressure increases, radiation monitor-ing records, selected portions of the building ventilation systems,and preliminary conclusions.
The report reflects extensive discussion with plant operating personnel.
It depends on the reader having the Burns and Roe, Inc., mechanical flow diagrams and general arrangement drawings included in the THI-2 " Green Book".
The report will be a good checkpoint for diagnostics af ter the plant is accessible because there is considerable detail for each pathway examined.
In Section 2.0, " Pathways for Transport of Radioactive Material Following the TMl-2 Accident," there is a comprehensive compilation of the potential pathways. There were copious details of relief. valve locations, setpoint pressures, and discharge recipients, it also included pertinent control room operators' logs and shift foremen's logs and relevant sequences of events.
In the preparation of this report to the President's Commission, information con-tained in the Met Ed/GPU report was checked by this author and found to be accurate.
^-'
1182 119
The appendices were well developed and rational. The description of the radiation monitors and the responses of the instruments were comprehensive.
The development of information to evaluate the potential siphoning of the reactor building sump to the auxiliary was well done. The conclusions appeared to be appropriate up to the point of development.
1182 120 A-2
APPENDIX B' DISCUSSION OF THE NUCLEAR REGULATORY COMMISSION'S EVALUATION OF RADIOACTIVE RELEASE PATHVAYS The Nuclear Regulatory Commission's evaluation of the release pathways for the TMI-2 accident is written as Details 11 Radiological Aspects, Section 3.1.1, Liquid and Gaseous Pathways.
It utilizes the same information as the Met Ed/GPU Preliminarv Report on Sources and Pathways of TMI-2 Releases of Radioactive Material.
The NRC discussion is a small part of NUREG-0600, investigation into the March 28, 1979,Three Mlle Accident by the Office of Inspection and Enforcement. To under-stand it adequately, one must firsi either read and understand the repc u.
(Appen-dix l-A, Operational Sequence of Events or Appendix Il-A, Radiological Sequence of Events), be completely knowledgeable of plant designs and operations at Three Mile Island, or be well familiarized from other descriptions of the accident.
T One specific item may be somewhat controversial.
It was stated that " Loss of seal water resulted in significant leakage from pumps WDL-P-5 and B, which take suction on the reactor coolant bleed tanks (Ref 127)." Loss of seal water should not automatically result in signficant leakage although it could if the seals were damaged.
The Met Ed/GPU report addresses the problem without conclusions.
Final resolution will not be available at least until cleanup of the auxiliary building is cccomplished.
In the overall comparison, there is reasonably good agreement about NRC pathways among Met Ed/ CPU and this author.
fll B-1
t APP [NDlX C COMPARISON OF RATING OF PATHWAYS OF RADI0 ACTIVITY TO THE ENVIRONS Metropolitan Edison /
Nuclear Regulatory Author of Report to Pathway General Public Utilities Corp.
Commission President's Commission Letdown /Make-Up System Major pathway to auxiliary bldg Same Same Reactor bldg sump to auxiliary 8,400 gal of low-act ivi ty water Same Same bldg sump Reactor coolant drain tank to Prot [ ably not significant Neither referenced Probably added to water in reactor coolant bleed tanks nor discussed.
ventory in auxiliary bldg.
Reactor coolant drain tank Discussed nominally Considered signifi-Considered most important vent to vent gas header cant in damaging vent in degrading of vent head-header system er components originally Reactor coolant drain tank vent Not of great significance Same Same to reactor coolant bleed hold-n up tanks Reactor coolant pump seal water Not discussed Not discussed Low probability to seal return coolers Letdown coolers cooling water Low probability Not discussed Low probability Leakage coolers cooling water Low probability Not discussed Low probability Make up pump seal leakage Low probability Not discussed Hot discussed; low proba-bility Relief valve by block orifice Probable path to bleed holdup tanks Mentioned only Considered important path-valve way to bleed holdup tanks Relief valves on purification Possible pathway Considered important Believe block orifice demineralizer of letdown /make-pathway to auxiliary valve relief valve MU-R1 u? system building drains would open first to prev-ent discharge from these valves.
Relief valve on make-up tank Happened Happened Happened outlet a
N N
Metropolitan Edison /
Nuclear Regulatory Author of Report to Pathway General Public Utilities Corp.
Commission President's Commi = ion Relief "alves on bleed holdup Believed no release Considered a signifi-Considered a significant tanks cant release pathway release pathway Vent ft>m bleed holdup tanks Not discussed Believed releases thru Believes releases thru to vent header vent header could have vent header could have been significant been significant Radwaste pumps seal water Possible contaminated water leakage Considered potentially Not discussed; not con-significant sidered signi ficant. Double seal failures are required
& little evidence of activ-Radiation Monitor HP-R-227 Not likely Not discussed
- 10 d scu ed.
Intermediate closed cooling
'Jo t likely Not discussed Not likely water system E
Reactor building Not likely Not likely Not likely Steam generator Discussed but not significant Considerable discus-Not discussed; not signi-sion; not significant ficant Liquid water Discussed but not signif. cant Discussed but not Not discussed; not signi-significant ficant Waste gas compressor Briefly discussed; no comment Dis:ussed; possible Not discussed; not con-pat 1way sidered particularly sig-nificant Other miscellaneous leakage Briefly discussed; not s i gni fi cant Briefly discussed; Not discussed; not con-paths not significant sidered signi ficant M
rx]
M
\\
U
REFERENCES Ref. No.
I NUREG-0600, investigation into the March 28, 19 79, Th ree M i le Island Accident by Of fice of Inspection and Enforcement, Nuclear Regulatory Commission, August 1979.
2 NSAC-1, Analysis of Three Mile Island - Unit 2' Accident, Nuclear Safety Analysis Center, Electric Power Research Institute, July 1979 3
Preliminary Annotated Sequence of Events, March 28, 1979, Metro-politan Edison / General Public Utilities Corporation, July 16, 1979, Rev. 1.
4 Strip Chart Recorder, FR 7100.
5 Preliminary Reoort on Sources and Pathways of THI-2 Releases of Radioactive Material, Metropolitan Edison / General Public Utilities Corporation.
6 Orawing Number 2024, Rev. 25, Burns and Roe, Inc..
7 Strip Chart Re co rde r, HP-UR-2900 8
Drawing Number 2644, Rev. 4, Burns and Roe, Inc..
9 Strip Chart Recorder, HP-UR-1907 10 Strip Chart Recorder, HP-UR-3236 11 Strip Chart Recorder, HP-UR-3264.
12 Strip Chart Recorder, HP-UR-1901 -
13 Strip Chart Recorder, HP-UR-1902.
14 Strip Chart Recorder, Auxiliary Building Exhaust Flow, FR 5313 and FR 5286 15 Strip Chart Recorder, Fuel Handling Building Exhaust Flow, FR 5709 and FR 5659 16 Strip Chart Recorder, Source and intermediate Range Power Level.
17 Discussion with Nuclear Safety Analysis Center Staf f.
18 Brunswick-2 Wa te r Chemis try, A. D. Miller, NEDO 21810, General Electric Company Report, February 1978.
19 Instruction Book for Heliflow Letdown Cooler, Babcock & Wilcox i ns truc t ion 620-0005 20 Drawing Number 2C64, Rev. 4, Burns and Roe, Inc.,
21 Drawing Number 2045, Rev. 19, Burns and Roe, Inc..
22 Discussion with L. Noll, Metropolitan Edison Co..
23 Telephone call to V. Fricke, Burns and Roe, Inc..
24 Drawing Number 2028, Rev. 26, Burns and Roe, Inc..
25 Drawing Number 2632, Rev 9, Burns and Roe, Inc..
1182 124.
26 Drawing Number 2060, Rev. 20, Burns and Roe, Inc.,
27 Drawing Number 2027, Rev. 24, Burns and Roe, Inc..
28 Telephone call to H. McGovern, Metropolitan Edison Co..
29 Drawing Number 15837, Bingham-Williametta Company.
30 Control Room Operator and Foreman Logs, Unit 2 31 Interview with Cont rol Room Operator C-(see Ref. Il 32 Log typer.
33 Alarm printer.
34 Interview with Radiation / Chemistry Technician D (see Ref. 1 ),
35 Make-Up Tank Level Chart Recorder.
36 Interview of Station Operations Supervisor (see Ref. 1) 37 Interviews of Shif t Supervisors C and E (see Ref. 11 38 Interview,K. Brian, Metropolitan Edison Co..
39 Interview, G. Hitz, Metropolitan Edison Co..
40 Drawing Number 2168, Burns and Roe, Inc..
41 Drawing Number 2169, Burns and Roe, Inc..
42 Reactimeter Data.
43 Chemical Engineers Handbook, John H. Ferry, McGraw Hill.
44 Burns and Roe, I nc., Memo DRMG flumber CCR 491 45 Discussion wi th J. Bremmer, Metropoli tan Edison Co..
46 50P 2Z, Metropolitan Edison Co..
47 Drawing Number 2065, Rev. 19, Burns and Roe, Inc..
48 Interview of B. Mehler, Metropolitan Edison Co.,
I