ML19338F286

Forwards Draft Annotated Chronology of Operational Aspects of TMI Accident Through 790401 Per NRC Request

ML19338F286
Person / Time
Site:	Crane
Issue date:	06/29/1979
From:	Johnston W NRC - NRC THREE MILE ISLAND TASK FORCE
To:	Frampton G NRC - NRC THREE MILE ISLAND TASK FORCE
References
NUDOCS 8010150637
Download: ML19338F286 (32)

Text

_

to t.

June 29,1979

~

1 ME!ORAllDUM FOR: G..Frampton FR0ft:

W. V. Johnston, Task Group 2 Leader

SUBJECT:

CHRONOLOGY OF THI EVEllTS f

Enclosed is an annotated chronology of the operational aspects of the j

TMI accident through April 1 per your request. The chronology for the first i

16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> was assembled principally from the IaE and Met Ed.

Chronologies i

available in May.

Beyond that time, log looks and IaE duty status reports s

(

are used. Additional analysis and interpretation was done by staff based upon interviews and other source documents.

y:

s William V. Johnston Leader, Task Group 2 l

TMI Special Inquiry cc:

M. Rogovin

~

K. Cornell B. Budnitz R. DeYoung i

Task Leaders t

i i

oenan v..Tas k.. 42.. Lead ec...............

eunuaua >

, H.VJohns lod....

............ ~ -

.~~~*-.-

.. ~ ~ ~ ~.

i as, c >

, 6 /..../. 7. 9. ~...

-.... ~ ~....

..... ~.....

...... ~.....

~. - ~ ~.. -

....~~~~

NIC FORM 318 (9 76) NaCM 0140 W w*.****a""*""'.n...,nes,u n.

.e. vee 8010150 h 9 g

~~

~

Draft Chronology of TMI-2 Accident - Task Force 2 (Based primarily on IE chronology of May 8 and GPU chronology of May 10)

Plant Status Immediately Prior to Turbine Trip Just prior to the turbine trip at 4 a.m. on March 28, Tai-2 was operating at roughly 97 percent full power (~930 MWe). The status of the Reactor Cnolant System (RCS) was nornal; the flow rate through the letdown line was about 70 gpm, and one mga eup pug (MW-P-1B) was operating WafG'

' O #b to supply % cooling 4 o thej[four reactor coolant pugs ('tCP) and to t

provide normal reactor coolant makeup. Leakage on the order of six gpm from the RCS was known to be coming from the two safety valves and/or the power-operated relief valve (PORY) at the top of the pressurizer.

Figure 1 shows a siglified schematic of the RCS.

O The stat.s of the plant secondary side also appeargto be essentially normal Work was underway on the polishing units to transfer spent resins hvh to a holding tank. Apparently, this work caused a loss of instru ' -

ment air to the outlet valves of the polishir.g units, causing them to M

shut. This(closure cut off the water supply to the condensate booster pugs, causing them to trip. The trip of these pumps caused the subse-quent trip of one of the condensate pugs, the main feedwater pumps, and the main turbine. Figure 2 shows a simplified schematic of the plant secondary side.

est e a

f C

W W

m.

4y-3 o

w H.

\\

h 4,7 _ =1 25 E

8 m2 E

or

=

l u<

~

=

- =.

r c.

-e 5

c g,

s =

3 U

M t

J Os e.

M 8

3 2

Nd i

a M

3 Z

2 1

N C

a r

l E

$2 1 9 lw ^

T 'I r

~

J

~

~

5 Ecb 8d g

1,.

=

n-(t 3

oc=

=,

co<

on

--g,

$5 1 E,kjE h'

S g

s w'

Jl 4

~

r o E

E E3

5 g

x y

g E

o 2

=

a a

Y U

C W

2 9

s a

a

~

V' h

h I

t e

u I

E 8

h 5

O e

A s

! kN 1..$ r N

~

s i

5

-s 1

o o

Q T

b x

>u u

]

w 5

m H

k

\\

q 3

5 s wN 4

4 6

3 5 >%

5 r

N m

"E E II wL 4

z o

.U c

- O 5,,)

l

'p' 9%

S(

l 0

s.

gi i

r.'-..

)

s ll 1

s) f e =*

\\

i 11 s

5 ss 33 j

. ~

T, 5

s s s

)

RE 3

~a

~W G

-*.1 C f1

-W g

,1 f x

t, y

r

-s 3f 7.

v-i rf 4

x a

so e

1 b

i

$3

~~-

w

s. v.

I [2 7. 5:5 =.- 504

!N l~

l m g

2 N

g

'a 10[ I !7 E ( v = 23 u 7 r~ X Ea v gx ? l e e, E (M# [ ~ 1{= 1 z i I C t t = 'u

3 ag o.

~ O E t I 1,3%.. h5 I]S j ha! T T E3 3:2 I ~3 3: plcapss. -x,v 58..s y L I O r is k f 'e * ~ -x v - b

s4

/J7

3 X

ij i. 52 ,F 31 =

=

i

5. i-s r

i _j 9

75 2

2a4 s& ' g =YT)$ ~ L g x J i < g ,J ; sj A 3 X Y

e 2 Approximate Elapsed Time Event 0 sec. Main feedwater (MFW) pump trip and turbi trip. (4:00:37 a.m.) Emergency feedwater (EFW) pumps act ed automatically. (EF-P1, EF-P-2A, EF-P-2B). 3 sec PORY opens on increasing RCS pressure (2355 psia). 8 sec Reactor trip occurs on high RCS pressure.(2455 psia). 13 sec. Letdown flow isolated, makeup pung MU-P-1 A actuated and HPI isolation valve MU-V16B opened (Met Ed chronology says MU V16A) by operator to help minimize reactor coolant shrinkage after a loss of feedwater trip. (Note that the IE chronology suggests that letdown nay not have been secured until about five ndnutes.) POP.V closure setpoint reached but valve does not clo'se. PORY status light in control room indicates that valve has closed (Note that status light only shows the electri-cal state of the valve, not the actual physical state of being open or closed). 30 sec. Alarm printer in control room prints out PORY and safety valve discharge line temperature alarns. (Operators state this is not unusual after a reactor trip.) 38-40 sec. EFW discharge line valves (EF-Vil A and EF-Y11B) ipen automatically on low steam generator level. EFW flow to steam generators preventd by closed EFW block valves (EF-V12A and EF-V12B). \\ i e y-__-,,--..w,- .c

3 PLANT STATUS The plant appears up to this tinie to be experiencing essentially a normal response to a loss of feedwater initiated turbine trip and reactor trip. Pressurizer pressure and level are decreasing as expected; steam generator level has decreased to the point where EFW should begin to flow into the steam generators. Makeup punps MU-P-1A and MU-P-18 continue to deliver water into the RCS. Two problems now exist which will begin to alter the course of the transient: the failure of the PORY to reclose and the closure of the EFW discharge line block valves (EF-V12A and EF-V12B). 4 / e 9

iff 4 i min Pressurizer level has reached a minimum and begins to increase. Pressurizer pressure continues to decrease. Alarm printer prints out safety valve discharge line temperature alarm. ~ 2 mi n. Actuation of the Emergency Core Cooling System (ECCS) in the High Pressure Injection (HPI) mode occurs on low RCS pressure (1600 psia). Makeup pump Mt'-P-1B is tripped; makeup pump MU-P-1C is started and aligned (along with makeup ousp MU-P-1 A) to draw water from the Borated Water Storage Tank (BWST) at a rate of about 500 gpm per pump, injecting through valves MU-Y16A, MU-V183, MU-Y16C, and MU-Y16D into the RCS cold legs and the vessel. 3-4 min. ECCS actuation signal bypassed by operator, allowing operator throttling of and/or tripping of makeup pumps. Makeup pump MU-P-1C subsequently tripped and makeup pung MU-P-1 A discharge valves (MU-V16A or B?) throttled to decrease the flow to RCS. Alarm printer prints out reactor coolant drain tank (RCDT) high temperature aldrm. 4 ,------n,,--.. .,--,c e

5 PLANT STATUS Pressurizer level has begun increasing while prssurizer pressure continues to decrease (see Fig. 3 for trends). RCS coolant continues to flow out the PORY to the Reactor Cool. int Drain Tank. Steam generator level has decreased such that the steam generators are essentially dry. ECCS actua-tion has occurred on low RCS pressure; subsequent oper. me actions have 2-reduced nakeup purp/HPI flow rates to the RCS. Pressurizer pressure and level indications now are apparently inconsistent; i.e., level has it. creased rapidly while pressure continues to decrease. Pressurizer level turnaround and increase apparently are a result of RCS coolant expansion due to the conbination of the open PORY at the top of the pressurizer, the large decrease in heat removal from the RCS because of steam generator dryout, and because of water flow into the RCS from the makeup pumps. Subsequent operator actions to decrease makeup / HPI flow suggest the reliance on pressurizer level as a true indication of RCS coolant levels. The intent of these operator actions was apparently to prevent the pressurizer fiom filling conpletely with water. Various high temperature alarms in the PORV/ safety valve discharge lines and the RCDT have now been printed on the alarm printer in the control Apparently the operators associated these alarms with the normal room. e opening of the PORY. 9 e

s , w s. .~s s m, 3., s 3 %.e is v 4* 3 Sco-- o,. IRESS vb'AES. bt\\1P.L... d 9 ~d \\\\ 0 xo s_ _.- ___-

• )

~ Op ..I . tl. _.. _ - Loo-4 l l O L 2 3 + r n 74KE MT *TC t. %..V.RP., a..PJ C __%_N..}- (MiwvTu) _. ~ _.... _ ~ _ ~E1. 0 0 - - - ~ ~ - - - ~ ~ - - ~ - - - - ~ -- * * - - ~ - - ~ ~ - - - - - - - - - - k - M.ODVY5___._ I

• -b

'A 0 0 0 - - *. = =. _.. _.. _.,.. v y W 3.. C O 9 Qoa. _.

5 min. Operator initiates letdown flow in excess of 140 gpm in attspt to stop prssurizer level increase (Note: Met Ed chronology indicates letdown begun at 7 min., after pressurizer water-solid and at a rate of 71.4 gpm). 6 mi n. Pressurize steam bubble lost - water level goes off-scale high. RCS hot leg temperature and pressure rach saturation conditions (580*F at 1350 psia), so that some bulk boiling of coolant in the core nay be expected to begin. 7-1/2 min. Reactor Building surp pung A )WDL-P-2A) starts automati-cally on high sunp level. Increasing level in the sump is due to water discharge from the RCOT relief valve ( WDL-R1 ). Apparently, this water was pumped to auxiliary building sung tank outside the reactor ~ ~ - building. ) l l l i 1

't. 7 8 min Operator discovers very low level indication for the steam generators and that EFW block valves (EF-VI2A and EF-VI2B) are closed; he opens them and begins the refilling of the steam generators. 8-10 mi n As steam generators bcgin to refill and cool the RCS coolant, RCS pressure and tegerature drop, and pressurizer level comes back on scale. Since saturation conditions still exist in the RCS, some boiling in the core would be expected, but adequate core cooling still exists. 10 mi n. The second Reactor Building sug pug (WDFL-P-28) starts on high sug level, puging water discharged from the RCDT relief valve to the auxiliary building sum tank. 101/2 min Makeup pug MU-P-1A tripped, restarted at about 11 1/2 minutes. Apparently, no makeup flow or seal injection occurred in this period. 11 mi n Reactor Building sug high level alarm received (on alarm printer?) indicating that flow into the sug is greater than the capacity of the 2 sug prvs (i.e., level in sug is still increasing). 15 min RCOT rupture disc bursts at about 190 psig releasing tank overflow to Reactor Building. Pressure indication (in control room?) shows decrease to about 10 psig. Pressure rise of about 1.5 psi (?) shown on Reactor Building pressure indicators. 18 min Reactor Building exhaust monitors show decade increase in radioactive reading (of what?) (increase from cpm to cpm?) Small rag increase in iodine radioactive reading in fuel handling building exhaust monitors.- This is a relatively minor increase resulting from radio-active material being released from the RCS coolant in the Reactor Building sug. 25 min Operator requests (from computer) tenerature in PORY discharge line which is indicated to be at 285*F. Operator believes this is attributable to initial opening and closure of PORV, not a continuous discharge. Radiation alarm received at radiation monitor panel (in S. control room?) from intermediate cooling system monitor (IC-R-1092) (monitoring letdown coolers?). Monitor is physically located next to Reactor Building Sump. (Met Ed says this alarm is normally received after a reactor trip). (Alarm level is cpm?). (Note: IE chronology has this alarm at 60 minutes.) l l

8 d6 min. EFW purp EF-P-2B tripped by operator after both steam generators were filled to the 35 inch level. Apparently, EFW-P1 had been tripped by the operator prior to this ti me. EFW pump EF-P-2A is still operating. 38 min. Reactor Building sump pumps turred off by the operators. 73 min. Alarm printer in control room fails because of signal overload returns to service at 2 hr. 47 minutes. (Note: Met Ed has different times for outage of printer 2 hr. 27 min., 5 hr to 5 hr.17 min.). Reactor Coolant Pumps RC-P-1B and RC-P-2B tripped by operator because of vibration alarns caused by pumping of two-phase edxture and loss of RCS coolant inventory. Procedures require RCP trip on such alarns. 1 80 min. Operator requests from (plant process?) conputer terpera- ) ture of PORY discharge line, which indicated 283 F. Radiation levels at letdown line monitor increase by j a decade (from to ). j i 86 min. Steam Generator B isolated by operator because of a sus-pected leak from the RCS into the steam generator. 90-100 min. Source rang. and intermediate range neutron instrumentation show fluxuations of up to a decade. Operators believed this was due to a reactivity (Neutron production rate) increase and ordered emergency boration to compensate. In reality, these instrumentation fluxuations were apparently due to changes in the. density of the coolant caused by boiling in the core. Steam generator A level has decreased such that secondary side is essentially dry until operator begins refilling as part of an atteapt to establish natural circulation. 100 min. Reactor coolant pumps RC-P-2A and RC-P-2B tripped by operator due to vibration alarns. Source range neutron instrumentation count rates dropped by about two decades after the trip of the loop A RCPs. S

9 PLANT STATUS Because of the stuck-open PORV, RCS water continues to be lost. As the water level in the RCS drops, the steam produced in the core is being less effectively cooled in,the steam generators. As a result the water entering the Reactor Coolant Pumps contains increasing amounts of steam. This ten causes cavitation and vibration problems in the pumps. Because of this the operator trips the B loop RCPs. When this occurs, additional water condenses and collects in the B loop to the detriment of the flowing t + A loop. The continually degrading flow in the loop A RCPs is thus further harmed; operator action to trip these RCPs is soon after required. With this action, the liquid levels collapse to the low points in the CS covering the core to a level of about one foot above the fuel. The neutron instru-mentation count rates decrease, indicating an increased liquid density in the core. At this time the RCS is routhly 1/3. liquid water (in the core region, in the bottom of the steam generators and in the pressurizer with the remainder of the RCS filled with steam. All forced cooling in the RCS stops. A number of problems have arisen in the plant secondary side. The opera-tors have had trouble maintaining the desired levels in the steam genera-tors; one EFW pump (EF-P-2A) is apparently being used to maintain level in both steam generators. These levels have been kept low by the opera-tors; steam generator A level dropped to essentially zero at about 95 minutes before operator action began to refill it. Because of suspected leakage, steam' generator B has been isolated. Additional problems were arising in the main feedwater system. The condenser hotwell apparently was overfull, and the problems were being experienced as the operators attempted to restore the condensate and condensate booster pumps to service. Up until the time of RCP trip, the core has been kept relatively cool and apparently has not experienced any significant damage. While 'a signi-i ficant amount of'RCS coolant ( nearly 2.3) has been lost and bulk coolant boiling is occurring, the RCP flow appears to have been sufficient to keep the core substantially covered and cool. Some radiation alarms have occurred but actual radiation levels are still relatively low. N O i i 1 f e ,m,. _. _ ,., _... _.. _ - _,. _., -. ~ _. _.. _ _ _.,. ,__.-,,,,_,,,__.--_r. .._.-_,_.m_nr,_ . _,,.,,,_--._.4

s. 10 1 hr. 40 min Source range and intermediate range neutron instrumentation to 2 hr.10 count rates increases by about two decades, indicating min. that significant boiling and voiding in the core coolant is occurring, and that water levels in the core region are dropping such that the fuel is beconnnng uncovered. RCS hot leg and cold leg temperatures immediately begin to diverge widely (see Fig. J), indicating a large temper-re rise through the core. + Decay heat generation rates are now h.'aut 1 percent of full power ( 27 Mw). RCS presesre is now about 900 psia and dropping. 2 hr.10 min. RCS loop A hot leg temperature goes off-scale high ( 620*F). 2 hr.17 min. Operator requests (from plant process corputer?) tempera-tures of PORY discharge line, which indicates 228 F. ~ 2 hr. 22 min. Operator c1oses PORY block valve (RC-V2), stopping loss of coolant out the PORY, and causing RCS pressure to begin to rise. 2 hr 28 min. RCS loop B hot leg temperature goes off-scale high ( 620*F). 21/2 hr to Operator raises steam generator coolant level and returns 2 3/4 hr.. it to service in preparation for start an RCP. Radiation monitors, alarms associated with letdown line indicate significant increase in count rate (factor of 100 increase). 2 hr 50 min. Reactor coolant pump RC-P-2B restarted, reestablishing flow in RCS. RCS pressure now approximately 2100 psia. Site emergency declared (because of ). Alarm printer returned to service (according to I&E chronology). i w j 0 ._,,-.,_,..,_,.,_____y__

0FF SCALE / b I 600 LOOP A T ll0T LOOP 11 T 110T. ; ,4-t L L PATil0T tg. 500 CUT pj PUMPS 23 0FF [5 li0b !'= w LOOP A TCOLD i-i t'; fj300 C a i u-i gi g y -1 I A LOOP [2 1ItEACT0lt 200 l COOLANT [LOOPBT W [ PUMP j I COLD 1 i 100 8 10 12 14 16 TIME l6 0 '2 4 AFTER TURBIf4E TRIP - Il00RS FIG 11 I i f B

12 PLANT STATUS This is the period when the core suffers major danage. The decay heat being generated in the core is 27 megawatts (1 percent of full power). At this power level, the waterr will boil out of the core at an average rate of about one foot every 4 to 5 minutes. As the water level drops, the uncovered portion of the full rod begins to heat up. Depending upon its local heat generation and the cooling by the steam passing over it from boiling below, the rods will heat up aty a rate of 50 to 100 F per -ni nu te. Ur. der these conditions the cladding of the fuel rods in the center of the reactor will fail in 18-20 minutes (1400-1500 F), releasing the gap inventory of fission products. the fuel rods at the outside edges of the core will fail in about 40 minutes. When temperatures of 1600-1700 F are reached, the steam will begin to react with the Zircaloy cladding i on the "uel rods forming hydrogen, producing an oxide layer and causing an incr ease in the heating rate. Some of this hydrogen subsequently escapes through the open PORY to the Reactor Building. As the temperature increases, more of the radioactive fission products are released from the UO pellets ~ inside the citdding after 35 minutes portions of the hottest rods have reached 3500 F and the small smaount of unoxidized Zircaloy remaining melts and reacts with the U0, pellets, forming a liquid phase and releasing additional fission products. (The melting point of UO along is 5500 F). This heating up is assumed to terminate after 50 minutes (2 hrs., 30 minutes) because the PORY block valve is closed, HPI flow is increased and neutron detectors show reduced count rates all of which cause or indicate a reduced tempere+ure rise. Based upon the above assumptions, the distributi)n No damage to the lower 5 ft. f any of cort damage is as follows: fuel assenbly (assenblies are 12 feet high). Severe damage to center assemblies to a depth of 5-6 feet. Severe damage to other assemblies (except outer ones) to depths of ~ ~- about four feet. Little damage or oAidation to Odter assemblies to aRy elevation eXLept for ballooning and rupture above about three feet from the top. The severely damaged assemblies have now suffered loss of regulator geondty and probably have taken the form of a rubble pile of 1/2 - 3/4 " pieces and shards, caused by the subsequent quenching of the hot, embrittled fuel and cladding when Reactor, Coolant Punp RC-P-2B was started. N l l l l = t

13 2 hr. 55 min.' Source range and intermediate range neutron instrumentation count rates drop sharply, indicating that RCS repressuri-zation and ECP actuation have collapsed steam voids in the core region. HP technician reports that letdown line sanple lines reading radiation levels of 800R. ~ 2 hr. 56 min. Steam generator B isolated by operator because of radiation alarms from condenser offgas monitor. 3 hr.10 rin. EFW pump EF-P-2A tripped by operator after levels in both steam generators had rised to about 50 percent level. 3 hr.12 min. Operator opens PORY block valve in an atempt to reduce pressurizer level, which had gone off-scale high, and causing RCS depressurization from 2200 psia. Operator trips Reactor Coolant Punp RC-P-2B after receiving vibration alarms, low motor current, and no-flow indication. i 3 hr.18 min. Intermediate' range neutron instrmentation show increasing count rates, indicating formation of steam voids (boiling) in t'he core. 3 hr. 20 min. Emergency core cooling system actuation in HPI mode occurs as a result of low RCS pressure (1600 psia). Makeup purp MU-P-1C started, makeup pung MU-P-1 A already running. 3 hr. 21 min. Source range and intermediate range neutron instrumentation show decreasing count rates, showing collapse of steam voids in core. 3 hr. 23 min. A General Emergency was declared because of radiation reading of 8R/hr. in the Reactor building. 3 hr. 30 min. Operator closes PORY block valve RCS pressure i.2d dropped to approximately 1400 psia. PLANT STATUS Attempts have been made by the operators to restablish flow in the RCS. Addtion of water from the makeup pumps has been offset somewhat by the opening of the PORY block valve by the operator. At this time, the core is probably partially covered; however, up to about 20-25 percent of the core may be so hot that water cannot penetrate to provide ccoling. This hot region cc:. Lined with the prsener of non-condensable hydrogen gas make it difficult to add water without substantially increasir.g RCS pressure. e ,~, - - --.--n.. ,--__,.,,_--,,,c,-..,---,,..,-.,w_,,,,-.,-.----~rw--.,-.-

14 5 hr. 35 min EFW pug EF-P-2A started by operator to counteract de-creasing level in steam generator A. 3 hr 37 min Makeup pug MU-P-1C tripped by operator because of rapidly rising pressurizer level. 3 hr 48 min Reactor building high range and fuel handling bridge radiation monitors indicate less than 1 mr/hr. ~ 3 hr 51 min PORY block valve opened by operator because of high pressurizer level. (RCS pressure now approximately 1400 psia.) 3 hr 55 min Radiation reports at North Gate and Rt. 441 less than 1 r./hr. Station Superintendent directs outside agencies to be contacted. Emergency core Cooling System, Reactor Building isolation, and Reactor Euilding Emergency Coolers (?) actuated by high Reactor Et:11 ding pressure. 4 hr 8 min Operator attegts to start Reactor Coolant Pump RC-P-1 A without success. 4 hr 17 min Makeup pugs MU-P-1 A and MU-P-1C trip (for reasons unknown at this time). 4 hr 20 min Radiation monitor at Reactor Building dome indicates 600 r/hr. 4 hr 21 min Makeup pug MU-P-1B started by operator. 4 hr 27 min Makeup pug MU-P-1C started by operator. 4-5 hr Thermocouple readings taken by instrumentation technicians indit: ate tegeratures greater than 2200 F in at least six regins above the core. PLANT STATUS Operators continue to have trouble in establishing flow arouri the RCS, apparently due to the very hot regions in the core, and the presence of hydrogen gas and steam in the loops. The very high thermocouple readings and hot leg temperatures are reported to the operating and supervisory staff, but apparently are either not considered to be 'real, or the igli-cations of the readings were not perceived. Evidence for neutron and other instrumentation suggest additional periods of core uncovery and possible damage to the fuel assenblies from thernal shock in this time period. e _-,y,

15A 4 hr 42 min EFW pung EF-P-2A tripped by operator apphrently after steam generator levels became acceptably high (Note: Met Ed makes no note of this). 5 hr 18 min PORY block valve closed by operator to repressurize RCS (Note: IE (May 8 version) says block valve opened at this point). RCS pressure now at apprcximately 1200 psig. ~ 5 hr 20 min Emergency Core Cooling System actuation signals due to high Reactor Building pressure cleared (Note: Met Ed sequence says this done at about 4 hrs.; also incon-sistent with makeup pung trips at 4 hr.17 min.). 5 hr 54 min Filling of steam generator A begun by operator in an attengt to induce netural circulation. Steam generator filled to 100 percent in about 1 1/2 hr. t 1 1 7 )

W* 16 PLANT STATUS In this time period, additional core damage appears to be less severe in scale. The closing Of the PORV and running of the Rc:: tor Coolant pung helped the initial refilling and cooling of the core. Two additional actuations of the Emergency Core Cooling System in the HPI mooe at 3 hours, 20 minutes and 3 hours, 55 minutes aided in the covering and cooling of the core. Cold leg temperatures dropped during this time period, apparently because of the influence of relatively cool ECC water; however, hot leg temperatures remained off-scale high (>620F). Since the hot leg tenperatures remained above saturation temperatures for the range of RCS pressures in this period, boiling in the core region (and possibly additional uncover-ing) would still be expected. For the time period up through about 71/2 hours, the operators apparently opened and reclosed the PORY a number of times in an attempt to control pressurizer level, causing additional losses of RCS water. Radiation levels in auxiliary building (Rad Waste Panel) measured by crew to be 10R/hr. ? ~ i l

= e 17 7 1/2 hr. Operators decide to depressurize the RCS in an atte @ t to use makeup pugs, core flood tanks and residual heat removal (RHR) system to cool down the core. MU-P-1B and MU-P-1C are apparently running. PORY and pressurizer spray valve opened by operator beginning the depressurization from RCS pressures of about 2100 psia. 7 hr. 42 min. Emergency Core Cooling System actuation channels bypassrd by operators according to plant log, apparently so that actuation would not occur during atte@t to depressurize to RHR pressures. (Note: Met Ed does not note this action). 8 hr.12 min. High level indication on core flood tank A (iglications, if any, unknown at this time). 8 hr. 31 min RHR pugs DH-P-1 A and DH-P-1B started by operation in anticipation of reaching RHR pressure levels. 8 hr. 41 min. RCS pressure reaches set point for core flood tanks, water from tanks (slowly) begins to enter RCS. 9 hr. 4 min. Makeup pug MU-P-1C tripped by operator. Makeup pu g MU-P-1B still running. 9 hr. 50 min. Pressure pulse of 26-28 psig occurs in Reactor Building, due to deflagration (rapid burning) of hydrogen. Hydrogen produced during

18 core uncovery apparently migrated out the PORY until concentrations in the Reactor Building were sufficient to allow deflagration. Reactor Building pressure rapidly returns to a few psig. ~ Reactor Building isolation occurs and Reactor Building spray system actuates on high Reactor Building pressure, causing spraying of water into Reactor Building. Emergency Core Cooling System actuation on high Reactor Building pressure. Makeup pum MU-P-iC started. (Note: ECCS actuation apparently should have start MU-P-1 A and tripped MU-P-1B, which was operating prior to ac'tua-tion signal.) Makeup punp MU-P-1C tripped by operator about 40 seconds after actuation signal starts it. 9 Hr. 55 min. 9.eactor Building spray pumps tripped by operator. --. 9 hr. 57 min. Residual heat renoval purps tripped by operator, because attempts to rach RHR pressures did not succeed. 10-11-1/2 hr. Various low level and high level alarms from core flood tanks received. (Implications, if any, unkn own at this time. ) 9 --n--,

19 PLANT STATUS In this time period, the operators attenpted to cool the core by opening the PORY and depressurizing the RCS to the point that water in the core flood tanks could flow into the core. They hoped to then continue depres-surizing to the point that the Residual Heat Renoval (RHR) System could be ) used to provide long term cooling. For reasons not fully understood at this ) time, the RCS pressure did not fall to the RHR System set pnint. However, since RCS pressure did fall below the core flood tank set point, some water from these tanks entered the RCS. It is not clear whether this addition of water had any significant impact in cooling the core; however leg tenperatures remained off scale high throughout this attenpt. Also during this time period a large pressure pulse occurred in the Reactor Building as a result of a hydrogen deflagration. The source of this hydro-gen was the chemical reaction between steam and the zircaloy clad in the' core; subsequent migratica out the PORY released it to the Reactor Building. The hydrogen concentration had apparently been increasing since the time of the initial core damage to the point that rapid burning could occur. The Reactor Building safety features actuated as designed on high building pressure. Operator interviews indicate that it was believed that the pressure rise was the result of instrumentation nelfunction rather than a physical occurrence. 1.ater analysis of the deflagration indicate that the hydrogen concentration in the Reactor Building just prior to the event would result from the oxidation of about 1/3 o. the zircaloy in the core. This is consistent with the minimum estimates of zircaloy raction l l which had occurred during the initial uncovering of the core. m. ~ v -.-,.,--,.,----.---p ,w-,,--, --,-o,-,,,.- m

20 10 hr. 28 min. RCS loop A hot leg tegerature decreases into instrumenta- , tion range, for reasons not clear,at this time. 10 hr. 34 min. RCS loop A hot leg tegeratures return to off-scale high. 10 hr. 36 min. Makeup pump MU-P-1C tripped by operator. 1 10 hr. 39 min. RCS loop A hot leg tegerature decreases again into instrumentation range (<620F). 11 hr. Pressurizer level decreases from 370 inches to 180 inches in about 18 minutes. 11 hr.12 min. RCS loop A cold leg tegerature begin increasing, indi-O cating circulation occurring through loop A steam generator. 11 hr.18 min Makeup pug MU-P-1C periodically started and stopped to 11 hr by operator to counteract pressurizer level changes. 35 min. RCS pressure now at approximately 450 psia. Loop A hot leg tegerature 500-550*F 11 hr. 36 min. Filling of steam generator B to full height begun by operator to help induce additional RCS circulation and cooling. EFW EF-P-3 used for about 16 minutes to N i accoglish refill. 9 -c. _n-

i 4 i 21 12 hr. 30 min. Pressurizer level increases off-scale high. j i 12 hr. 48 min. Pressurizer level returns to scale. 4 i l 13 hr.15 min PORV block valve closed by operator to repressurize the RCS and collapse the remaining steam bubbles. i

l j

Condenser vacuum reestablished. i 13 mr. 23 min. Makeup pump MU-P-1C started by operator in order to re-pressurize RCS. l 13 hr. 45 min. Steaming of steam generator A to condenser through turbine bypass valves, providing additional cooldown capability 4 I for RCS. I i 15 hr. 33 min. Reactor coolant pump RC-P-1 A started and stopped 10 seconds i [ later to verify proper starting currents. j RCS pressure drops from 2340 psig to 1440 psig, causing i ECCS actuation. Makeup pump MU-P-IC starts, tripped i (by operator) almost immediately. RCS pressure returns i rapidly to approximately 2200 psia. 15 hr 49 min. RCS loop B hot leg temperature decreases into instrumentation 1 range (<620F). i 15 hr. 50 min. Rea'ctor Coolant Pump RC-P-1 A started by operator, reestab-lishing forced flow through the core and enhancing fuel ) cooldown. l

I i 22 PLANT STATUS Heat removal through steam generator A began after natural circulation was achieved through that steam generator. (It is not clear at this time how natural circulation was actually achieved). Attempts to cool down the core using the core flood tanks and the RHR system had been unsuc-cessful. The decision is made by higher management (apparently in coordination with BSW) at 13 hours,15 minutes to repressurize the RCS and start a Reactor Coolant Pump. Preparations began to repressurize the RCS and start one Reactor Coolant Purp. These preparations included closing the PORY and start' makeup pump MU-P-1C to repressurize the RCS, reestablishing the condenser vacuum, refilling steam generator B and " bumping" Reactor Coolant Purp RC-P-1 A to check its operability. At 15 hours and 50 minute, RC-P-1 A was restarted and run, establishing forced cooling of the core.

• m.

I / e l

23 i Plant Status - Wednesday, March 28, 1979 ) 5 hrs. kt this time reactor coolant pump 1A was operating with the reactor coolant 3 mins. system pressura at 1165 psig. Pressurizer temperature was 5510F as maintained 2000) by the pressurizer heaters with the pressurizer level at 397 inches. Core cooling was being achieved by steaming from steam generator A to the main condenser and steam generator B was isolated. Reactor coolant makeup pump B (MU-P-18) was operating to supply the reactor coolant pump seal injection flow. The primary system temperatures were: loop A cold leg temperatures 2560F; loop B cold leg temperature 2520F; loop A&B hot leg temperatures were 0 off-scale low, that is, they were less than 520 F. Prior to starting reactor coolant pump 1A at hour 15:32:42, 7 incore thermocouples were reading well above the saturation temperature in the primary system. When the pump was started, flow was established in loop A and a number of the thermocouples dropped significantly in their temperatures. ~ Two dropping immediately below the saturation pressure of the primary system, with a general downward trend being established for the remaining thermocouples. The operation of the reactor coolant pump 1A established flow in loop A with backflow being established in loop B. This process resulted in a gradual l 0 cooldown of the primary system of approximately 10 an hour initially over the first four hours. At this point, there were two main concerns: (1) the size of the voided region (the " bubble") in the reactor vessel and the mecns by which it might be removed ar.d (2) the appropriate means for continued long term cooling of the core. At this time planning efforts were being made to prepare to put the plant on decay heat removal. That is, to cool down, depressurize and go on the decay heat removal system. At this time the reactor building temperature and 0 pressure had stabilized at approximately 150 F and atmospheric pressure. The rapid fluctuations seen earlier in the accident were not apparent at this time. 7 hrs. In preparation for placing the decay system into service, the manual blockvalve 5 mins. for the decay heat pump supply to the pressurizerspray line (DHV 187) was opened. 2125) At this time the plant operators were conducting what they considered to' be a slow plant cooldown and depressurization. This cooldown/depressurization was 0 maintained until 2230 hours at which time a temperature of 300 F and 1000 psi pressure were maintained.

24 . 3 hrs. The control room log notes that at this time a bubble was established in the 3 min's., pressurizer. Further information needs to be gathered to understand the 2218) events at this point because previous information indicated a pressurizer bubble. ' 3 hrs. At this point letdown flow was lost. It appears that the filters in the let-5 mins. down system had become plugged. From this time until about 6 o' clock in the 2235) morning, various attempts were made to reestablish letdown flow. These were unsuccessful until 0631 in the morning at which time an increase of the let-down coolers intennediate cooling water temperatura to about 1000F permitted flow of approximately 25 gpm letdown.* Thursday, March 29, 1979 0 hrs. The ventilation for the auxiliary and fuel buildings was secured. l 5 mins. 0055)

.2 hrs. Ventilation for the auxiliary and fuel building was restarted. At this time l 0 mins. an alarm was received for high particulate and gas concentration in the control

'0210) room and all control room personnel put on respirators. The control room personnel stayed in respirators until 0315 in the morning at which time the alarm was cleared and they were removed. -4 hrs. Makeup tank 1 was vented to the waste gas vent header. At this point there '5 mins. should be an attempt to correlate with any offsite releases that might have 3435) occurred.

7 hrs.

Began pumping the auxiliary building sump to WOL-T-8B tank. This activity 5 min. occurred intermittantly over the next several days. 3715) 7 hrs. Reactor coolant letdown was shifted from the MU-T-1 to B reactor coolant drain 4. 5 min. tank. Because of an increase in pressure in the makeup tank, and a recognition 0716) that when the tank was vented, radiation levels in the aux building were increasing. 0

5 hrs. The reactor had been cooled down slowly to a cold leg temperature of 280 F with

0 mins, primary system pressure at 1050 lbs. Boron concentration at this time is entered in the log book at 1521 parts per mil' lion.

i

• It was later confirmed that significant flow blockage in the letdown coolers had' resulted from boron precipitation.

Increasing the cooling water temperature desolved some of the boron.

25 3 hrs. Makeup tank was vented to the waste gas vent header again to lower the makeup 2 min.. tank pressure to about 55 psig. 2030) l 1 hrs. Pressurizer level transmitter 2 failed high and then returned to normal. This 5 min. is the first indicated failure on the pressurizer level transmitters. This type 2115) of intennittent failure would continue for several weeks and eventually all three level transmitters would fail. It was believed at the time the transmitters were being severely affected by the radiation fields inside containment. These instruments had not been required to be qualified for any accident environments. 2 hrs. Shift personnel continued to monitor the OTSG/B level for any indications of 3 mins. leakage from the primary to the secondary side. There was no indication of 2200) any further leakage after the initial pa:t of the accident. Friday, March 30, 1979 4 hrs. NRC gave the licensee permission to resume releases of a slightly contaminated 5 min. industrial waste to the Susquehanna River. This action was coordinated with 0015) the Office of the Governor of Pennsylvania and a press release was issued by the State. 5 hrs. Vented makeup tanks to the waste gas decay tanks. Decay tank pressures at J min. this time were: Tank A - 50 lbs; Tank B - 75 lbs. 3150) 7 hrs. Once again the makeup tank was vented to the waste gas decay tank. 3 min. 3330) ' hrs. Again vented the makeup tank to the waste gas decay tank header. 3 min. 3710) .~ At this time there were only 2 core thennocouples reading temperatures above the saturation temperature for the primary system which at that time was 545 psig. At this time the licensee expressed his view that it was prudent to remain in the

• T.here was apparently a high concentration of noncondensible gases desolved in the primary coolant.. Continued letdown to the makeup tank was degassing most of these gases into the tank vapor space resulting in a frequent need to vent

the execss tank pressuie.

26 prastnt ecoling moda. Shifting to tha decay heat removal system for cooling could result in leakage of highly radioactive coolant from the decay heat re-boval system into the auxiliary building, the movement of noncondensible gases I into the reactor coolant loop and boiling in the core when the reactor coolant pump is shutdown. Preliminary assessment of the extent of fuel damage from a reactor coolant sample taken at approximately 5 p.m. on March 29. :adicated significant releases of iodine and noble gases from the fuel. ~ The staff was informed by the licensee on Friday morning that examination of containment pressure data for March 28 indicated a pressure spike up to approxi-mately 30 psi occurred at approximately 1:50 p.m. This was the first the NRC was informed of this event. 5 hrs. The Chairman of the NRC suggested to Governor Thornburgh of the Cannonwealth of 2 min. Pennsylvania that pregnant women and pre-school children in an area within 5 1130) miles of the plant site be evacuated. At this time there was a general concern about the increasing pressure in the waste gas decay tanks, by 1440 on the 3011. Tank A was at 84 psig and Tank B at 83 psig. It was determined at that time that i .it was recessary to vent these tanks into the reactor building and with some difficulty, this was finally accomplished and the tank pressures were reduced to 80 psig. In an attempt to identify the cuantity of non-condensible gases in the pressurizer vapor space, pressurizer heaters were secured and the pressurizer pressure rate of decrease was ok.erved. It is not clear that this method provided usable infonnation to the control room operations, however, at 2128 on the 30th the pressurizer vapor space was vented to the reactor coolant drain tank and ultimately to the containment by opening RCV 137. The pressurizer vos continued to be vented and at time 2353 the spray line was turned on and degas-sification of the primary system began. . hrs. The Director of the Office of 'NucleiF~ Reactor Regulation, the Director of the ~ 2 min. Region 1 Office of Inspection and Enforcement and the Director of the Division 1400) of Operating Reactors arrived at the site to direct the activities at the site and site vicinity. ] hrs. The licensee released less than 50,000 gallons of slightly contaminated industrial i 3 min. wastes. This release was terminated at NRC request because of concerns expressed '500) by State representatives. 9 _y-- _ _, _, -., -. - - - - - -, - - -, - _ _,,,,,,,. ~-... -., -

^ ---~ W)27 Saturday, March 31, 1979 khe hydrogen recombiner is in an operable status, however, shielding of the piping and complements was not fully installed at this time. Calculations l of hydrogen in containment showed that the present concentration was less than 4%.

~ hrs.

Secured pressurizer degassification by closing pressurizer vent valve (to 3 min. Reactor Coolant Drum Tank) (RCV 137) this' was done to wait for the hydrogen

315) recombiner to be placed in service.

It is not clear when this form of degassing the primary coolant was initiated. o3 hrs. Began spraying and degassing in the pressurizer again and venting to the con-5 min. tainment building. 2755) l 5 hrs. Plant status - containment fan coolers,4 out of 5 were operating; Steam generator ? min. A being used for cooldown,B isolated due to a potential leak from primary system; ,?S00) Containment pressure negative 1 psig; Containment sump level 2 feet; Borated water storage tank lavel - 18 feet; Reactor coolant loop A inlet temperature 280 F; j Reactor coolant pressure 120 psig; Pressurizer level 230 inches; Pressurizer temperature 550*F; Boiling water storage tank level is 18.5 feet. There continued to be considerabia concern over the volume of non-condensible gases in the reactor vessel. At 2:40 p.m., the licensee calculated the size of the bubble at 880 cubic feet at 875 psig. He later recalculated this value and arrived at a value of 621 cubic feet at the same pressure. Evaluations and analyses of this'co'ncern continued. Sunday, April 1,1979 0 5 hrs. Plant status - primary temperature 28J F; Pressurizer pressure 1028 psig; Pressurizer min. level 193 inches; Steam generator pressures A 26 psig, B 31 psig; Containment pressure

130) negative 1.1 psig, average temperature 90 F; l'atest containment air sample hydrogen 2.4 percent, oxygen 19.1 percent, nitrogen 79.5 percent; Highest reading thermo-0 couple at 0850 in the morning was D-10 reading 497.

2 hrs.-It was generally agreed between the NRC senior staff onsite and GPU representatives 3 min.

that the volume of the noncondensible bubble in the reactor vessel was 400 + 200 '300) cubic feet at 875 psia. It appeared that it was shrinking at a rate'of 34 cubic feet per hour. Based on analyses performed by B&W it was estimated that the bubble would be gone by approximately noon on April 2 if not sooner. _ _ _ _.,}}