Text

Amend 13 to Restart Rept
	ML19290E013
Person / Time
Site:	Three Mile Island
Issue date:	03/03/1980
From:	; METROPOLITAN EDISON CO.
To:	;
Shared Package
ML19290E008	List: ML19290E007; ML19290E013;
References
	NUDOCS 8003040007
	Download: ML19290E013 (136)
	v • d • e

_ TABLE OF CONTENTS - Continued Paga 9.0 DRAWINGS 9- 1 10.( . dS REFERENCE TO ORDER RECCMMENDATIONS 10-1 10.1 Introduction 10-1 10.2 Short-Term Recommendations cnd Met-Ed Responses 10-1 10.3 Specific Responses to Recommendations 10-4 10.3.1 Response to IES79-05A, Item 2 10-4 10.3.2 Performance Testing for PWR Relief and Safety valves 10-5 11.0 TECHNICAL SPECIFICATIONS 11-1 11.1 Introduction 11-1 11.2 Technical Specification Cnanges 11-1 11.2.1 Reactor Trip on Loss of Feedwater or Turbine Trip 11-1 11.2.2 Position Indication of PORV and Safety Valves, Setpoints 11-2 11.2.3 Emergency Power Supply Requj.rements -

Pressurizer Heaters 11-S 11.2.4 Post - LOCA Hydrogen Recombiner System 11-7 11.2.5 Containment Isolation Modifications 11-8 11.2.6 Instrumentation to Detect Inadeq ua te Core Cooling 11-11 11.2.7 Emergency Feedwater System Modifications 11-13 11.2.8 Post Accident Monitoring 11-16 11.2.9 Reactor Coolant Pump Trip on Coincident ESFAS and Coolant Voiding 11-18 11.2.10 TMI-1/TMI-2 Separation 11-20 vii Am. 13 E D d 36 +ooO 7

1 TA2LE OF CONTENTS - Co. inued Page 11.2.11 Lou Reactor Coolant System Pressure l Channel for HPI/LPI Initiation 11 ^1 11.2.12 Raising the Low Reactor Coolant System Pressure Trip Setpoint 11-22 viii Am. 13

To provide further assurance that emergetyy feedwater can be delivered when requirsd, tha failure mode of control valves EF-V30 A/B is being enanged. Currently these valves fail half open on loss of electrical centrol signal and fail 'as-is" on loss of instrument air. The change consists of modification to the operator such that on loss of air, the valves will fail in the open position and remain in this position.

Control valves EF-V30 A/B ere controlled by the Integrated Control System. The design of this sytem is described in chapter 7 of the IMI-l FSAR. Upon loss of all reactor coolant pumps, and/or bcch feedwater pumps, the IC positions the control valves to maintain steam generator wate- level. If reactor coolant pumps are available, the ICS controls are set to maintain a 30 inch water level on the start-up range level indicator. If reactor ccalant pumps are not available, the ICS maintains steam gener-ater water level at 50% on tha operating range level indicator.

The Integrated Control System is a control grade system. It does, however, receive power from the Class lE power system. Specific-ally the ICS is supplied from Distribution Panal ATA. This panel can be powered from the station batteries thru inverter lA and Panel VBA or from ES Control Center lA through Panel TRA.

Manual Control of the emergency feedwater control valves can be taken from the control room. When manual control is selected all active components of the ICS are bypassed except for the raise /

lower voltage circuit. As further assurance that control of the emergency feedwater control valves are available to the operator, an additional nanual control station is being provided for each valve. The -controls will be located in the control room and will be totally separate from the ICS. Power from the redundant por-tion of Class IE power system will be provided to the back-up controls. A functional diagram of the new manual controls is shown in Figure 2.1-3. A new manual loader station for each control valve will be mounted on the control room console. This will allow the operator to manually set a +10 volt control signal into the voltage / pneumatic converter in order to control the position of the EFW control valve. An adjacent selector switch connects the signal from the manual loader station to the voltage /

pneumatic converter and also replaces the ICS "EL" power supply with an independent 115 volt, 60 hz supply. Thus , if the EFW controls are disabled due to a failure in the ICS or failure of the "EL" power supply, the operator will have the ability to control flow to either steam generator entirely independent of the ICS.

Each of the emergency feedwater supply lines has also been pro-vided with two redundant loops of flow monitoring devices. Each loop consists of flow transducer, flow display computer (i.e.,

flow transmitter) and control room mounted indicator. The trans-ducers and computers are sonic flow devices as manufactured by Contrelotron. The transducers are installed downstream of the control valves before the emergency feedwater lines enter the 2.1-22 Am. 13

containment building and the computers are installed in their respective (Train A or B,i diesel generator building conttol panel area or another area environmentally suitable for the computers..

The transducers are suitaole for a steam line break tempera ture environ =ent. They are qualified for a continuous operation in 250*F ambient or prccess conditions following half an hour ex-posure to 350*F ambient or process. The computers are suitable for -20 to 110*F ambient temperature. The special coaxial cable between each set of transducers and associated computer will be suitable for class IE application and meet applicable require-ments of IEEE-383-1974 Additional cable between each computer and associated indicator plus the locp power supply cable shall also be suitable for class lE application. The flow devices have been seismically qualified. The output of each flow computer will transmit the signal to the main control room where the associated indicator will be installed to read flow directly with an overall loop accuracy of better than or equal to 5%. The in-dicators to be installed will have seismic qualification. Cabl-ing will be routed as described in Section 7 of the TMI-l FSAR.

The power supply for the instruments will be derived from the vital 120 V power system. Redunden Power supplies will be used for redundant instruments.

A diverse means of monitoring emergency feedwater flow is provid-ed by the steam generator level indicators. These measuremsnts are derived from Bailey type "BY" transmitters which, subsequent to their installation at TMI-1, have been seismically qualifiec and qualified for operation in a post LOCA containment enviren-ment. Cne start-up range and one operating range transmitter have been raised higher above the reactor building floor to avoid flooding in a post-accident situation and have had their elec-trical connections protected to prevent degradation due to moisture. The level instruments are supplied from lE on-site power sources and their wiring is run in raceways which have been analyted to assume heat. They wil] withstand a seismic event.

2.1.1.7.4 System Operation The TMI-l Auxiliary Feedwater System is a stand-by plant system which is not used during nor=al plant start-ups, shutdowns or ope ration. The system is maintained in stand-by during plant operations and is automatically actuated upon loss of both main feedwater pumps or loss of all four RC pumps. The following table gives actuation time for the system:

Event Turbine-Driven Motor-Driven a) Loss of Feedwater or Immedia te 5 Sec.

Loss of RC Pumps b) Above with loss of Immediate 15 Sec.

off-site power (LOP) c) Above with ESAS but Immediate 20 See no LOP d) Above with ESAS and LOP Immediate 30 See Start up and test data indicates that the turbine driven pump requires 18 seconds to reach full flow. The motor-driven pumps should be capable of accelerating to full speed in less than 10 seconds. Therefore under worst case conditions emergency feed-water flow should be established within approximately 40 seconds.

2.1-23 Am. 13

Control of auxiliary feedwater flow following initiation is accom-plished by the ICS. The ICS controls the injection of auxiliary feedwater to maintain water level in eacn steam generator to one cf two setpoints depending on whLther RC pu=ps are or are not available. Under forced cooling ccaditions, the ICS controls level to 30 inches on the start-tp range since this is sufficient to provide core cooling. However upon loss of forced RCS cooling the ICS controls steam generator level to 50% on the operating range to promote natural circulation with the Reactor Coolant System.

Manual controls in the control room are available for the opera-tor to take control of the EFW flow to either steam generator wnen needed or in the event of ICS failure.

2.1.1.7.5 Design Evaluations Table 3-1 (supplement I to this RESTART REPORT) indicates that the heaviest loading on one diesel generator during an ESAS actuation would be 2913 Kw and during a loss of offsite power only, the load would be 2817 Kw. The total load in either case is below the 2000 hour0.0231 days 0.556 hours 0.00331 weeks 7.61e-4 months rating of 3000 Kw. Since no credit has been taken for the reduction in pumping requirements following a LOCA and since the diesels 2000 hr rating is not exceeded, the diesel operability will not be affected. A detailed loading study has also verified this f act and testing will be performed to further verify this fact.

2.1.1.7.6 Safety Evaluation (See also responses to Questions 8 and 10 of Supplement 1, Part 1)

Safety analyses performed on the 177 Fuel Assembly B&W plants have determined that the emergency feedwater systems for a 2772 Mw plant must be capable of delivering 550 gpm (total to both generators). The basis for this criteria is contained in Volume 1; Section 6 - Supplement 3 of B&W's report entitled, " Evaluation of Transient Behavior and Small Reactor Coolant System Breaks in the 177 Fuel Assembly Plant". The analysis submitted by B&W is applicable to TMI-1. Several studies have also been performed by 3&W for the 177 FA plants on loss of main feedwater transients.

These analysis have demonstrated that 500 gpm or lower auxiliary feedwater flow is adequate following upset transients such as loss of power and the loss of normal feedwater flow. (SeeSupple-ment 1, Part 2responsetoquestion3.)Therefore,thesmallbreak LOCA conditions with a 20 minute delay in auxiliary feedwater ini-tiation sets the minimum emergency feedwater capacity requirements.

Considering that TM1-1 is only a 2535 Mw plant, a minimum emer-gency feedwater capacity requirement of 500 gpm is equivalent to 550 gpm for a 2772 Mwt plant.

As discussed in paragraph 2.1.1.7.3 above, the TMI-1 emergency feedwater system is comprised of two 460 gpm capacity electric pumps and one 920 gpm capacity steam driven (turbine) pump. The addition of the motor driven pumps (automatically) to the diesel block loading sequence and the turbine-driven pump start circuit ensures that a single failure will not result in less than the minimum required pump capacity being available under all condi-tions including loss-of-off site power. That is at least two motor driven or one motor driven and the turbine pump will be available under all single failure conditions.

2.1-24 Am. 13

Regulatory Guide 1.97 Rev. 2 (Dec. 1979) shall be utilized in the design guidance of high range ef fluent monitor. Vital bus power shall be employed for each system's modular assembly with the normal power supplying the monitor pumps with diesel gener-ations as back ups. Further descriptions of increased range capabilities are provided in Section 2.1.2.

High Range Effluent Radio Iodine & Particulate Sampling Analysis -

The existing sampling system uill oe expanded and will include the addition of silver zeolite cartridges. The system design and operation will both decrease the activity on the cartridges so they can be handled and will decrease the xenon to iodine ratio. Counting of tne cartridges will be by use of Na1 crystal connected to a single or dual channel analyzer with appropriate window and discrimination settings for th 364 Kev gamma cf I-131, or by use of a CEL1/MCA syste=. The expanded r srtion of the sampling system would be placed in service follow-ing an accident anc will be located in an applicable area exhibit-ing low background. The system will be on site and operable by 1 January, 1981.

Prior to incorporation of the expanded sampling system, procedures will be developed for the use of silver zeolite cartridges and normal particulated filters for sampling with a NaI detector and a single or dual channel analyzer for iodine and gross particulate release rate determination. Specific details to insure exposures are maintained as low as reasonably rchievable will be incorporated into the procedures.

These procedures will be available for NRC review prior to restart or 1 October 1980 whichever occurs first.

2.1.2.2 RCS Venting 2.1.2.2.1 System Description Power operated vents will be provided for the reactor coolant system in order to ensure that natural circulation and adequate core cooling can be maintained following an accident. The vents will be from the top of the pressurizer and the top of both hot legs using existing connections on the reactor coolant piping.

The discharge from the hot leg vents will be directed to the con-tainrent atmosphere. The system is shown schematically in Figure 2.1.-11.

The vents to the containment atmosphere will tie into existing hot leg vent piping outside the secondary shield wall. As part of this modification, remote operation of the vent valves in the existing vent line from the pressurizer to the reactor coolant drain tank will be provided and the system will retain the exist-ing venting capability. Control and position indication for the power operated vent valves will be provided in the control room.

2.1-31 Am. 13

Pending the availability of the required safety grade equipment to accomplish this modification. implementation can be completed by January 1, 1981.

Instrumentation will be provided for determining when hot leg venting is required and for determining when the venting is c omp le t e. The details of this instrumentation will be provided later.

2.1.2.2.2 Design Basis Small break loss of coelant accidents (LOCA's) can lead to RCS depressurization where steam and/or non-condensible gases may accu =ulate in the reactor vessel head, the upper portion of the hot legs and in the pressurizer. Following repressurization of the RC5 by high pressure injection (HPI), the steam bubbles collapse and remotely controlled vents on the upper hot legs and pressurizer can be used to vent non-condensible gases to pro-mote water solid na: ural circulation for core cooling.

The function of the reactor coolant venting system is to permit venting, from a remote location, of gasss trapped at high points in the reactor coolant system and pressurizer when post-accident radiation and contamination levels will not pe rmi t access to systems inside the containment.

The hydrogen generation design basis for the system will be loss of coolant accident (LOCA) with hydrogen generation rates cal-culated in accordance with NRC Regulatory Guide 1.7.

The system will be capable of venting a volume of non-condensible gas equivalent to one-half of the reactor coolant system volume in one hour. The system in performing its design function will not degrade nor defeat any features of the existing reactor coolant system. The remote operated vent valves will be solenoid operated (except for the normal " degas" valve from the pressurizer to the RCDT which is motor operated). The power supplies and any instrumentation for the vent valve operators will be Class IE and from on-site power sources.

The pipe size selected for the vents will be such as to preclude challenges to the high pressure injection function of the Make-Up and Purification system. The vent lines will be sized so that an inadvertent opening of a pair of vent valves in a single vent line will not result in out-flow greater than the make-up capacity of a Make-up Pump. On this basis a LOCA analysis will not be required.

2.1.2.2.3 System Design The venting system will be designed to assure reactor coolant system integrity and the capability to vent the RCS to the containment atmosphere following an accident.

2.1-32 Am. 13

The key elements of system design are as follows:

a) Piping will be designed in accordance with ANSI (USAS) B31.7

" Code for Pressure Piping - Nuclear Power Piping."

Piping from the reactor coolant system hot legs and pressur-i:er to the power operated vent valves will be class N Piping downstream of the vent vsives will oe class N-2. All vent valves will be class N-1.

b) The solenoid vent valves, their operators, and the vent piping will be seismically designed and analyzed in accordance with the requiremen:s of Seismic Class I.

c) Vent piping and valving will be designed and si:ed such that the failure to completely close off any one of the vent paths will not cause a loss of reactor coolant at a rate in excess of the normal capability of the makeup system at full RCS design pressure.

d) The effluent flow from the hot leg vent points will be routed directly to the containment atmosphere. The region into which the discharge is diverted will be selected to enhance mixing and dilution so as to minimize the potential for regions within the reactor building reaching flammable hydrogen gas concentra-tions. The design will take advantage of existing ventilation and heat removal systems for mixing and dilution. Discharges will be routed and directed so that the effluent will not ad-versely affect any structures, systems or components important to safety, e) Vent piping and valving will be designed to the same conditions as the reactor coolant system. Pipe and valve materials will be compatible with all anticipated fluids. These include water, saturated steam, steam water mixture, superheated steam, fission product gases, helium, nitrogen, boric acid solution and hydrogen.

f) Spark free solenoid operated valves will be employed for vent-ing. The valve operators will be qualified for normal and post-accident reactor building conditions.

g) The system will be capable of venting a volume of non-condensible gas equivalent to one-half of the reactor coolant syste= volume in one hour.

h) The hydrogen generation design basis for the system will be a loss of coolant accident (LOCA) with hydrogen generation rates calculated in accordance with NRC Regulatory Guide 1.7.

i) The system will retain local manual vent capability to the Reactor Coolant Drain Tank (for normal system operation venting.)

2.1-33 Am. 13

j) The remote operated vent vciver will be soleacid operated (except for the normal pressurizer "cegas" valve which is motor cperated). The solenoid valves will be energized to open and fail closed on less of power.

k) There will be two remote actuated valves in series in each vent line to provide redundancy 60 that c single active failure of a valve to close will not degrade the RCS integrity.

Each valve will have its own control switch and position indicating lights.

1) All valves for any one hot leg vent nozzle will be powered fro = a safety grade supply independent of that which powers the valves for any other hot leg vent nozzle so that any single power supply f ailure cannot cause a failure to vent at least one hot leg vent nozzle.

m) Each venting point will be individually operable, independent of any other vent point.

n) Control of vent valves will be remote manual from the control room. Direct indication of actual valve positions will be provided in the control room.

o) Both vent valves at a vent point will be powered by the same power source, but controlled by independent switches. An alarm in the main control room will indicate when the valves are energized.

p) The piping and valving for the venting system will be routed, oriented and protected to preclude damage from pipe whip, jet impringement and missiles caused by small LOCA's and steam /

feed ruptures.

q) Pipe routing, orientation and evaluation will assure that all remotely operable valves are located well above the maximum anticipated level of water in the containment following an accident. Each solenoid operated vent will be designed to remain functional after all design basis events except large LOCA's, evacuation of the Main Control Room, and loss of actuation power.

r) Existing system.

nozzles in the RC System will be used for the venting capability.

No new nozzles will be added to achieve the venting 2.1.2.2.4 System Operation This system will be used following a loss of coolant accident that generates volumes of non-condensable gas which may become trapped in the pressurizer and in the hot leg high point piping in suffi-cient volume to jeopardize water solid natural circulation. The system is operated manually from the main control room. Tw' spring loaded (to close) key-locked operating switches must be manually held in the open position, s imu l taneous ly , in order to establish a vent path to the containment atmosphere from any one hot leg vent point.

When the key-locked operating switches are released 2.1-34 gn, 13

or when power to che vent valve operators is cut off, the valves close. Specific operating procedures will prevent the operator from opening more than one vent path at a time. Failure to closs off a vent path will result in a loss of coolant rate which will be within the normal make up capability of a Make-up Pump.

2.1.2.2.5 Design Evaluation The RCS high point venting modification will utilize, in part.

existing hign point vent piping in the RCS. Existing high point vents are used on a routine basis to vent the RCS during normal preoperational filling procedures. This modification will make possible the remote venting of gases, from high points in the RCS, to the containment atmosphere following an accident.

Consideration has been given to the addition of a vessel head vent. Venting of the vessel head is not considered necessary or desirable for the following reasons. Firs t, the vents on the pressurizer and top of the hot leg piping provide the capability necessary to degas the RCS and to ensure that natural circula-tion flow is not interrupted by non-condensible gases. When RC pumps are operable, non-condensible gases are removed by de-gasing the RCS through pressurizer venting with pressuriser hecters and spray operting. Moreover with RC pumps available, non-condensibles will not prevent core cooling. If RC pumps are not available, non condensible gases could interrupt flow if gases collected at the top of the hot leg. Because of the RCS geometry gases would migrate to the top of the hot leg and not into the core. Therefore any desirable venting can be done from this location.

With no RC pumps operating, gas could collect in the vessel head region. This gas , however, does not interfere with natural circu-lation unless the quantity becomes sufficiently large to expand down to the elevation of the outlet nozzles. In this case the ex-cess gas would than rise to the hot leg high point and could be vented at that location. In addition, it should be noted that a temporary interruption of natural circulation is not a problem, since B&W analysis has shown that the core will continue to be adequately cooled as long as it remains covered.

Besides not being necessary, there are other disadvantages to the addition of a head vent. First there will be fewer open-ings in the RCS system that could be inadvertently opened.

Second, the available head vent location is on the top closure of the control rod drive mechanism. Babcock & Wilcox has in-dicated that venting of the CRDM closure may have some ad-verse effects on the safe and reliable operation of the drives.

2.1.2.2.6 Safety Evaluation The RCS remote operated system will be designed as a pipe Class N-1, seismic category S-1 system with safety grade power supply.

2.1-35 Am. 13

RCS Integrity will be maintained by two (2) valves mounted in series in each vent line.

Based on hydrogen generation rates calculated in accordance with Requiatory Guide 1.7, no further analysis of hydrogen release to the containment atmosphere as result of operating this system is required, since TMI-1 is being modified to install a hydrogen recombiner system capable of handling hydrogen volumes in accordance with Regulatory Guide 1.7.

2.1-35(a) Am. 13

E The power supply to the respective hot leg vent valves will come fron different sour:es so as to provide a degree of redundancy of power to the different hot leg vent valves.

In orcer to prevent pos sible loss of :oolant accidents, that '

could result from inadvertent va lve actuation. the power supply to the vent valves will be normally de-energized. Also, adminis-trative controls, including key locked operating switches in the main control room, will be provided to preclude inadvertent operation.

Since the only portions of the system to penetrate the containment will be electric and ins trumentation systems , there is no potential for a direct release to the outside environment as a result of the operation of this system.

2.1.2.2.7 Startup Testing and Inservice Testing / Inspection Requirements Besides the normal inservice inspection requirements imposed on design by ASHI B&PV Code Sect. XI for system design and inspection, the following surveillance will be conducted on the solenoid actuated vent v:ives:

The vent valves will be exercised during each refueling outage ,

in accordance with existing TMI-l requirements for reactor cooling l system venting.

Also, provisions will be made for testing all portions of the venting system during the TMI-l restart startup and test program.

Testing will consist of the following:

a) Flow indication to show that flow is present. Such testing will likely be done during initial fill, b) Confirmation of vent shutoff capability will be established during pre-service hydrostatic testing.

During plant operatica flow indications which have been previously tested will be monitored to assure that gross inadvertent venting is not occurring during normal reactor operation.

2.1.2.2.8 Instrumentation The remote actuation station for the solenoid vent valves (located in the main control room) will be controlled by means of key-oper-ated switches. The switches will be two position, spring return to the "CLOSE" position. The actuation station will be equipped with the following for each solenoid valve:

a) Key-locked Operating switch b) Position indicating lights (open-red / closed green) 2.1-36 Am. 13

Valve position shall be derived from stem operated limit switches or comparable means.

Annunciation (visaal and audio), for the not closed position, will be provided in the main centrol room. '

Control of valves for any one vent point will be independent of l the control for salves for any other "ent point.

Instrumentation will be provided to allow the operator to deter-mine when venting is reauired. Also, ins trumentation will be provided to indicate the presence of flow in the vent lines.

2.1.2.3 Plant Shielding Feview 2.1.2.3.1 Design Review of Plant Shielding The Section 2.1.6b in NUREG-0578, "TMI-2 Lessons Learned Task Force Status Report and Shcrt-Term Recommendations", requires Met-Ed to perform a radiation and shielding design review of the spaces around systems that may, as a result of an accident, con-tain highly radicactive materials and also to identify locations of vital areas and equipments in which personnel occupancy may be unduly limited or safety equipment may be unduly degraded by the radiation fields during post-accident operations of these systems.

A report, TDR-121, " Plant Shielding Design Review Report", has been prepared and will be submitted separately to provide greater detail concerning the above requirement. The content of the report is based on the following guidesines:

a) The post-accident release of radioactivity in the evalua-tion should be equivalent to the source terms recommended in Regulatory Guide 1.4 and NUREG-0578.

b) The post-accident dose rates in areas requiring continuous occupancy should not exceed 15 mn/ hour, c) The post-accident dose rate in areas which do not require continuous occupancy should be such that the dose to an individual during a required access period is less than 5 Rem whole body or its equivalent.

The report has utilized the well-accepted cyllindrical source methodology for the calcu!ation of dose rates at different vital aras. The doses te be received during ingress and egress and while performing a given operation in those areas have been determined by multiplying the dose rate by the travel time and the time assumed to perform the operation respectively.

As a result of the review, several areas requiring post-accident access have been identified as areas that would exceed the guidelines mentioned above. The report addressed mitigating measures to be taken for the problem areas so that the functions can be performed in those areas without exceeding the safety gu idelines . The Table 2.1-6 depicts the required manual actions, 2.1-37 Am. 13

areas that have been considered, radiation doses assuming no corrective action, and the corrective action under considera-tion. The report does not, howaver, identify the equipment requiring improved environmental qualification, which wculd ,

be delineated in a separate report.

2.1.2.3.2 Design Basis The source term to be used for shielding calculations shall be as follows:

Licuid Systems:

Noble Gas - 100% of core inventory Halogens - 50% of core inventory Others - 1% of core inventory Containment Air:

Noble Gas - 100* of core inventory Halogens - 25% of core inventory The criteria for limiting general area radiation levels in order to assure personnel access to vital equipment will be as follows:

Arens requiring continuous occupancy - <l5 mr/hr Control Room Operation Support Center (TMI-l Health Physics)

Tecnnical Support Center (Mod / comp room and cooldown from outside control room panel)

Areas requiring possible frequent access - <100 mrem /hr (Once or more per each 8 hour9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months shif t)

Radicchemistry Laboratory H2 Recombiner Control Panel Liquid Waste Disposal Panel For all other areas, shielding will be provided as required to keep personnel exposures less than 10CFR20 and to maintain the integrated dose to vital equipment below that for which the equipment has been qualified. The integrated does to vital components and equipment will be determined using the calculated radiation levels and the required length of service of each component and piece of equipment post accident. ,[

2.1.2.4 Post Accident Samoling Capability 2.1.2.4.1 System Description Post accident analysis of reactor coolant samples and the c.?r-tainment atmosphere is recognized as a means to better define core damage and anticipate the need for remedial actions. The TMI-l capabilities for post accident sampling will be modified as necessary to provide key sample results on an on-line basis and to provide backup confirmatory sampling capability within 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months of directing that a sample be taken.

2.1-38 Am. 13 '

The key parameters to be monitored with on-line ins trumentation incluce containment hydrogen concentration, reactor coolant boron concentration and letdown failed fuel monitors. The on-line hydrogaa monitcring capability has been previously described in this restart report. An on-line boronmeter will be installed.

The conceptual design and schedule for installation will be forwarded to the NRC by March 1,1980. The existing reactor coolant system letdown monitors will remain on scale with up to 10% failed fuel based on the FSAR definition of failed fuel.

This existing monitor is deemed adequate as an indicator thac significant core damage has occurred.

A design and operational review of the reactor coolant and containment atmosphere sampling systems shall be performed.

Modifications shall be completed as necessary to ensure that personnel can obtain samples under accident conditions without incurring a radiation exposure to any individual in excess of 3 rems to the whole body and 13 3/4 rems to the extremities. The source terms to be considered shall be those previously listed under the Design Basis of 'lant Shielding. In addition, a design and operational review of :he radiological spectrum analysis facilities and the chemical analysis facilities will be conducted in order to identify any additional design features or shielding required to ensure that confirmatory samples can be obtained and analyzed within the 8 hcur period previously mentioned. The chemical analyses to be considered shall include both boron and chloride analysis. The results of the design and operational review and conceptual design for required modifications will be forwarded to the NRC by March 1, 1980.

RCS sampling can be accomplished within one hour and analy: sed in an additional hour under normal circumstances. Met-Ed's experience sampling under accident conditions at TMI-2 has taught us that slow deliberate steps are necessary to prevent personnel overexposures.

Since an early indication of significant fuel failure is obtained from the failed fuel monitor and an on line baronometer is being provided it is considered unnecessary and imprudent to attempt to draw a confirmatory sample in less than 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . All necessary and appropriate operator and emergency actions can be taken based on the early indications. In a :dition chloride analysis provides inform-ation which is only useful in the long term and therefore is not immediately necessary.

2.1.2.5 Reactor Coolant Pumo Trio on HPI 2.1.2.5.1 System Description The purpose of this proposed modification is to provide automatic trip of the Reactor Coolant Pumps when degraded primary system conditions associated with a LOCA have been detected. This will be accomplished by requiring that RCP trip be initiated when the Engineered Safeguards System has actuated Safety Injection and an increasing RC void fraction has been detected as indicated by low RC pump motor current. The proposed logic will preclude RC pump trip during those events such as severe overcooling or very small breaks where maintenance of forced cooling is very 2.1-39 Am. 13

desirable. The conceptual design described in this section is being submitted for NRC .ceview and cocment and will be implemented subject to concurrence of the NRC Staff.

2.1.2.5.2 Design Bases Analysis has shown that a certain range of small primary breaks may result in unacceptable clad temperatures if the R.C. Pump are tripped at a time when the R.C. System void fraction has achieved a high level. To prevent these detrimental consequences, the proposed control scheme will promptly trip the R.C. Pumps when R.C. system conditions indicate that a small break in this range may be in progress. (Until this modification is in place procedures will specify operator action to manually trip the RCP's upon actuation of Safety Injection). The system shall actuate when High Pressure Injection has bee.. initiated and the R.C. System void fraction has reached a nominal value which indicates that a high void fraction may develop. It is also very desirable, although not necessary, to avoid initiation during transients sucr. as overcooling and very small breaks where R.C. pump trip is not required so that forced R.C. System circulation can be =aintained. The proposed system will meet both of these criteria.

2.1.2.5.3 System Design The R.C. Pumps will be tripped on a coincident detection of High Pressure Injection by the ESFAS and low R.C. Pumps Motor Current in at least two of the four R.C. Pumps. This means that for the special case of two pumps operation, the pumps will be tripped on HPI alone. Reduncant sensors will be used for pump current on each R.C. Pump Motor. Redundant trip signals will be derived for each motor. Electrical separation will be provided for redundant signals. No surge failure in the proposed system shall prevent a trip when required. No single failure in the system shall result in the trip of more than one R.C. Pump. The actuation system will be designed to be operable after a seismic event. However, the R.C. Pu=p Motor Switchgear is not seismi-cally qualified. Provisions for on-line surveillance testing will be included. The operator will be able to restart a pump after trip by manual means.

2.1.2.5.4 Design Evaluation The proposed design concept assumes that pump motor current can be used in combinaion with HPI actuation to detect the need for an R.C. pump trip. Supporting evidence for this concept has been generated by two EPRI sponsored test programs.

In 1973, Babcock & Wilcox in conjunction with the Bingham-Willamette Company, conducted a test program to investigate the single and two phase performance of a one-third scale reactor coolant pump using air-water mixtures. The result of these tests were reported under a contract with EPRI in 1977 (Ref. 1).

Testing performed by CREARE under an EPRI contract using a 1/20 scale pump also shows a substantial decrease in torque at void fractions above 20%. Preliminary results of this testing were reported in the 6th Water Reactor Safety Research Information Meeting in November, 1978 (Ref. 2).

2.1-40 Am. 13

Since torque is directly related to pump motor current. the use of pump motor current as an indicator of the fluid void fraction, based on ene referen;ed experimental data, is appropriate.

The proposed design meets the dual requirements of reliability initiating a trip when required, and not degrading plant availa-bility through inadvertent trips.

2.1.2.5.5 Safety Evaluation The system will be functional during a seismic event (except for the R.C. Pump Motor Switchgear), will be testable and will meet single failure criteria for actuation. No single f ailure will result in trip of more than one pump. Redundant circuits will be separated. Where the new system interf aces within existing safety systems, care will be taken in the design to assure that there vill be no degradation of existing safety functions.

2.1.2.

5.6 REFERENCES

1. 1/3 Scale Air-Water Pump Program, Pump Performance Data, EPRI NP-160, Vol. 2, Oct. 1977.

2. EPRI/CREARE 1/20-SCALE TWO PHASE Pl}P PERFORMANCE RESULTS, P. W. Runs tadker, Jr. and W. L. Switt, CREARE Incorporated, Present at the 6th Water Reactor Safety Research Information Meeting, Nat. Bureau of Standards, Garbersburg, MD, Nov. 6-9, 1978.

r 2.1-41 Am. 13

2.1.2.6 Auxiliarv Feedwater Svstem Auto start of the emergency-feedwater (EFW) System is being implemented in two phases: 1. Control Grade Auto Start - This is a non-safety related initiation as described in paragraph 2.1.1.7 and it is a short-term approach, 2. Safety Grade Auto Ste : - Inis will be a long-term modification where the initia-clon will meet ene requirements for Class lE system and the system is functionally described below.

1. The saf ety grade EFW auto start when implerented will automatically initiate the system on presence of the following conditions with or without the availability of the of f-site power:

-* Loss of both normal feedwater pumps, or Loss of all four reactor coolant pumps, or Low differential pressure between the normal feedwater and main steam lines at each steam generator, The system initiation on low steam generator '.evel will even-tually be added. This will be done af ter the necessary analysis and engineering has been completed to insure that this signal will give a satisf actory actuation and vill not interact with other -lant functions. Loss of normal feedwater pumps is detect ud by differential pressure switches across each pump (two svitches per pump, i.e., one switch per train).

Th e model of diffe: 3ntial pressure switches used for this appln scion has been seismically tested. These switches have te=peratcre limits of -60 to 200*F. Since they will be loca-ted in Tu bine Building which is a non seismic building, the switches will be tied into their respective EFW initiating circuits (Train A&B) through buffer devices and thus the switches will be treated as safety grade items to the extent possible.

2. All cables associated with the initiating logic will be quali-fied for Class lE application and the initiations will be designed to meet single failure criteria. All circuits will meet the regulatory criteria for separation of Class lE cir-cuits.

3. The initiating logic will include hardware for the following purposes:

Latching mechanism to seal-in the actuation Manual Reset Capability Testability of the initiating circuit

4. Indication will be provided in the control room to identify the source of the initiation.

5. Annunciation will be provided in the control room to alarm:

Auto start of the EFW system. This will be a common alarm for both the trains.

2.1-42 Am. 13

2.1.2.7 Increased Range of Radiation Moritors (2.1.8.b) 2.1.2.7.1 The existing Radiation Monitoring System provides in-line monitoring capability for effluents from:

a) Auxiliary and Fuel Handling Building (RM-AS) b) Reactor Building Purge (RM-A9) c) Condenser Off-Gas (RM-AS)

Discharge from k'aste Gas Decay Tanks are monitored by RM-A7 prior to combination with other exhaust and after dilution by RM-A8. The Reactor Building Hydrogen Purge System discharge is monitored by the normal purge system monitor RM-A9.

The monitors, RM-AS and RM-A9 are manufactured by Victoreen, Inc. and consist of:

a) A fixed filter particulate monitor; Beta scintillation detector; sensitivity approximately 1.5 x 10 10 com/ min Ci/n 6

based on SR-90; full range 1 x 10 cp,,

b) A Fixed Charcoal Filter Iodine Monitor; NaI detector with fixed window; Sensitivity approximately 1.3 x 109*{

6 fill range 1 x 10 cpm, c) A gross gaseous monitor; Bata Scintillation detector; Sensitivity approximately 4 x 10 *" full range f

6 1 x 10 gm.

d) Air sampling pump with normal sample flow of approximately 1 cubic foot per minute-Radiation monitor RM-A5 has only a gross gaseous monitor (c above) situated on the discharge of the condenser vacuum pumps, exhausting to the suction of the vacuum pumps. Flow through the monitor is regulated to maintain approximately 500 cc/ min. All monitors have Control Room readout and recording.

2.1-43

/

Am. 13

2.1.2.7.2 Long Term Modificaticas Increased range capabilities will be furnished for eacc of the ef fluent monitors described above (RM-A8, RM-A9, RM-A5) and the Main Steam lines. For the Long Term Modification addi:icnal monitoring ranges will be provided utilizing ionization chambers for the Reactor Building Purge Exhaust, the Condenser 0FF-GAS Exhaust, the Main Steam Lines. The Auxiliary and Fuel Handling Building Exhaust will have extended monitoring ranges incorp-orating a G.M. device. The sensitivity of the individual units will be determined by standard vclume source calculations.

The sensitivity will assure that release rate of:

5,600,000 Ci/sec from Auxiliary & Fuel Handling Bulding.

2,300,000 Ci/sec from Reactor Building Purge.

1400 Ci/see from Condenser Of f-Gas based on maximum flow rates from each release path.

2500 Ci/sec from a single steam generator can be detected.

The installation of each menitor will include evaluation of the position of the monitor relative to other potential radiation sources and shielding necessary to minimize the effect of sources other than sample lines on the response of the monitor and recording.

For each of the monitors described, the following applies:

Each will be powered from vital power, thereby providing redundancy in power supply.

Establishing sensitivities will be correlated to solid source calibrations. Procedures defining calibration method and frequency will be written to assure proper response of the instruments.

Emergency procedures will be written to the use of the radiation instrumentation in conjunction with flow infor-eation to determine release rate.

Emergency Plan implementing procedures describe the dissemination of information obtained from monitors.

Procedures and evaluations will be available for NRC review prior to 1 January 1981.

2.1-44 Am. 13

TilREE MILE ISLAND UNIT NO. 1 -

TABLE 2.1-4 LEAKAGE REDUCTION PROGRAM TEST

SUMMARY

SYSTEMS TEST METil0D ACCEPTANCE CRITERIA

1. Makeup and Purification The Makeup and Purification System is leak checked 1 GPM in combination with the Reactor Coolant Syatem leak rate calculations per Surveillance Procedure #1303-1.1. Makeup and Purification System flow paths taken into consideration by SP 1303-1.1 are (a)

Makeup, (b) Letdown, (c) liigh Pressure Injection up to the MU-V16 valves, (d) R. C. Pump Seal Injectiou and Seal Return, (e) Makeup and Purification De-mineralizers and (f) attached portions of chemical addition and sampling. The allowable leakage for this system in combination with RCS cannot exceed 10 gpm and one gpm unidentified (Tech. Spec.

3.1.6). Surveillance Procedure 1303-1.1 utilizes a combination of the methods listed in the Tech.

Spec. 3.1.6 to detect, measure and identify system leakage.

2. Liquid Waste Disposal A new surveillance procedure SP 1303-11.28 has been Note 1 drafted to collect and measure leakage on the Reactor Coolant letdown piping from MU-V8 to the Reactor Cool-ant Bleed Tanks each refueling interval. The affec-ted flow path will he isolated and hydrostatically pressurized to a nominal pressure of 75 psig. While at pressure, the affected flow path will be inspected for evidence of leakage. Aay leakage detected will be measured, recorded and submitted to the station engineering staff for evaluation.

3. Decay IIeat Removal Leakage of this system is collected and measured by 6 GPil Surveillance Procedura #1303-11.16 and meets the requirements of Tech. Spec. 4.5.4.

4. Reactor Building Spray Surveillance Procedure #1300-3A/A/B la being revised Note 1 to include a 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months leak rate determination with the Building Spray pumps recirculating to the Borated Water Storage Tank. With the BS pumps operating, the flow path and unisolable piping up to the R.B.

Am. I 'l

TilREE MILE ISLAND UNIT No. 1 TABLE 2.1-4 (Cont'd.)

LEAKAGE REDUCTION PROGRAM TEST SUMiiARY SYSTEMS TEST METIl0D ACCEPTANCE CRITER_IA

4. Reactor Building Spray (Cont'd.) isolation valves will be inspected for evidence of leakage.

5. Waste Gas Disposal A new surveillance procedure SP 1303-11.29 has been Note i drafted to determine the leakage rate of the entire system using nitrogen pressurization in lieu of helium. The procedure is divided into two phanes.

Phase I checks the low pressure piping upstream of the Waste Gas Compressors at 2.0 psig. Phase II checks the Waste Gas Compressors, Waste Cas Tanks, and associated high pressure piping at a pressure of 65 psig.

6. Reactor Coolant Sampling A new surveillance procedure SP 1303-11.30 has been Note 1 drafted to determine leakage rate of the reactor coolant sample piping in conjunction with a Reactor Coolant sa.aple on a refueling interval f requency.

The leak check will be performed at a time when a primary sample is required while the plant is at normal operating temperature and pressure valves, piping and components between and inclusive of CA-V2. CA-V35 and CA-V16 will be pressurized to nor-mal operating pressure, and maintained at that pressure while being inspected for evidence of leakage.

7. Reactor Building Containment Containment leakage integrity is provided by the As per Ref. Pro-performance of surveillance procedure 1303-11.18, cedure Reactor Building, Local Leak Rate Testing; #1303-6.1, Reactor Building Integrated Lea; Rate Testing; 1303-11.24, Reactor Building 1.ocal Leakage Pene-tration Pressurization.

8. Containment Monitoring Leaka;a measurements are provided by the perfor- As per Ref. Pro-mance of surveillance procedure 1303-11.18, Re- cedure actor Building Local Leak Rate Testing on a re-fueling interval frequency.

Am. Il

TilREE HII.E ISLAND UNIT NO. 1 TABl.E 2.1-4 (Cont'd.)

LEAKAGE REDUCTION PROGRAM TEST SIRDIARY SYSTEMS TEST MET 110D ACCEPTANCE CRITERIA

9. Fluid Block Leakage measurements are provided by the perfor- As per Ref. Procedure

~

mance of surveillance procedure 1303-11.23, Re-actor Building Local Leakage Fluid Block.

10. Miscellaneous Waste Leakage measurement incorporated ith nitrogen Storage Tank test of the Waste Gas Disposal system of #5 above.

NOTE 1: During the init al performance of each new surveillance procedure, an effort will be made to identify and reduce any .eakage to as low as practicable. The nature and location of the leak, as well as the magnitude of the leakage, if any, will be evaluated as required. Allowable leakage rates will the n he established based on initial performance results following any repairs, if required. Procedurec will be revised to incorporate the newly established criteria. With the exception of Reactor Bu11 ding Spray and Reactor Cooling Sampling, the established criteria will be defined prior to restart of Unit No. 1. The allowable leakage criteria for Reactor Building Spray and Reactor Coolant Sampling systems will be established upon performance of these procedures during Restart Testing at ilot Shutdown.

Am. Il

TABLE 2.1-6 Major S~urce of Corrective Action Manual Action Location ose Radiat:on Dose Under Consideration

To read the waste 305' El. of the 2.3 rem Letdown relief line and Relocation and shielding of the gas tank pressure Aux. Building Aux. Building IIVAC Ex- 7.tdown relief line in the corri-To perfore waste haust Duct dor on El. 305' north of the rad-305' El. of the 4.7 rem waste control panel transfers Aux. Building To open d . cup and 305' El. of the 3300 rem for Makeup and purification Changing the valve HUV-198 to re-purificaHon sys- Aux. Building operation and system piping and the mote operation and adding bypass tem valve HUV-198 80 rem for waste gas tank vent to arround filter and .!cmineralizers to bypass the seal travel the containment and let-injection filter down relief line To reset any thrown 305' El. of the 3300 rem for Makeup and purification installing a shield wall to iso-circuit breakers Aux. Building operation and system piping and the late Motor Control Center IA from at Motor Control 80 rem for waste gas tank vent to the seal injection and high pres-Center lA travel the containment and let-- sure injection piping plus unknown down relief line, seal action to protect from gas line injection piping and high pressure injection piping To reset any thrown 305' El. of the 80 rem during Makeup and purification Installing a chield wall plus un-circuit breakers at Aux. Building occupancy and system piping and the !

known action to protect from gao Motor Control 64 rem for Center IB waste gas tank vent to line to isolate Motor Control Cen-travel the containment and let- ter IB from the seal injection and down relief line, seal high pressure injection piping injection piping and high pressure injection piping To operate manual 3600 rem 281' El. of the 11akeup and purification changing valves DilV-15A & B, Dil7-valves DilV-ISA & B, Aux. Building system and decay heat 19A & B, DilV-38A & B to remote g DilV-19A & B, and piping operation; revision of 1104-4 DilV-38A & B for procedure concerning post 1.0CA

[ Boron Precipitation Boron Control so that valves DilV-Control and for 15A & B and DilV-19A & B remuin continued operation locked open of the decay heat system

0ther sointion<; may ninn ha i d on r i f i n.1 a n.1 e a t ad -t in 14 ~. af '2~

7_

~

Major Source. of Corrective Action Manual Action Location Dose Radiation Dose Unpler Crzusidstign* ____ _

To unlock and open 281' El. of the 6500 rem Makeup and purification Changing valves DilV-12A & B and valves DHV-12A & B Aux. Building system and decay heat DilV-64 to remote operation and DilV-64 for Boron piping Precipitation Con-trol Control room, Tech- 355' El. of the Negligible ------

None nical Support Center Control Complex since dose and Operation Sup- rate is port Center Opera- <0.5 mr/hr tions Radiochemistry Lab 305' El. of the Negligible -

None Operations Control Complex since dose rate is

<0.5 mr/hr Nuclear Sample 305' El. of the Room Operations Control Complex Refer to Section 2.1.2.4 of the Restart Report operation of hydro- 305' El. of the Negligible ------

Recombiner control panel in the gen recombiners Intermediate since the Intermediate Building would be Building dose rate shielded f rom hydroge n recombiners is <10 mr/hr themselves so that etcessive doses are not received during operation of the recombiners Diesel Generator Negligible None

Rooms since the g dose rate is

~

<10 mr/hr u.

To reset thrown 281' El. of the 40 rem for Makeup and purification Adding a motor operated bypass to the Circuit breakers Aux. Building occupancy and system piping to and from makeup and purification sjstem fil-in Motor Control 220 rem for letdown line filters tors Center 1C travel

n e 6,,. ,, n i . . ra ,,,, ,, .. .,1 ,, ,, i ,, t . .,., e a r e ,.i .i---i . ee. e i

10 IIC DT TO 8tr,or Y(tJ EVJ) RC-VG ' T( HEW) IIC - V 47 a & LOCKED CLOSED LothE0*1LOSLD g$2 'W*-la i 5 ']t &

-SEEONDAIW SillELO WALL-*- --

- l- -

rS1

-- -l'" ItC- Vl3 A P

RC- Visn *J ?

LJ j ,(EXISTlHM 5

~

(E XIST lHG[-

/ I(t0CKED OPEM) (LOCKED OPEN)I.

~

RC-V4t A H C- V40 A ' fic-V 1013 fic v4lB

l . _ _

l A "

E E DPI -

L J d0 V 2 B (E K1581 "ph -

=~

(NEW ) RC V45-Lor.nto Optn (in W) l c vi4 (i] / '

~~

SECutJDAR( [i , [bP HIELD WALL 8 IIC'VIl ~

__ (DELE T E D) j ~ r7 es i RI.v's a ,rie'vff 10 K01 V TO REALT OR BLD6. BCW 18"4 ^ H i* E xiii.) [lC-J.i TO RE ACTOR Alt.10 S. 14 VICINITY IA WNI 10 f43 4t3 OtitLDitJ6 OF ll{V RETURt1 RCOT SUPPLY Ail.10 SPHERE DUCT (TYR) ,

PRESSURl2Elt s - ' '

b-c 1 DE F EllE NC E DV/G. [I I]

(.a Al OWG. 419 2-C-302- < >

P (o50 " R E AC.T OR COCL ANT SYST EM If u

r u

" 4 r g' iI Ik "

Fl_0W DI AGRAM" y m

K 3r FtGURE NO. "L .1- 11 TO Olla # #E ^

PUMP SULTION W ii NO. l .

R C. 5 VEN11NG SYS T E M

7.3.2 Flant Shieldini 7.J.2.1 General:

The purpose of this section is to identify the necessary action to be taken for providing additional shielding, procedures or plant modifications in vital areas and to equipment in which personnel occupancy may be unduly degraded by the large radiation fields in operating these systems in a post-accident environment.

Systems, components, and areas ceasidered subject to post-accident large radiation fields include: decay heat removal, reactor building spray recirculation, make up and let-down, vaste gas, Rad Waste Control Center, and Liquid and Gas Waste Panel.

7.3.2.2 Design Review:

In order to determine which areas and stems require additional shield-in or plant modifications, a review of existing design will be completed which will identify and recommend the cor-rective actions needed in vital areas throughout the Unit.

The review will encompass the accessibility and operability of the above stated systems and area in a post-accident environment.

Bases to be used in comparing present design to post accident environment. Bases to be used in cocparing present design to post-accident radiation levels will be the fission product release as described in Regulatory Guides 1.3 and 1.4 (see Section 2.1.2.3).

7.3.2.3 Near Tern Modifications:

Those near term codifications, identified by the design review, will be completed prior to startup of Unit 1.

7.3.2.4 Long Term Modifications:

Over the longer term modifications not completed before startup of Unit I will be completed by January 1,1981.

7.3.3 Auxiliarv Building Ventilation System 7.3.3.1 General The Unit-1 auxiliary Building Ventilation System is combined with the Fuel H&:.dling Building Ventilation System prior to filtration

~

7-12 Am. 13

d. After each complete or partial replacement of charcoal adsorber bank by verifying that the charcoal adsorbers remove > 99* of a halogensted hydrocarbon refrigerant test gas when they are tested in place in accordance with ANSI N510-1975.

7.3.3.3 Imolementation Schedule The testing schedule as previously described will be in place prior to the restart of Unit-1.

7.3.4 Nuclear Sacoling 7.3.5 Nuclear Sacoling Capabilities 7.3.5.1 Post-Accident Samoling 7.3.5.1.1 General Purpose of this section is to identify the necessary action to be taken to facilitate the capability to perform timely sampling of Reactor Coolant and Containment Atmosphere in a post-accident environment.

Action to provide the capability to perform timely (within 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months ) sar.pling of reactor coolant and containment atmosphere include:

a design review of existing systems and component location, near term and long term modifications.

7.3.5.1.2 Design Review:

To determine what modifications are required, a design review of the existing systems and components will be completed. This design review will identify and recommend the corrective actions necessary to provide the capability to sa=ple in a post-accident environment.

The design review will enco= pass the accessability and operability of the reactor coolant sampling and containment atmosphere sampling sampling systems in a post-accident environment. The design bases to be used for this review will be the fission product release in an accident as assumed in Regulatory Guides 1.3 and 1.4 (see Section 2.1.2.4).

7.3.5.1.3 Near Term Modifications:

Those near term modifications, identified by the design review, will be completed prior to startup of Unit I.

7.3.5.1.4 Lone Term Modifications:

Over the longer term modifications not completed before startup of Unit I will be completed by January 1, 1981.

e 7-14 Am. 13

4 k

i rA i w .

-% i

' { ' *

[1-I ' I te l' t

4:

s I }h 5 $!

u . 1 .

="a s a *i E -

- . , : e Wje ; L =

I b.5B u- e. wE 5 '

~.i.e

3 6C s "e

I s* c"l'" - > - '

" j- :=

3 _

A ;; . 'i e.

" 5 t .en e . : -

' .s'

..es- . -

5 r

4. - - - . - I A ,= :;Mj

- , i3- $ ': a Q s

, =1 g ; w i . g -

( ![

. ri iiiiiiii;i,l= =

/

3 x

,; 5 lS illiii'llil iliilli '

2

,- h D:K:T24*' N=%20=6aN6%'NM h i !, ]~

_ _I212l __ _ _12_l _!:l!I_ 2 'A

! ( _l2!: _ _ ' _' : '.j/

i I i-g 5

, = = = = = ,=,i : : e

$3' 5

~

l 2

e 2 l' ; .

y i

i m_

w ',

O

,,,,J 3'

- - - =

' , a e Z ':!

i :Y '@ 3 Q 4 I . f I g - ;(*! -

l

' O A

g

! si id* O "

. V~. !

2 l

sI:

i I I < 0 i

2 I

i l

,s a o g

nlIe&_r t i: l a.

l l' d i . .

U3 1

- ~ J _ _. - _ w j .

I

ss 1,,

n W, W ~ g-

\ h' ] 5 l ss _,,,

Fa/

r -

i I - sC N

3 5 T w <

l C g

/I. ,*- 5 3 lI t* h w W <

w$

~ ~

E / (*) 3 C

~

. , k h$ ~ ~

0 -!

'g@~ G , : h'TN M :\N e=\N- - .,\@*d a - ,

, o 1

.l i*".. t + +

m >

s o-o i . g Mi.' I e i(;, #

m

@@ =.

i<

-t e 8a v

3 Am. 13

Reconcendation Response Financial Qualifications To be Submitted Separately

8. TMI-2 Lessons Learned Recou-mendations - SUREG 0578 2.1.1 Section 2.1.1.3 and 51 P1 Qll

& 14 $1 P2 Q18 & 30.

2.1.2 See Section 10.3.2, also see S1 P1 Q16 and S1 P2 Q19 2.1.3.a Section 2.1.1.2, S1 P1 Q13 & 15 and S1 P2 Q23, 36 6 37 2.1.3.b Section 2.1.1.6, S1 P1 Q17, IS, 19, 20 & 39 and S1 P2 Q39, 92, 93, 94 5 95 2.1.4 Section 2.1.1.5, S1 P1 Q21-28 2.1.5 Section 2.1.1.4 2.1.6 Section 2.1.1.8 2.1.7 Section 2.1.1.7 & 2.1.2.6 2.1.8a Section 2.1.2.4 llh 2.1.8b Section 2.1.2.7.2 2.1.8c Section 7.3.5.3 2.1.9 Sections 3.1.1, 6.0, 8.1 2.2.1.a S1 P1 Q40 and $1 P2 Q25 2.2.1.b Section 5.0, S1 P1 Q42 and S1 P2 Q27 2.2.1.c Section 3.0, S1 P1 Q43, S1 P1 Q2d and Q29 2.2.2 Section 4.0, Section 10.3.3 and S1 P1 Q44 2.2.3 Not Applicable until NRC Regulations are revised 10-3 Am. 13

While there are many relief and safety valves in service there are only a few generic types which define the specifics of the test program.

Since existing f acilities preclude full-scale testing at this time, a two phase program is being undertaken by EPRI:

1. Existing test f acilities will be used for perf ormance testing of small safety / relief valves. Testing will occur under steam, water, and appropriate two phase conditions to ascertain whether safety / relief valves will open, close, and relieve sufficient fluid to protect the primary system pressure boundary.

2. In parallel with phase 1, f acilities that will allow testing of large safety / relief valves will be designed and constru:ted.

Present schedules call for scoping tests on relief valves which require the minimum in test facilities to be initiated during April, 1980 followed by safety valve tests, and generic safety /

relief valve system tests, to be completed in July, 1981. The expanded valve test facility will be in place by July, 1981.

Scheduling of test facilities and other uncertainties could result in a longer schedule. Met-Ed believes, however , that substantive test data can be obtained by July 1981.

Additional detail concerning the test program is contained in a letter from W. J. Cahill, Jr., to Harold Denton, Regarding:

" Program Plan for the Perfor=ance Verification of PWR Safety /

Relief Valves and Systems, December 13, 1979," forwarding letter updated.

10.3.3 Onsite Technical Suoport Center in response to the requirements of the Section 2.2.2.b, in NUREG-0578 as clarified by Mr. H. Denton's letter, dated October 30, 1979 to the licensee, Met-Ed has addressed clari-fications 13, 1D, IE, IF and 1G by providing the following information:

Response to the clarification IB: The requirement to provide plans and procedures for engineering / management and staffing of the Onsite Technical Support Center would be delineated in Emergency Plan Implementation Procedure (EPIP). EPIP 1004.31 (Activation of the Technical Support Center) is currently being written. Other EPIP's which will be involved in the TSC have been or are being written. The above requirement would be satisfied when the NRC cocpletes reviewing these EPIP's.

Response to the clarification 1D: Portable monitors have been installed in the TSC and the required action levels will be contained in the appropriate EPIP's.

,10-6 Am. 13

t Response to the clarification IE: Because of Fire Loading Limits in the area of the TSC. records and drawings which describe the ,

as-built conditions anc layout of structures, systems and co=per- I ents have been assembled and will be placed in the TSC when Fire Freef filing cabinets become available. Appropriate EPIP's would '

include this item.

Response to clarification IF: The EPIP's would also include the procedures for performing accident assessment function from the l control room should the TSC become inhabitable.

Response to clarification IG: As far as a longer range plan for upgrading the TSC is concerned, considerable engineering effort is required to complete it. Current plans call for devices to g read out plant computer information in the TSC. However, it has l not been determined if all the required information, including Radiation Monitoring data (new and existing) and meteorological data can be put into the computer. If it cannot be done, then additional equipment may be required in the TSC. The long range effort would be completed by January, 1981.

On page C3-5 of the NRC's Status Report On The Evaluation of Licensee's Compliante, dated January 11, 1980, the licensee is required to address all the provisions of the proposed rule on emergency planning and is expected to report on those matters; and on any further upgrading of the Emergency Plan, in a supplement to this evaluation prior to any restart of TMI-1. The proposed rule change is to upgrade 10CFR Part 50 Appendix E to incorporate requirements from existing documents and to require NRC concurrence with state and local plans prior to issuing operating licenses or as a condition for licensee's to maintain operating licenses. A review is cur-rently underway on this proposed rule change. It is antic-1 pated that the TMI-l Emergency Plan provided to the NRC will not require changes to be in compliance. State and local emergency plans are being upgraded at the present time. The state plan has been reviewed and corrections are now underway to being the state and county plans into compliance.

10-7 Am. 13

10.4 Transient Analysis and Procedures for Management of Small Breaks As part of the NRC Order of August 9, 1979, long term actions recom-mended by the Director of Nuclear Reactor Regulation were included.

Item two of that list recommended that the Licensee:

"... give continued attention to transient analysis and procedures for managecent of small breaks by a formal program set up to assure timely action of these matters;"

As part of the development of the TMI Generation Group, a new section was established as part of the Systems Engineering Department. This section, known as the Plant Analysis Section, is charged with responsibility for a continuing review of plant performance. This will include an on-going technical evaluation of the overall plant performance as well as the performance of major systems and components.

The Plant Analysis Section is also charged with the responsibilities of reviewing key information from other nuclear plants. This infor-mation will be obtained from the review of Licensee Event Reports, general industry survey information, industry contacts, regulatory body documents, owner's group activities, and standards committees.

Several of these sources of information will be provided to the Plant Analysis Section by other functional sections within the TMI Generation Group. The overall industry approach to the interchange of such information is currently under development and is expected to utilize organi:ations such as the Nuclear Safety Analysis Center and the Institute for Nuclear Power Operations. GPU fully intends to participate in these information interchange activities, when developed.

Recommendations resulting from these reviews may be used to modify equipment design, operating and maintenance methods, operator training programs, procedures, or other aspects of plant operation.

Additionally, two other sections with the Systems Engineering Department i provide more specialized analytical functions that, on a continuing '

basis, improve understanding of transient analysis and procedures for the management of small breaks.

The Control and Safety Analysis Section has the in-house capability to perform analysis of the type shown in Section 8 of this report.

They also contract for and review the results of such work performed by the NSS supplier or other consultants. The Control and Safety Analysis Section has been deeply involved in analysis of programs for small break LOCA operator guidelines and anticipated transient operator guidelines. This work is closely associated with and has had a direct input to operator training programs and procedures.

The Nuclear Analysis Section pe-forms or reviews fuel and core related aspects of transient and accident analysis, including reload evaluations. This work is interfaced with the activities in the Control and Safety Analysis Section.

10-8 Am. 13

Draft TMI-1 Technical Specification 3.4 provides Limiting Conditions for Operation for the emergency feedwater system. Guidance on oper-ability of the emergency feedwater System is contained in IE Bulletin 79-05A, Item 8, as follows:

" Prepare and implement immediately procedures which assure that two independent steam generator auxiliary feedwater flow paths, each with 100% flow capacity, are operable at any time when heat removal from the primary system is through the steam generators. When two independent 100% capacity flow paths are not available , the capacity shall be restored within 72 hours8.333333e-4 days 0.02 hours 1.190476e-4 weeks 2.7396e-5 months or the plant shall be placec in a cooling mode which d7es not rely on steam generators for cooling within the next 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months .

When at least one 100% capacity flow path is not available, the reactor shall be =ade subcritical within one hour and the facil-icy placed in a chutdown cooling mode which does not rely on steam generators for cooling within 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months or at the maximum safe shutdown rate."

The guidance contained in IE Bulleti- 79-05A has been incorporated in Draft TMI-1 Technical Specificati a 3.e.1 with the exception that the restoration time for the emergency feedwater system has been reduced from 72 hours8.333333e-4 days 0.02 hours 1.190476e-4 weeks 2.7396e-5 months to 48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months as a result of subsequent requirements from the NRC. Existing Technical Specifications 3.4.3 and 3.4.6 have been rewritten to incorporate remedial action in the esent that the condensate storage tanks and/or the main steam safety valves are inoperable. For both the condensate storage tank and the main steam safety valves, remedial action has been proposed that is consis".ent with NRC guidance as reflected in the B&W Standard Technical Specifications.

With regard to surveillence requirements, draf t Technical Specifica-tion 4.9, " Emergency Feedwater System," has been modified as follows:

(1) Existing Technical Specification 4.9.1.1 which requires testing of the e=ergency feedwater pumps every three months, as modified, would require pump testing every 31 days snd also require verification of specific pump flow values during the testing. The flow testing would be based the requirements of the ASME Boiler and Pressure Vessel Code,Section XI, Article IWP-3220, and would confirm that the emergency feedwater system can deliver at least 500 gpm to either steam generator flow path.

(2) Draf t Technical Specification 4.9.1.2 would require that, during testing of the EFW, a qualified, dedicated individual in communication with the control room, be maintained at the EFW manual valves. In the event that the EFW is required to function, the individual would promptly realign the manual valves from test to operational positions.

11-14 Am. 13

t l

(3) Draft Technical Specification 4.9.1.3 would require valve verification (correct position and locked, if appropriate) for valves in the emergency feedwater system, every 31 days. I In addition, locked salves could only be maintained in an un-locked condition under administrative control.

(4) Draf t Technical Specification 4.9.1.4 would require a test, each 18 months, of automatic pump start logic and automatic valve lineup following cn emergency feecwater actuation signal. In addition, the operability of the manual control valve station would be verified.

(5) Draf t Technical Specification 4.9.1.5 would require testing of the EFW injection valves on a quarterly basis.

(6) Itez 10F of NRC's October 26, 1979 letter to Mr. R. C. Arnold requires a, "... Technical Specifications to assure that prior to plant startup following an extended cold shutdown, a flow test would be performed to verify the normal flow path from the pri=ary EFW system water source to the steam generators.

The flow test should be conducted with EFW system valves in their normal alignment. " This test is incorporated in draft Technical Specification 4.9.1.6 where the term " extended cold

I shutdown" is interpreted as "a cold shutdown of longer chac 30 days' duration."

(7) Existing Technical Specification 4.1.2 (Table 4.1-2), would I recuire a functional test of the Backup Instrument Air Supply System (backup air supply for the emergency feedwater control valves), every refueling period.

(8) Existing Technical Specification 4.1.1 (Item 50 in Table l

4.1-1), as modified, would require a check each shif t, monthly testing, and calibration each refueling period, for the emergency feedwater flow instrumentation. The " check" and " test" surveillances would not be required when TAVG is less than 200*F since the reactor would be shutdown and this safety function not needed.

(9) Existing Technical Specification 4.5.1.1, as modified would I incorporate the motor driven feedwater pumps into the list of equipment whose operation is verified during the testing of the emergency diesel generators. In this case, only operation of the interlock would be verified since the pumps do not actually start on loss of. AC power (the actual start signal is on loss of main feedwater or loss of reactor coolant pumps.)

Conclusion In conclusion, with regard to the modifictions to the emergency feedwater systems and associated Technical Specifications:

11-1S Am. 13

(1) The probability or consequences of accidents previously evaluated nave not increased. The more conservative Limiting Condition for Operation and Surveillance Requirements for the emergency feedwater syste= previde increased assurance that the system will operate properly, when required.

(2) No accidents, other than those previously considered, will be introduced. The modifications to the emergency feedwater system could only effect the non-operation or spurious operation of the system; both of these conditions have been previously evaluated. The only aspect of the emergency feedwater system modification with the potential ror effecting other systems involves the loading of the motor driven feedwater pumps on the emergency diesel generators. An antlysis of the diesel generator loading indicates that the maximum load, with the emergency feedwater pumps is 2817 Kw compared to the 2000 hour0.0231 days 0.556 hours 0.00331 weeks 7.61e-4 months rating of 3000 Kw. The proper diesel generator loading sequence with the emergency feedwater pumps will be verified prior to startup and every 18 months thereafter. Other aspects of the emergency feedwater system will be tested prior to startup, and periodically thereafter.

(3) No safety margins have been reduced. The modifications to the emergen y feedwater system did not involve any changes which resulted in a decrease in capacity of this system to perform its designed function.

Based upon the above, we conclude that the modifications to the emergency feedwater system, and associated Technical Specifications, do not involve any unreviewed safety questions with regard to the criteria of 10CFR Part 50, Section 50.59(a)(2).

11.2.8 Post Accident Monitoring Introduction Section 2.1.8 of, "TMI-2 Lessons Learned Task Force Status Report and Short-Term Recommendations," NUREG-0578, July 1979 makes the following recommendations with regard to the increased range of radiation monitors:

" Provide high range radiation monitors for noble gases in plant effluent lines and a high-range radiation monitor in the contain-ment. Provide instrumentation for moltoring effluent releases lines capable of measuring and identifying radiciodine and par-ticulate radioactive effluents under accident conditions." In addition, the NRC recommended that facilities " Provide instrumen-tation for accurately determining in plant airborne radioiodine concentrations to minimize the need for unnecessary - < - of res-piratory equipment. In an August 13, 1979 ACRS memor adum to the NRC, the ACRS recommended the following additional post-accident instrumentation: (1) containment pressure, (2) containment water level, and (3) on-line monitoring of hydrogen concentration in the containment.

11-16 Am. 13

The post-accident monitoring instruments to be installed at 241-1 are responsive to the recommendations of the NRC and the ACRS.

Evaluation Section 2.1.2.1 of " Report in Response to NRC Staf f Rr. commended Requirements for Restart of Three Mile Island Nuclear Station Unit 1" describes the post-accident monitoring instrumentation to be installed at TMI-1. The post-accident instrumentation, in conformance with Regulatory Guide 1.97, consists of the following:

(1) Containment Pressure - the range will be - 5 psig to three times the containment design pressure; (2) Containment Water Level - a narrow range monitor will measure containment sump level while the wide range monitor will measure from the bottom of the containment to a 10 ft.

level; (3) Ccntainment Hydrogen Indication - continuous reading of the concentration of hydrogen in the containment, from 0 to 10%,

will be available in the control room; (4) High Range Containment Radiation Monitor - two monitors with a range to 107 R/hr will be provided; (5) High Range Effluent Monitors:

(a) Undiluted Containment Exhaust - 105 C1/cc (b) Diluted Containment Exhaust - 104 Ci/cc (c) Auxiliary and Fuel Handling Building Exhaust 103 C1/cc (d) Condenser Off Gas - 102 C1/cc (e) High Range Effluent Radio Iodine & Particulate Sampling and Analysis - silver :eolite cartridges.

Although the above instrumentation does not actuate saf ety equipment, nor is it required by safety analyses, it is appropriate to provide Surveillance Requirements to assure reliable post-accident performance of the instrumentation. Surveillance Requirements for post-accident monitoring ~ instrumentation is incorporated into TMI-l Draft Technical.

Specification 4.1.1 (Table 4.1-1) as follows:

(1) Item 13 of Table 4.1-1, "Eigh Reactor Building Pressure," is provided with a footnote to include the post-accident instru-mentation in the existing containment pressure instrumentation surveillance program; (2) Item 28 of Table 4.1-1, " Radiation Monitoring Systems," is provided with a footnote to include the post-accident instru-mentation, described in item (5)(a) thru (5)(d) above, in the existing radiation monitor system surveillance program; 11-17 Am. 13

(3) Ite: 37 of Table e.1-1 " Reactor Building Sump Level" has been changed to " Reactor Building Sump and Containment Level." A foot-note has also been added to include the post-accident intrumentation in the sump level instrument surveillance program.

(4) A new item 52, " Reactor Building Hydrogen Concentration,"

has been added to address Surveillance Requirements for the reactor building hydrogen concentration instrumentation. The

" check" and " test" surveillance need not be performed when TAVG is less than 200*F since the reactor would be shutdown and this safety function not needed.

Conclusion With regard to TM1-1 post-accident monitoring instrumentation, and associated Technical Specifications, since the instrumentation does not actuate safety equipment, nor is it required by the safety analysis:

(1) The probability or consequences of accidents previously evalu-ated have not increased.

(2) No accidents of a type, not previously evaluated, will occur, and (3) No saf ety margins have been reduced.

Based upon the above, we conclude that the post-accident monitoring instrumentation, and associated Technical Specifications, do not involve any unreviewed safety issues with regard to the criteria of 10CFR, Part 50, Section 50.59 (a)(2).

11.2.9 Reactor Coolant Pump Trio on Coincident ISFAS and Coolant Voiding Introduction The IE Bulletin Nos.79-05C and 79-CAC, July 26, 1979 states that, "Recent preliminary calculations periermed by Sabcock & Wilcox, Westinghouse and Combustion Engineering indicate that, for a certain spectrum of small breaks in the reactor coolant system, continued operation of the RCPs can increase the mass lost through the break and prolong or aggravate the uncovering of the reactor core.

The damage to the reactor core at IMI-2 followed tripping of the last operating RCP, when two phase fluid was being pumped through the reactor coolant system. It is our current understanding that all three of the nuclear steam system suppliers for PWRs now agree that an acceptable action under LOCA symptoms is to trip all oper-ating RCPs immediately, before significant voiding in the reactor coolant system occurs."

With regard to reactor coolant pump trip, IE Bulletin Nos.79-05C and 79-06C recommends the following long-term action:

11-18 Am. 13

" Propose and submit a design which will assure automatic tripping of the operating RCPs under all circumstances in which this action may be needed."

Section 2.1.2.5 of " Report in Response to NRC Staff Recommended Response to NRC Staff Recom= ended Requirements for Restart of Three Mile Island Nuclear Station Unit 1" contains a description of the reactor coolant pump trip that is proposed f or TMI-1.

Evaluation The logic for the reactor coolant pump trip receives inputs from the High Pressure Injection (HPI) signal from the ESFAS, and redundant pump current sensors from each of four reactor coolant pu=ps. The pump trip will occur on concurrent HPI and low current in two of the four reactor coolant pumps. Non-operating reactor coolant pumps have eff ectively tripped current sensors.

With only two reactor coolant pumps operating, therefore, these pumps will trip on RPI. The cutput of the trip logic provides a trip signal to each reactor coolant pump. Limiting Conditions for Operation and Surviellance Requirements for the Reactor Coolant Pump Trip are addressed below.

Limiting Concitions for Operation for the Reactor Coolant Pump Trip are presented in TMI-l draft Technical Specification 3.5.7.

The draft Technical Specification requires that an HPI actuation channel and one pump current channel f rom each operable pump be available to the reactor coolant pump trip logic. In the event that the reactor coolant pump trip is inoperable, hot shutdown must be achieved within 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . The draft Technical Specifica-tion only allows continued reactor operation if the pump trip is in a condition which assures reliable operation. The remedial action is specified to allow a reasonable time to restore the reactor coolant pump trip to operability or achieve an orderly shutdown. The Surveillance Requirement for the reactor coolant pump trip is contained in 241-1 draft Technical 9pecification 4.1.1 (Table 4.1-1). A new item, number 51, proposes a pump trip channel check each shift, a test each month, and a calibration each refueling period. The draft Surveillance Requirement for the pump trip is consistent with the surveillance for other safety instrumentation channels. The " check" and " test" surveillances need not be performed when TAVG is less than 200*F since the reactor is shut down and this safety function is not needed.

Conclusion With regard to the reactor coolant pump trip, the logic is designed to provide high assurance that the reactor coolant pumps will be triped when required. Any single failure within the reactor coolant pump trip logic will result in only a single reactor coolant pump being tripped. The draf t Limiting Condition for Operation for the reactor coolant pump trip prevents extended reactor operation if the reactor coolant pump trip is significantly degraded. The draf t Surveillance Requitecent for the reactor coolant pump trip provides assurance of reliable operation.

11-19 Am. 13

11.2.10 TMI-1/TMI-2 Separation Introduction Item II.4 of the NRC's August 9, 1979 " Order and Notice of Hearing,"

requires that, "The licensee shall demonstrate that decontamination and/or restoration operations at TMI-2 will not affect safe operations at TMI-1. The licensee shall provide separation and/or isolation of TMI 1/2 radicactive liquid transfer lines. Fuel handling areas, vent-ilation syscens, and sampling lines. Ef fluent =onitoring instruments shall have the capability of discriminating between ef fluents resulting from Unit 1 or Unit 2 operations."

Section 7.2 of " Report in Response to NRC Staff Recommended Require-ments for Restart of Three Mile Island Nuclear Station Unit 1" de-scribes a plan to separate 241-1/TMI-2 interfaces that have the potential of transferring significant quantities of contamination as a result of restoration activities at TMI-2.

Evaluation The two major pathways for potentia! transfer of contamination from TMI-2 to IMI-l are the waste zanage=ent interconnections and the common air space of the Fuel Handling Building. The following TMI-1/TMI-2 waste management interfaces have been identified:

(1) Unit 2 Reactor Coolant Bleed Holdup Tank - Unit 1 Reactor Coolant Waste Evaporator.

(2) Unit 1 Miscellaneous Waste Evaporator - Unit 2 Evaporator Condensate Test Tank.

(3) Unit 2 Nuetralizer Tanks , Contaminated Drain Tanks , Reactor Coolant Bleed Holdup Tanks, Auxiliary Building Sump Tanks and Miscellaneous Waste Holdup Tanks - Unit 1 Liquid Waste Disposal System.

(4) Unit 1 Evaporator Concentrate - Unit 2 Evaporator Concentrate.

(5) Unit 1 Spent Ion Exchange Resin - Unit 2 Spent Ion Exchange Resin.

Draft IHI-1 Technical Specification 4.1.2 (Table 4.1-2, Item 13) requires the isolation devices (valves , blank flanges , etc.) on the above tielines to be verified to be isolated, by visual in-spection, on a monthly basis. Draft IMI-l Technical Specification 3.19 requires that, if an isolation device is found to be open with-out prior NRC authorization, a " Thirty Day Written Report" must be prepared per TMI-l draft Technical Specification 6.9.2.B(5). In addition, TMI-1 draft Technical Specification 3.19.2 requires NRC approval prior to creation of additional EdI-1/TMI-2 system inter-ties that can transfer potentially significant quantities of con-tamination.

11-20 Am. 13

With regard to the separation of the air space in the Fuel Handling Building, the details of this modification have not been finalized.

Additional evaluations and preparation of draft Technical Specifica-tions will be undertaken, if appropriate, following finali:stion of the design details of the Fuel Handling Building isolation system.

Conclusion The draft Thl-1 Technical Specifications 3.19 and 4.1.2 for the TMI-1/

TMI-2 interties provide assurance that:

(1) System intenties that could potentially transfer significant quantities of contamination from TMI-l to IdI-2 will re=ain closed.

(2) If permission is received f rom the NRC to open system interties, these intercies will be used in accordance with plant pro:edures.

(3) No new system interties, with the potential for transf erring significant quantities of contamination from TMI-2 to IdI-1, will be created without prior NRC approval.

The above controls limit releases from TMI-l to materials under control at IdI-1 and thus to previcusly evaluated quantities and concentrations of contamination.

11.2.11 Low Reactor Coolant System Pressure Channel for HPI/LPI Initiation Introduction The Low Reactor Coolant System Pressure Channel setpoint, which is used as input to the ESFAS logic, is determined based on a generic LOCA analysis. The generic LOCA analysis for TMI, referenced as "ECCS Analysis of B&W's 177-FA Lowered-Loop NSS ," BAW-10103, has referenced the Low Reactor Coolant System Pressure setpoint as 1600 psig compared with the Technical Specification value of 1500 psig. The setpoint actually used in the BAW-10103 calculations, however, was 1350 psig.

Evaluation The TMI-1 Technical Specification 3.5.3.1, " Engineered Safeguards Protec tion System Actuation Setpoints ," requires the Low Reactor Coolant System Pressure HPI/LPI initiation setpoint to be > 1500 psig. Draft TMI-1 Technical Specification 3.5.3.1 would require the Low Reactor Coolant System Pressure HPI/LPI initiation setpoint to be raised to > 1600 psig. In the event of a LOCA, the only impact of the 100 psig increase in the minimum Low Reactor Coolant System Pressure setpoint would be to initiate actions, based on this signal, at an earlier time in the accident (e.g., in conjunc-tion with the 4 psig High Reactor Building Pressure, both HPI and LPCI pumps would start earlier in the accident.)

11-21 Am. 13

Conclusion With regard to the 100 psig increase in ene minimu= Low Reactor Coolant System Pressure HPI/LPI initiation setpoint:

(1) The probability or consequences of accidents previously evalu-ated have not increased. The potential initiation of engineered safety feature equip =ent at an earlier time in a LOCA is not expected to have a significant impact on peak clad temperature anc other LOCA limits (any changes would be expected to be in direction of a less severe accident).

(2) No accidents of a type not previously evaluated will occur.

The proposed change in the Low Reactor Coolant System Pressure Setpoint woulc have only a s=all impact on the severity of the LOCA, in the conservative direction, rather than change the nature of the accident.

(3) No safety margins have been reduced. The applicable LOCA calculations continue to be those for which the Low Reactor Coolant System Pressure HPI/LPI initiation setpoint is 1350 psig; operationally, raising the minimum setpoint to 1600 psig would slightly increase the LOCA margins.

Based upon the above, we conclude that raising the minimum Low Reactor Coolant Pressure setpoint from 1500 psig to 1600 psig does not involve sny unreviewed safety questions with regard to the criteria of 10CFR Part 50, Section 50.59(a)(2).

11.2.12 Raising the Low Reactor Coolant System Pressure Trip Setnoint Introduction The 2dI-l Technical Specification 2.3.1 (Table 2.3-1, Figure 2.3-1) provides a value of 1800 psig for the R?S Low Reactor Coolant Pressure trip setpoint. The B&W generic ECCS analysis , "ECCS An .1-ysis of B&W's 177-FA Lowered Loop NSS," BAW-10103, Rev. 2, Aprii 1976, referenced a value of 1900 psig for the Low Reactor Ce .c Pressure Trip setpoint. Draft Technical Specification 2.3.< would increase the Low Reactor Coolant System Pressure Trip Setpcint from 1800 psig to 1900 psig.

Evaluation The principal reason for the Low Reactor Coolant System trip set-point is to maintain thermal margins for the fuel by preventing the minimum DNB ratio from decreasing below the safety limit of 1.3; the transient analysis for TMI-l is based on an 1800 psig Low Reactor Coolant System trip setpoint. The Low Reactor Coolant System trip setpoint is also credited in the ECCS analysis since a reactor trip is part of the assumed LOCA scenario.

11-22 Am. 13

By increasing the Low Reactor Coolant System Pressure setpoint from 1800 psig to 1900 psig, the reactor would trip earlier in the LOCA scenario and thus the decay heat would be slightly less when the ECCS functions. Increasing the Low Reactor Coolant System trip setpoint also has the effect of increasing the margin to DNS following a trip on low pressure; the reactor would trip earlier on low pressure and enus the final minimum DNS would be higher (more conservative) than if the reactor tripped at 1800 psig.

Conclusion With regard to increasing the Low Reactor Coolant System trip setpoint from 1800 psig to 1900 psig:

(1) Tne probability or consequences of accidents previously con-sidered have not increased. For any accident that involves a pressure decrease, the reactor will trip earlier in the trans-ient and thus the result of the accident will be more conser-vative.

(2) No accident of a type not previously evaluated , will occur.

The increasing of the Low Reactor Coolant System trip setpcint wil? not have any effect other than tripping the reactor at an earlier time in pressure reduction transients.

(3) No safety margins have been decreased. It is expected that for pressure reduction transients, DNB following the reactor trip will be higher (more conservative) and for the LOCA, the peak clad temperature and other system parameters will be more favorable.

Based upon the above, we conclude that increasing the Low Reactor Coolant System trip setpoint from 1800 psig to 1900 psig does not involve unreviewed safety questions with regard to the criteria of 10CFR Part 50, Section 30.59(a)(2).

11-23 Am. 13

Leaf: Technical Specification Corresponding to See:1on 11.~."

as Revised by Amendment 13 s

e

3.4 DECAY HEAT REMOVAL - TURBINE CYCLE Aeolicability Applies to the operating status of equi.pment that functions to remove decay heat, utilizing the secondary side of the steam generators.

Obieetive To define the conditions necessary to assure im=ediate availability of the auxiliary feedwater system and main steam safety va.lves.

Soecification 3.4.1 With the reactor coolant system temperature greater than 2500F, three independent steam generator emergency feedwater pumps and associated flow paths

shall be OPERA 3LE with:

a. Two emergency feedwater pumps, each capable of being powered from an OPERA 3LE emergency bus, and

b. One emergency feedwater pump capable of being powered from an OPERABLE steam supply system. With one emergency feedwater pump or flow path" inoperable, restore the inoperable pump or flow path to OPERA 3LE status within 45 hours5.208333e-4 days 0.0125 hours 7.440476e-5 weeks 1.71225e-5 months or be in COLD SHUTDOWN within the next 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months . With more than one emergency feedwater pump or flow path" inoperable, restore the inoperable emergency feedwater pu=ps or flow paths" to operable status or be suberitical within 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months , in at least HOT SHUTDOWN within the next 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , and in COLD SHUTDOWN within the following 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months .

c. Four of six turbine bypass valves are OPERABLE.

3.4.2 The condensate storage tanks (CSTS) shall be OPERABLE with a mini-mum of 150,000 gallons of condensate available in each CST. With a CST inoperable, restore the CST to operability within 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months or be in at least HOT STANDBY within the next 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , at least HOT SHUTDOWN within the next 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , and COLD SHUTDOWN within the next 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

"For the purpose of this requirement, an OPERABLE flow path shall mean an unobs tructed path from the water source to the pump and from the pump to a steam generator.

3-25

3.4.3 Wi:h the reactor coolant syste= :empera:ure greater than 2500F, all eignteen (18) main steam safety valves shall be operable er.

if any are not operable, the maximu= overpower trip setpoint (ree Table 2.3-1) shall be rese: as follows:

haximu= Nu=ber of Maxi =um Overpower Safe:y Valves Disabled on Trip Se: point Anv Steam Generator (: of Rated Pcwer) 1 92.4 2 79.4 3 66.3 With more :han 3 main s: cam safety valves inoperable, restore at leas: fif:een (15) main steam safety valves to operable status within 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months or be in a: least hot standby wi:hin the .sext 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months and in cold shutdown within the following 30 hours3.472222e-4 days 0.00833 hours 4.960317e-5 weeks 1.1415e-5 months .

Bases A reactor shutdown following power operation requires removal of core decay heat. Normal decay heat removal is by the steam generators with :he steam dump to the condenser when system te=perature is above 250'F and by the decay hea: removal system below 250*F. Core decay heat can be con:inuou. sly dissipated up to 15 percent of full power via the steam bypass to thee condenser as feedwa:er in :he steam generator is converted to steam by heat absorption.

Normally, the capability to return feedwater flow to the steem generators is proviced by the main feedwa:er system.

The main steam safety valves will be able to relieve to atmosphere the to:al stea= flow if necessary. If main stea= safety valves are inoperable, the power level must be reduced, as stated in Technical Specification 3.4.3, such that the remaining safety valves can acco=modate the decay heat.

In the unlikely event of complete loss of of f-site elec:rical power to the station, decay heat removal is by either the s: cam-driven emergency feedwater pump, or two half-si:ed motor-driven pumps. Steam discharge is to the atmosphere via the main s:eam safety valves and con: rolled at-mospheric relief valves, and in :he case of the turbine driven pump, from :he turbine exhaust.(1)

Both motor-driven pu=ps are required initially to remove decay heat with one eventually sufficinF. The minimum amount of water in :he condensate storage tanks, contained in Technical Specification 3.4.2, will allow cooldown :o 250*F with steam being discharged to the atmosphere. After cooling to 250*F, the decay heat removal system is used to achieve further cooling.

An unlimited emergency feedwater supply is available from the river via ei:her of the two motor-driven reactor building emergency cooling water pumps for an indefinite period of time.

3-26

The requiremen:s of Technical Specifica: en 3.6.1 assure that before the reac:or is heated to above 250*F, adequate auxiliary feedwater capacity is availaole. One turbine driven pump full capaci:y (920 gp ) and the two half-capaci:y motor-driven pu=ps (460 gps, each) are spe:ified. Ecwever, only one half-capacity notor-driven pump is necessary to supply auxiliary feedwater flow to the s:cas generators in :he onset of a s all break loss-of-coolant acciden: (Reference 2).

The requirements of Technical Specifica: ion 3.4.1 assure tha: a: leas:

920 gp is available at all times to both s:eam generators giving re-dundan: capacity except for a limited time of 72 hours8.333333e-4 days 0.02 hours 1.190476e-4 weeks 2.7396e-5 months to allow for cce-panent maintenance. Further degrada:1on of :he emergency feedvater sys:e=

requires the reactor :c be suberi:1 cal within I hour.

The feedvater line break accident performed for TMI-2 (Reference 3) shows satisfae:1on of core ther=al power limits and reactor coolant sys:e:

pressure limi:s assu=ing full auxiliary feedwater flow within 40 se:onds.

The Technical Specifica: ion 3.4.1 provides assurance that this flow will be available with auto =atic initia: ion following loss of be:h =ain feedwa:er pumps.

REFERENCES (1) FSAR, See:1on 10.2.1.3 (2) " Evaluation of Transient Behavior and Small Reactor Coolant Syste=

Breaks in the 177 Fuel Assembly Plant," Volume I and II, Babcock and Wilcox, May 7, 1979.

(3) Three Mile Island Nuclear Station - Uni 2, Final Safety Analysis Report, USNRC Docke No. 50-320.

3-26a

TABL .1-1 (Continued)

CilANNEL DESCRIPTION CliECK TEST CALIllRATE REMARKS

38. Steam Generator Water Level W NA R

39. Turbine Overspeed Trip NA R flA

40. Sodium Thiosulfate Tank Level NA NA R Indicator

41. Sodium Ilydroxide Tr.nk I,evel flA I:A R Indicator

42. Diesel Generator Protective NA h R Relaying ,

43. 4 KV ES Bus Undervoltage Relays NA H(l) R (1) Relay operation will be checked (Diesel Start) by local test pushinit t ons.

44. Reactor Coolant Pressure S(!) H R (I) When reactor coolant system is Dil Valve Interlock Bistable pressurized above 300 psig or Taves Is grea t er alean 200*F.

45. Insa of Feedwater Trip S(l) H(1) R (1) When reactor > 101 power.

46. Turbine Trip / Reactor Trip S(i) !!( I) R (I) When reactor > 201 power.

47. Pressurizer Code Safety Valve and Electromatic Relief valve del ta P/ flow S(l) R R (1) When TAVC is greater than 2OO"F.

48. Pressurizer Electromatic Relief Valve acoustic / flow S(i) R k (1) When Tayq is steater than 200"F.

49. Saturation Margin Heter S(I) H(l) R (l) When T Ayc Is greater ihan 2OO"F.

50. Emergency Feedwater Flow Instru- NA H(t) mentation R (1) Emergency Feedwater is not normally in operation.

S - Each Shift T/W - Twice per week R - Eacle Refueling I'eriod D - Daily B/H - Every 2 months flA - Not Applicable W - Weekly Q - Quarterly n/W - Every two weeks N - Honthly P - Prior to each startup -

if not donc previous week

TABLE 4.1-2 MINIMUM EQUIPMENT TEST FREQUENCY I:em Tes: Freouenev

1. Control Rods Rod drop times of all Each refueling shutdown full length rods

2. Con:rol Rod Movemen: of each rod Every two weeks, when reactor Movement is critical

3. Pressurizer Safe:y Se:poin:* 50% each refueling period Valves

4. Main Stea= Safe:y Setpoint 25% each refueling period Valves

5. Refueling Syste= Fune:ional Start of each refueling period Intericcks

6. Main Steam (See Section 4.8)

Isolation Valves

7. Reactor Coolant Evalua te Daily, when reactor coolant Sys:em Leakage system temperature is grea:er than 525*F C; Charcoal and high DOP test on EEPA Each refueling period and at efficiency filters filters, f reon test any time work on filters for Control Room, on charcoal filter could alter their integri:y and R3 Purge uni t s Fil:ers

9. Spen: Fuel Cocling Functional Each refueling period prior to System fuel handling

10. Intake Pu=p House (a) Silt Accumulation- Each refueling period Floor Visual inspection of (Elevation 262 Ft. Intake Pump House Floor 6 in.) (b) Silt Accumulation Quarterly Measurement of Pump House Flow

11. Pressurizer elec- Setpoint Each refueling period tromatic relief valve

12. Back-up instrument Functional Each refueling period air supply system

The setupoint of the pressurizer code safety valves shall be in accordance with ASME Boiler and Pressurizer Vessel Code,Section III, Article 9, Winter, 1968.

4-8

4.5 EMERCENCY LOASING SEQUENCE AND POWIR TRANSFER, EMERCENCT CORE COOLIFG SYSTEP ANC REACTOE B"lLDINC COOLING SYSTEM PERIODIC TESTING 4.5.1 EMERCENCY LCA21NC SEQUINCE Acrlicability Applies to periodic testing requirements for safe:y ac tuation sys te=s.

Objective To verify tha: the E=ergency leaiing sequence and au:omatic power ::ans-fer is operable.

Soecifications 4.5.1.1 Secuence and Power Transfer Tes:

a. During each refueling interval, a tes: shall be condue:ed to demon-strate that :he emergency loading sequence and power transfer is operable.

b. The test will be considered satisfac:ory if the following pumps and f ans have been successfully s:arted and the following valves have co=pleted : heir travel on preferred power and transferred to the emergency power as evidenced by the control board component operating lights, and either the sta: ion computer or pressure /

flow indication or, in the case of :he Motor Driven Emergency Feedwater Pumps, the pump Interlock.

- M. U. Pump

- D. H. Pu=p and D. H. Injection Valves and D. H. Supply Valves

- R. B. Cooling Pump

- R. B. Ventilators

- D. H. Closed Cycle Cooling Pump

- N. S. Closed Cycle Cooling Pump

- D. H. River Cooling Pu=p

- N. S. River Cooling Pump

- D. H. and N. S. Pump Area Cooling Fan

- Screen House Area Cooling Fan

- Spray Pump. (Ini:iated in coincidence with a 2 out of 3 R. B. 30 psi Pressure Test Signal.)

- Motor Driven Emergency Feedwater Pump Interlock.

4.5.1.2 Secuence Test

a. At i.tervals not to exceed 3 months, a test shall be conducted to de nonstrate that the emergency loading sequence is operable, this :est shall be performed on either preferred power or emer-gency ,ower.

b. The test will be considered sa:isf actory if the pumps and fans listed in '. 5.lb have been successfully started and the valves listed in 4.5.1.lb have completed their travel as evidenced by the control board component operating lights, and either the station computer or pressure / flow indication.

4-39 f

Lases The Emergency loading sequence and autceatic power transfer controls the operation of the pumps associated with the e=ergency core cooling syste= and Reactor Building cooling system. A successful test of the e=ergency loading sequence and automatic power transf er is a prerequisite to any system test of the emergency core cooling syste= or reactor building cooling systes.

References (1) FSAR Section 7 (2) FSAR Section 1.4 4-40

4.9 EMERCENCY FEEDWATEP. SYSTEM PERIODIC TESTING Aeolicability Applies to the perodic testing of the turbine driven and two motor-driven emergency feedwater pumps, assocaited actuation signals, and valves.

Obiective To verify that ene auxiliary feedwater system is capable of performing its design function.

Specification 4.9.1 TEST 4.9.1.1 Whenever the Reactor Coolant System temperature is greater than 2500F, the emergency feedvater pumps shall be tested in the recirculation mode in accordance with the t !quirements and acceptance criteria of ASME Section XI Article IWP-3000. The test f equency shall be at least every 31 days ;J days of plant operation at Reactor Coolant Temperature above 2500F.

4.9.1.2 During testing of the emergency feedwater system when the reactor is in STAR. TUP or POWER OPERATION, if one steam gen-erator flow path is made inoperable, a dedicated qualified in-dividual who is in communication with the control room shall be continuously stationed at the EFW manual valves. On in-struction f rom the control room, the individual shall realign the valves from the test mode to their operational alignment.

4.9.1.3 At least ence per 31 days each valve listed in Table 4.9-1 shall be verified to be in the status specified in Table 4.9-1.

4.9.1.4 At least once per 18 months, during shutdown, verify that:

(a) each emergency feedwater pump starting logic actuates upon receipt of an auxiliary feedwater actuation signal, and (b) valves in the emergency feedwater flow paths

actuate to their l correct position on an emergency feedwater actuation signal and that the manual control valve station functions properly.

4.9.1.5 On a quarterly basis, the valves which are a part of the emer-gency feed system discharge (EFV-30A and 308) will be checked for f proper operation by cycling the valve over its full stroke.

4.9.1.6 Prior to start-up, following a cold shutdown of longer than 30 days' duration, conduct a test to demonstrate that the motor l driven emergency feed pumps can pump water from the CST to the steam generators.

For the purpose of this requirement, an OPERABLE flow shall mean an unobs tructed path from the water source to the pump and from the pump to a steam generator.

4-52

4.9.2 ACCEPTANCE CRITERIA These tests shall be considered satisf actory if control board indication and visual observation of the equipment demonstrates that all components have operated properly.

Bases The 31 day testing frequency will be sufficient to verify that the turbine driven and two motor-driven emergency feedwater pumps are operable and that the associated valves are in the correct alignment. ASME Section XI Article IWP-3000 specifies requirements and acceptance standards for the testing of nuclear safety related pumps. Co=pliance the normal acceptance criteria of IW?-3000 assures that the emergency feedwater pumps are operating as expected. The test frequency of 31 days (nominal) has been demonstrated by the B&W Emergency Feedwater Reliability Study to assure an appropriate level of reliability. If testing under Article IWP-3000 indicates that the flow and/or pump head for a particular pump is not within the normal acceptance standard, Article IWP-3000 requires that an evaluation of the pump perfcrmance shall be completed within 96 hours0.00111 days 0.0267 hours 1.587302e-4 weeks 3.6528e-5 months or the pump declared inoperable. For the case of the emergency feedwater system, the system shall be considered operable if under the worst case single pump f ailure, a minimum of 500 gpm of emergency feedwater can be delivered when steam generator pressure is 1050 psig and one steam generation is isolated. A flow of 500 gpm, at 1050 psig head, ensures that sufficient emergency feedwater, demonstrated to be acceptable for plant cooling requirements under transient and accident conditions, can be delivered to either steam generator flow path. The 18 month surveillance requirements ensure that the overall emergency feedwater system functional capability is maintained comparable to the original design standards.

4-52a

f Table 4.9-1 Status of EFW Valves Valve No. Status CO-V-10A Open l CO-V-10B Open EF-V-1A Open EF-V-1B Open E F-V- 2A Open EF-V-23 Open MSV-2A Open MSV-23 Open EF-V4 Locked Closed

EF-V5 Locked Closed

EF-V6 Locked Open*

EF-V10A Locked Open*

CF-V10B Locked Open*

EF-V-16A Locked Open*

EF-V-16B Locked Open*

EF-V-20A Locked Open*

EF-V-20B Locked Open*

EF-V-22 Locked Open*

Locked Valves, if Maintained in a status other than indicated, shall be under the administrative control of a dedicated qual-ified, individual who is in communication with the control room.

The individual shall be continuously stationed at the EFW manual valves end, on instruction from the control room, shall realign the valves from the test mode to their operational alignment.

SUPPLEMENT 1 - P.GT 1 QUESTI0F:

8. Describe and justify the method used to deter =ine the minimum required emergency Ieedwater flow ccpacity (sizing criteria). Verify that the cini=u: emergency feedwater is consistent with your safety analyses for all anticipated accident and transient conditions as-s' 'ng a single failure of any systeL component.

RESPONSE

See Section 2.1.1.7.6 and Supplement 1, Part i response to Appendix A of Question 10.

Am. 13

SUFPLEMFNT 1 - PAF.T 1 RESPONSE 10 APPENDIX A 0F OCESTION 10 RESPONSE TO APPENDIX A, QUESTION la The design basis event for sizing the Auxiliary Feedwater System (AFWS) is Loss of Feedwater (LOFW) with a concurrent Loss of Offsite Power (LOOP),

and subsequently loss of reactor coolant pumps. The pertinent parameters for this accident relative to the AFWS are design flowrate and required time to full AFWS flow. These parameters reflect the functional require-ments of the AFWS to a) remove decay heat, and b) provide a smooth reactor coolant flow transition from RC pu=p operations to natural circulation.

The design values which resulted from this analysis are 720 gpm deliverable to the steam generators within 40 seconds of the initiation signal. The 40 second time was chosen to allow the AFWS to inject feedwater and begin in-creasing SC level to the 50% operating range level, required for natural circulatior. prior to completion of the RC pump coastdown. At that time, the design clewrate was selected to be equal to or greater than the decay heat generation rate. Since decay heat rate changes with time, other values than 40 seconds and 720 gpm could have bean used and been acceptable. All other transients which either require or assume the availability of AFW in the Safety Analysis use the design values derived from the LOFW analysis.

The results of these other analyses are acceptable and are referenced in Table 1 attached.

Accidents 1,1, and 3 cf Table 1, which specifically require AFW for =iti-gation, were analyzed using the AFWS performante criteria established by the LOFW accident. The results of these analyses were acceptable and are described in the FSAR sections noted in Table 1. The other accidents listed in Table 1 (4-12) do not require AFW for mitigation though the availability of the AFWS, as defined by the perfor=ance criteria established by the LOFW accident, is assumad. The results of those analyses were acceptable and are described in FSAR sections noted in Table 1.

The events included in the NRC Appendix A question la which have not been included in Table 1 are discussed in Section 8. In addition, it should be noted that:

a) LMFW v/ loss of onsite and offsite AC power - This event was not a design basis of the plant and, consequently, is not included in Chapter 14 of FSAR.

b) Plant Cooldown - Plant cooldown with AFW is a new issue as stated in Reg.

Guide 1.139 and not a design basis for this plant. The NRC has not indi-cated how Reg. Guide 1.139 is to be applied to operating plants. The --

tent of plant cooldewn for which the AFWS is designed is discussed in IS*"

Section 14.1.2.8.3d.

c) Turbine Trip with and without bypass - This event does not affect the AFWS unless MFW fails in which case the loss of MFW event previously addressed would bound the AFWS design.

Main Steam Isolation Valve Closure - Again, this event does not directly affect the AFWS unlass MFW is lost as discussed above.

Am, 13

t SUPPLDIIITf 1 - PART 1 RESPONSE TO APPENDIX A OF OUESTION 10 EESPONSE TO APPENDIX A, QUESTION la (Continued) d) Main Feed Line Break - This event was not e required analysis for this plant and is not included in FSAR Section 14. Main Feedline Break is a mora abrupt case of LOFW and results of an analysis would be approxi-

=ately the same.

e) Small Break LOCA - The AE criteria assured for this event is described in Topical Report RAW-10052 updated by Letter Report J. H. Taylor B&W to S. A. Varga NRC 7/16/78 and the recently submitted B&W Report titled

" Evaluation of Transient Behavior and Small Reactor Coolant System Breaks in the 177 FA Plant," 5/7/77.

See also Supplement 1, Part 2 Response to Question 3.

Am. 13

SUPPLEMENT 1 - PART 1 RESPONSE TO AFPENDIX A 0F OUESTION 10 ,

RESPONSE TO APPENDIX A, OUESTION lb The design basis event for sizing the AFWS is LOFW as discussed in response to Question la. The acceptance criteria for the other transients which in-clude or assume AFW are given in Table 1.

The RCS cooling rate is not a 14-4 t relative to accident acceptance criteria.

The safety liait for all transients which use AFW for citigation is that the core re=ain cooled with ultimate acceptance criteria being those addressed in Table 1. For transients which result in draining the pressurizer or for which natural circulation is slowed or interrupted, restoration of pressurizer level and subcooling is accomplished by swelling due to core heat input and inven-tory restoration by EPI.

Steam Generator level is not based on decay heat removal rate or cooldown capability. SG level is set low for decay heat removal and high for natural circulation. It is also set high for Small LOCA as described in Topical Report BAW-10052, and in the B&W Report " Evaluation of Transient Behavior and Small Reactor Coolant System Breaks".

Am. 13

SUPPLEHENT 1 - PART 1 l RESPONSE TO AFPENDIX A 0F CUESTION 10 l t

TABLE 1 (1)

ACCIDENT DESCRIPTION FSAR SECTION ACCEPTANCE CRITERIA

1) Loss of Coolant Flov 14.1.2.6 A, B

2) Loss of Electric Power 14.1.2.8 A, B

3) Stean Line Break 14.1.2.9 D

4) Uncompensated Operating Reactivity 14.1.2.1 A, B Changes

5) Start-Up Accident 14.1.2.2 A, B

6) Rod Withdrawal Accident at Rated 14.1.2.3 A, B Power Operation

7) Moderator Dilution Accident 14.1.2.4 A, B

8) Cold Water Accident 14.1.2.5 A, B

, 9) Stuck-Out, Stuck-In, or Dropped 14.1.2.7 A, B Control Rod Accident

10) Stean Generator Tube Failure 14.1.2.10 B, D

11) Rod Ejection Accident 14.2.2.2 C, D

12) Loss of Coolant Accident 14.2.2.3 D, E NOTE: (1)

KEY ACCEPTANCE CRITERIA TECHNICAL BASIS A Max. RCS Press. - 1102 Design ASME Code B 1.3 with BAW-2 SRP 4.4 C 200 cal./gra= fuel 11=1: Reg. Guide 1.77 D Acceptable Doses 10CFR100 E Fuel Cladding 2200 F 10CFR50.46 Am. 13

SUPPLDIENT 1 - PAR! 1 RESPONSE TO APPENDIX A OF QUESTION 10 l

RESPONSE TO APPENDIX A, QUESTION 2 As discussed in response to la) above, the design 'aasis event regarding AWS design require =ents is loss of main faeewater with concurrent loss of RC pumps; the analysis assu=ptions for this event are listed below keyed to the letters of the question. Corresponding technical justifica-tions where not specifically listed below, is based on licensing require-ments and prudent engineering judgment at the time of the analysis.

a) Max. Rx Power - 10C*

'b) Time delay initiating event to Rx trip - The reactor will trip on high RCS pressure approxi=ately 5-10 seconds af ter a LOW event.

The initiation signal for AW is loss cf main feedwater.

c) AWS initiation signal and time delay - The AW initiation signal for the LOW event is loss of both main feed pumps as sensed by stea= inlet valve position on the two nain feed punp turbines.

The design basis time delay fre: initiation event to full flow of AW flow into SG is 40 seconds.

d) SO level at initiation event - Stea= Generator inventory is dependent o< power level. In the most restrictive case, AW will be fed into iteam generators before they boil dry.

e) SG inventory and decay heat - For discussion of water inventory see d) above. Reactor decay heat rate is shown in FSAR Table 14-13.

f) Max. SG Pressure - 1103 psig g) Min no. of SG - The nu=ber of generators was not specified in the analysis, heat re= oval capability is the pertinent parameter and can be accomodated by 1 SG.

h) RC Flow Condition - Both natural circulation and RC pump operation were analy:ed.

1) Max. AW inlet te=perature - The e ar%r AW inlet temperature assumed was 90 F.

j) Steam, Feed Line Break time delay - The feedvater line break was not a required analysis for this plant. Refer to FSAR Section 14.1.2.9 for steam line break analytical infor=ation.

k) Main Feed Line volume and te=perature between SG and AWS - R/A -

There is not piping connection between the .M'S and AWS.

1) SG Normal Blowdown - N/A - The OTSG's do not have a blevdown system.

Am. 13

4 SUPPLDiENT 1, PART 1 RESPONSE TO APPENDIX A OF QUESTION 10

(

m) Water and metal sensible heat used - Plant cooldown capability was not a design basis for ATWS. Ir10" BTU / F was used for renoval of sensible heat from power operation to the 0 power reactor trip se: point.

n) Time at hot standby etc. relative to AFW inventory - The AFW inventory was sized for decay beat removal for 1 day af ter Rx trip as discussed in FSAR Section 14.1.2.8.3d. The design basis for AFWS is not plant cooldown; the NRC Req. Guide 1.139 requirements for operating plants have not yet been established.

t Am. 13 4

SUPPLEMENT 1 - PART 1 f RESPONSE TO APPENDIX A 0F QUESTION 10 RESPONSE TO APPENDIX A. OUESTION 3 See response to Question 3 Supplementi, Part 2

\

Am. 13

SUPPL E ."? 1. PART 1

(

OUESTION

45. (Order Ice: 1(d))

Ycur response to this item indicates that procedures have been or are still being revised. Provide the procedures develcped to de-fine operator action during small break LOCA's.

RESPONSE

EP 1202-6 has been provided for your review. This procedure defines operator guidance during small break LOCA. In addition, the B&'d Guidance (B&W Docu=ent 69-1106001) used in developing the small break LOCA proce-dures is attached. This guidance and its supporting annlyses have been reviewed and are applicable to IMI-1.

Am. 13

SUPPLEMENT 1, PAR 7 2

, QUESTION t

2. Your response to Question 9 is not complete. Answer the indicated concerns. In addition:

A. Provide design drawings of the modified instru=ent air system.

B. Provide the test plan (procedure) and results for the proposed EFW control valve failure mode verification test.

C. Provide the B5W evaluation on the consequences of overfilling the steam generator.

D. Provide the calculations which indicate that operator action is required within 7 to 15 minutes to prevent potentially adverse steam generator overfill conditions. Include sufficient information to allow us to identify the allowable time delay beyond which the consequences would produce unacceptable effects.

Describe each manual action required.

E. Provide the revised procedures for preventing steam generator overfill conditions, and indicate that adequate operator training in these procedures has been co=picted.

RESPONSE

A. The design drawings of the modified instrument air system were provided to R. Fitzpatrick and S. Newburry of the NRC during a meeting on November 19, 1979. The drawings submitted were:

ECM Package WO-023 D-215-044 S-212-007 CH1102 E-215-053 S-212-007 C01059 E-215-013 SS-208-712 E-115-011 B-210-527A C-302-271 B-210-528A C-302-272 B-201-044 E-304-275 B-201-043 E-304-277 .

In addition it should be noted that the modified instrument air design is such that, if the backup instrument air compressors are lost, EF-V30 A/B can be controlled in excess of 5 minutes. This 5 minutes is a=ple time, in conjunction with the time to overfill once control is lost (10 minutes), to allow the operator to take appropriate corrective action by procedure. This action is described on Page 4 to the response to this question.

B. The test procedure will be submitted prior to restart testing and the test results will be submitted after completion of the tests.

Am. 13

SUPPLDfENT 1 - PART 2

(

QUESTION

3. Your response to Question 8 is not complete. Provide the B&W study on transients such as loss of feedwater and . loss of offsite power which verify that minimum EFW flow requirements meet the 550 gpm technical specification co==itment. Provide the revised IMI-l Technical Specifications for our review prior to restart. Justify the applicability of the B&W study to TMI-1.

RESPONSE

Attached is the B&W study (document identifier 86-1102587-00) on the auxili-ary feedwater flow requirements following a loss of main feedwater. The analyses performed included the following assumptions:

1) The initial power level at the initiation of transient was 2772 Mwt.

2) The reactivity feedback coefficients used were representative of approximately 100 EFPD operation.

3) The ANS 5.1 decay heat curve was used with a 1.0 safety factor. The key input parameters used are documented in Table 3.2-1 of the B&W Report to the NRC dated May 7, 1979 and entitled " Evaluation of

' Transient Behavior and Small Reactor Coolant System Breaks in the 177 Fuel Assembly Plant." The input parameters assumed in this study are applicable for IMI-l which is only a 2535 Mwt plant. As noted in the B&W study, auxiliary feedvater flow rates as low as 370 gpm were found to provide satisf actory performance.

Additional work has also been done by B&W to de=enstrate that 500 gpm auxiliary feedwater flow is adequate following upset transients such as the loss of offsite power and the loss of normal feedwater. This analysis is also attached. The Loss of Main Feedwater (LOFW) with Loss of Offsite Power (LOOP), and subsequent Loss of RC Pumps, is the Design Basis Event for sizing the Auxiliary Feedwater System (AFWS). The LOFW event places the most stringent capacity (flow) and actuation time requirements of any event which requires AFW. The acceptance criteria are to initiate within sufficient time, with sufficient flow so that decay and sensible heat re-moval will be adequate to prevent reactor coolant swell from filling the pressurizer. For this event, the Design Basis Flowrate was set by B&W at 720 gpm deliverable to the steam gcnerators within 40 seconds of the ini-tiation signal. The design flowrate was selected with conservatism to be equal to or greater than the decay heat generation rate at time of Laitial AFW injection.

More recently, B&W analyses have been made that use lower flowrates than the Design Basis flow. These analyses show that the pressurizer will not go solid for a LOFW Event with or without LOOP. Specifically a 500 gpm minimum flowrate at 40 seconds has been shown to be acceptable. Flowrates below 500 gpm have not been analyzed by B&Wfor the LOFW event; however, Am. 13

SUPPLEMENT 1 - PART 2 RESPONSE TO QUESTION 3 (Continued) lower flowrates might also be seceptable in meeting the defined acceptance criteria.*

Note that while the RC pu=p heat could not be accounted for dirt:tly in the attached CADDS analysis, hand calculations were done to confirm that inclu-sion of the pu=p heat would allow the acceptance criteria (pressurizar does not fill solid) for the LOFW event to be met. These hand calculations suoer-imposed the RC pump heat on the CADDS analysis.

One assumption was not specifically addressed in the analysis was worst case OTSG pressure. Heat removal is controlled by steam pressure; a higher steam pressure will remove less heat than a low steam pressure. The highest pres-sure possible in the steam lines is the safety valve set pressure (1050 psi).

The heat removal capability of emergency feedwater may be represented in terms of energy removal versus time by taking the product of the flovrate and the feedwater enthalpy rise through the generator. For this analysis, the feedwater te=perature at the entrance to the generator was assumed to be 100F, with a corresponding enthalpy of 71 Bru/lbm. The enthalpy of saturated steam does not change appreciably between 900 psia and 1100 psia (1196 Btu /

lbm to 1189 Beu/lbm) . This indicatas that the effect of pressure change en enthalpy difference is small. The transient simulation assumed that af ter initial fluctuations, the stea= pressure re=ained at 1015 psig. Review of plant post-trip data indicates that this is a realistic value. The differ-ence between 1015 psig (1192 Beu/lbm) and a possible worst case of approxi-mately 1050 psig (1150.5 Beu/lbe). (Based on safety valve setpoints) is, therefore, insignificant and would not impact the analytical conclusion re-garding the adequacy of 500 gpm.

It should also be noted that this analysis was for the purpose of showing the adequacy of 500 gpm and not to determine the minimu= AFW acceptable for the transient. Therefore, it is possible a lower flowrate would produce acceptable results given the same assumptions and acceptance criteria.

The events which require emergency feedwater are:

1) Loss of Coolant Flow

2) Loss of Offsite Power (LOOP)

3) Main Steamline Break (MSLB)

4) Small Break LOCA's (Certain sizes)

5) Loss of Feedwater (LOFW)

The most demanding event in terms of the need for heat removal via EFW is the LOFW without LOOP since this event requires the removal of RC pu=p heat as well as decay heat. Therefore, based on the above discussion, the 500 pgm EFW flow available under single failure assumptions is adequate

The acceptance criteria for the minimum auxiliary flow race were that (1) the pressurizer does not go solid and (2) the electromatic relief valve

( does not actuate. An auxiliary flow rate of 500 gpm was found to meet these criteria. The assumptions used for these analysis are conservative for TMI Unit #1.

Am.' 13

SUPPLE'.ENT 1 - PART 2

(

RESPONSE TO OUESTION 3 (Continued)

(720 gpm is available under non-single failure conditiens).

Main Feedline Break is a sacewhat more abrupt case of LOFW. However, frem a long term cooling standpoint, the heat removal requirements are identical to those occurring during a LOFW. Therefore, although the initial respense to a feedline break may be more severe, the emergency f eedwater sizing recuire-ments based on LOFW considerations assures sufficient heat removal capacity to mitigate the line break accident. Therefore, the results of a LOFW and Feedline Break Accident are essentially the same.

Am. 13

(

SUPPLDIENT 1, PART 2 OUESTION I 14 Your response to question 10j provided in Amendmen: 5 is not co=plete.

Provide legible arrangement drawings for the EFW system showing the location of all system pumps, pipin; and valves. Provide qualifi-cation documentation which assures that the motor driven EFW pumps will start and remain operational under the environmental conditions (humidity and temperature) resulting from a postulated break in the main steam supply line to the turbine driven EFW pumps. Further, verify that the EFW control valves and actuators are qualified to function under these environmental conditions. Also, provide an analysis which justifies the environmental conditions (323*F) assumed as a resul: of the postulated steam line break.

RESPONSE

As described in response to Question 10j (Supplemen: 1, part 1) the subject break was not considered probable enough to warrant detailed design consideration at the time TMI-1 was licensed. Since TMI-1 was licensed, NRC acceptance criteria for EFW systems has been modified and the EFW system has taken on new importance. In recognition of this fact, Me:-Ed has initiated a complete design review of the EFW system to upgrade it to the current licensing criteria to the extent practicable. This review will consider and resolve the type of concerns raised by questions 12,13 and 14 above. We believe that this approach is preferred over an item resolution of issues. Nevertheless, a response to your specific concerns is given below.

t The EFW pumps have been certified to withstand the calculated environment.

A copy of the motor qualification certification is attached, together with the calculations which support the environmental conditions (323*F). En-vironmental qualification of EF-V30A/B to 323*F was not invoked as part of the original purchase order for these valves. We have determined, after extensive review of the v.1ve operators with the vendor, that certain ele-ments of the operator cannot withstand the necessary accident environment.

These elements can be readily replaced with parts suitable for operation in the accident environment. These parts will therefore be replaced to upgrade EF-V30A/B before restart.

Arrangtment drawings showing the location of i=portant EFW valves and piping was provided separately on November 28, 1979 (

Reference:

E-304-086, Rev. 14).

t Am. 13

SUPPLEMENT 1, PART 2

( OUISTION 25 E. S_iculator Training The team concept for casualty control was stressed. The shift super-visor was evaluated in his command role.

F. TMI Transient Constructive criticism of operator action during the transient was stressed in this portion of their training.

The ele =ents of this program are being incorporated in the Shif t Supervisors Development Program. Any person who will be subsequently assigned to the position of Shift Supervisor will be required to receive this training prior to the assumption of the duties and responsibilities of that position.

Position 4 Review Poliev The administrative duties of the Shift Superviser will be reviewed by ap-propriate members of the THI Generation Group Staff in order to identify functions that detract from or are subordinate to the management respon-sibility for assuring safe operation of the plant.

The results of this review and reco==endation will be documented by De-ce=ber 31. Appropriate reco=mendations will be approved by the Senior Vice President - Met-Ed, responsible for plant operations and will be implemented prior to March 1, 1980.

Am. 13

Docusant No. 56-11C5508-07 i

ATTACHMENT TO QUESTION 45, SUPPLEMEC 1, PART 2

o ANALYSIS

SUMMARY

IN SUPPORT OF INADEQUATE CORE COOLING GUIDELINES FOR A LOSS OF RC3 INVENTORY

(

B ABCOCK AND k'ILCOX l

( CONTENTS

1.0 INTRODUCTION

2.0 ANALYSIS

SUMMARY

2.1 Correlation of Cladding Temperature to Reactor Coolant Pressure-Temperature Conditions.

2.2 Excore Neutron Detector Behavior.

2.3 Behavior of Loop Flow Indica: ion.

t 9

l.0 INTRODUCTION I The TM1-2 Lessons Learned Task Forte Status Report, NUREG-0578, contains two sections addressing inadequate core cooling.

First, S2ction 2.1.9 requires that Licensees provide the analysis, emergency procedures, and training needed to assure that the reactor operator can recognize and respood to conditions of inadequate core cooling. Secondly, Section 2.1.3 requires that:

" Licensees shall provide a description of any additional instrumentation or controls (prinary or backup) proposed for the plant to supplement those devices cited in the preceding section giving an unambiguous, easy-to-interpret indication of inadequate core coeling. A description of the functional design requirements for the system shall also be included. A description of the procedures to be used with the proposed equipment, the analysis used in developing these procedures, and a schedule for installing the equipment shall be provided."

In response to NUREG-0578, an extensive program for inadequate core cooling has been developed. The objectives of this progran are as follows:

1. Develop operating guidelines that vill allow the reactor operator to recognize and respond to conditions of inadequate core cooling under the following conditions:

a. Power Operation with portions of the core in DNB.

b. Loss of RCS inventory without the reactor coolant pumps operating.

c. Loss of RCS inventory vich the reactor coolant pumps operating.

d. Loss of the Decay Heat Removal Syste= and Loss of RCS Inventory During Refueling Operations.

e. Loss of natural circulation due to loss of heat sink.

2. Provide recommendations for any additional instrumentation required to indicate inadequate core cooling under the conditions listed above. Included with the recommendaticus vill be:

a. A description of the ftuctional design requirements for the additional instrumentation.

b. A description of the Operating Guidelines to be used with the proposed equipment.

t

c. A description of the analyses used in developing these guidelines.

d. Installation schedules for additional instrumentation (if required).

To date, operating guidelines and supportive analyses are complete for the following conditions within the scope of the Inadequate Core Cooling Program:

1. Loss of RCS inventory without the reactor coolant pumps operating.

2. Loss or RCS inventory with the reactor coolant pumps operating.

3. Loss of natural circulation due to a loss of heat sink.

(

_3 This report is thus a partial submittal. Additional guide-lines / supportive analysis for pcVer operation - DNS condition and refueling operations will be submitted, along with instrumentation-rela..ed reconmendations in a subsequent report.

For the inadequate core cooling conditions examined herein, guidelines for operator action and a description of the plant behavior, for use in operator training sessions, have been prepared. This infor=ation is provided in the revisions to Parts I and II of the Small Break Operating Guidelines References 3, 4, and 5. Supportive analyses and information relating to the expected behavior of the out of core detectors and loop flow measurements during inadequate

core cooling conditions is provided in Section 2.0.

1

(

2.0 ANALYSIS

SUMMARY

Guidelines for inadequate core cooling and a description of the plant behavior to support additional operator training are presented in Parts I and II of the Snall Break Operating Guidelines. These guidelines are in part based on the operators ability to assess the transie_*. Section 2.1 describes the basis and results of an analycis perf erned to correlate fuel rod conditions based on the pressure and tenperature conditions of the RCS. This information provides a neans to detect and to initiate corrective actions for an inadequate core cooling event.

In addition, Section 2.2 and 2.3 provide a qualitative assessnent of the behavior of the source range out of core i

neutron detectors and loop flow neasurenents during periods of inadequate core cooling.

(

~5-2.1 Correlation of Cladding Te=perature to Reactor Coolant Pressure-Temperature Condition During the small break LOCA, without the reactor coolant pu=ps operating, core cooling is accomplished by keeping the core covered by a steam-water mixture. However, should the core uncover, the uncovered portion of the fuel rod would be cooled only by the steam produced by boiling in the covered portion of the rod. Under this situation, elevated cladding temperatures, which could r' c in cladding rupture and/or a significant productior nydrogen due to metsi-water reaction, would result. The inadequate core cooling guide-lines have been developed to allow the operator to determine if core uncovery has occurred and to define appropriate actions to prevent significant cladding damage and/or hydrogen 1 generation.

The core exit thermocouples, which measure the core outlet fluid temperature, are the most direct indicators available to the operator for deter =ining the core status during a small break LOCA. If these thermocouples indicate superheated fluid conditions, core uncovery is in progress. This behavior allows an assessment of core cooling. To develop operator guidelines, a series of computer calculations were performed to develop a correlation between the measured core outlet fluid temperatures and the peak cladding temperature. Using the above correlation, various levels of inadequate core cooling were de.'.ned and appropriate operator actions were developed (see Appendix).

(

1

-6+ --

The approach taken for this analysis was to non-mechanistically reduce the Reactor Ceolant System Inventory in order to develop the correlation between clad temperature and outlet fluid temperature. Core decay beat, based on 1.2 times the 1971 ANS standard for infinite operation, at 200 seconds after scram was utilized fer this evaluation. . Core uncovery for small breaks should not occur any earlier than 200 seconds; thus this assumption vill.=ax1=1:e power and the peak cladding temper 2. a in the uncovered portion of the fuel rod. Five power shapes, given in Figures 1 through 5, were analy:ed to cover a reasonable spectrum of core conditions and to ensure that an outlet fluid temperature indication used for operator action would correlate to a peak cladding temperature less than a selected value. Radial naaking I

factors were chosen such that the maximum local power was equal to the LOCA limit value.

The FOAM 1 code was utilized to predict the peak cladding temperature and core exit fluid temperature. Table 1 erovides a summary of the cases analyzed. A brief outline of the procedure utilized in the FOAM code is as follows:

1. Using the input total core power, axial power shape, system pressure, and solid water level, the core mixture height is determined. This mixture height is based on a radial peaking factor of 1.0 and reflects the average core swell level.

(

I 2. Assurtng thnt all decay heat is removed in the covered portion of the fuel rod, the cora steaming is calculated.

As with the core =ixture level, the steaming rat' is based on a radial peaking factor of 1.0.

3. Using the average core steaming rate, the fluid temperature, in the uncovered portion of the fuel rod, for the hot pin is computed. This calculation uses the input radial peaking factor. In determining the fluid temperature, as a function of elevation in the core, it is assumed that all the core energy is removed by the stea=.

4. Using the core steaming rate and the local fluid properties in the uncovered portion of the fuel rod, a surface heat transfer coeffici t, based on the Dittus-Boelter correlation 2, is calculated.

5. Steady-State, hot pin cladding temperatures are then determined based on the local finAd properties obtained by Step 3 and the surface heat transfer coefficient obtained by Step 4.

Figures 6 and 7 su=narize the results of the calculations performed for the five power shapes analyzed. These curves correlate the calculated core exit fluid temperatures for peak cladding temperatures of 1400F and 1800F, respectively.

From these results, a bounding set of curves, shown on Figure 8, was obtained for use in the operating guidelines.

_e_

The small break operating guidelines include a provision for prompt tripping of the RC pumps upon receipt of a low pressure ESFAS signal. If the RC pumps cannot be tripped, ccre cooling vill be provided by the continued forced circulation of fluid throughout the RCS. There are two ways that inadequate core cooling can occur for a small break with the RC pumps operating. First, with the RC pumps operating, the fluid in the RCS can evolve to a very high void fraction. Should the RC pumps trip at a time when the system void fraction is greater than approximately 70%, the amount of water left in the RCS would not be s s _~ . z e i e n t to keep the core covered and an inadequate core cooling situation may exist. For this situation, the analysis described in the previous paragraphs apply directly.

Secondly. if little or no ECCS flov is provided to the RCS, the fluid being circulated by the RC pumps vill eventually become superheated stea= due to the continued energy addition to the fluid provided by the core decay heat. Under these circumstances, an inadequate core cooling situation vill start to exist. Due to the forced circulation of the superheated steam through the core under these conditions, even with only one RC pump operating, the heat removal process is better than the steam cooling mode described for the pumps off situation. Thus, the indications and operator responses determined for no RC. pumps operating are appropriate for controlling an inadequate core cooling situation with the

[

RC pumps operating.

_9_

2.2 Excore Neutron Detector Behavior

(

The extore source range neutron detectors are available te provide indications of anomalous incore behavior, although they cannot uniquely quantify inadequate core cooling. A departure frem expected response is anticipated for conditions that lead to inadequare core cooling. Therefore, the behavior of the source range detectors say,in some instances, be used to confir= other indications of inadequate core cooling.

The behavior of the neutron flux following a reactor trip is monitored and recorded by the source range count rate instrumentation following reactor trip. An example of this trace is presented in Figure 5. Normally the detector count rate falls at rates characteristic of the various mechanisms f

of neutron production that exist following the trip. During a trip, the neutron flux undergoes a prompt decrease associated with the negative reactivity of the control rods.

Following the promp t decrease the neutron flux decays with an 80 second period, characteristic of the decay of the longest-lived delayed neutron group. The neutron flux continues to decay at this rate until it approaches the level produced by neutron sources and suberitical multiplication. Two types of neutron sources are important in the determination of neutron level following delayed neutron decay, namely:

(

(1) Fixed startup sources (2) Natural sources The most important of the natural sources is the photoneutron production (y , n) resulting from the interactions of high energy fission product gammas with deuterium (D;O). The photonautron level decteases consistent with the decay of fission products (primarily Kr and La ).

The source range detectors will respond to a decrease in water density through several mechanisms.

(1) Reduced water density will enhance neutron transmission from core to detectors.

(2) Reduced vacer density vill decrease the neutron sources (i.e., photoneutrons from the y, n reaction in D 20).

I (3) The reduced water density will decrease the core multi-plication factor due to the negative moderator coefficient.

Scoping calculations with a 1-dimensional transport code have shown that the dominant effect is the improved neutron trans-mission from core to detector. Thus, the source range detector count rate vill increase or the rate of decrease vill be altered, depending on the magnitude of the change in water density, even though the core is becoming more suberitical and the photoneutron source strength is decreasing.

The source range detectors cannot unambiguously detect inadequate core cooling because voiding in different regions of the core vill have different effects on the ezcore flux

_11 . - _ _ _ _

levels. If the reactor coolant punps are operating while

(

the prinary systen is partially voided, the stea= voids are expected te be evenly distributed. Under these conditions, the source range detectors are expected to read a higher

.than nornal count rate. If the reactor coolant punps are not operating, the stean and water vill separate. In order for the core to be inadequately cooled, the water level must drop below the top of the core. When this happens the source range detector count rate should increase.

However, as the level continues to drop, the continued decrease in the quantity of available water could reduce the photonautron production and suberitical nultiplication to the point'where the source range detector output could begin to decrease. Because of this conplex behavior, the source range detector should only be used to confir= other indications of inadequate" core cooling.

A correlation has been nade between the source range detector response and several key events that followed the TMI-2 accident on March 28, 1979. This correlation is shown in Figere 10. The following is a discussion of the significant events referenced to the source range detector behavicr shown in Figure 10. As was discussed above, there is consider-able uncertainty in interpreting the behavior of the source range detectors. Any interpretation should therefore be used with caution.

(

_12

1. Time 0400 - The neutron power in the reacter core de-(

creased rapidly to the source range, as is typical of reactor trip.

2. Time 0408 - E=ergency feedvater was established to the steam generators approxicately 8 minutes after reactor trip. The PORV had stuck open, and it continued to relieve reactor coolant.

During the first 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months following reactor trip, high-pressure injection (EPI) was initiated automatically several times as system pressure decreased. Each time the automatic systa= activated, the plant operators took manual control of the HPI system.

Infor=ation on HPI flow rates and times of injection was not available and had to be inferred from makeup tank levels, operator interviews, and the alar = printer.

3. Time 0420 cc 0540 - Reactor ecolant pumps were circulating a saturated, two-phase flow. The void fraction was increasing ,

due to loss of coolant through the PORV. From the SR plot, circulating two-phase ficw can be inferred from the noisy, gradually increasing signal prior to point A. The noise in the signal is due t the turbulent action in the two-phase fluid. A decrease in moderator density results in an increase in SR level due to reduced attentuation. At 0514, the RC pumps in loop B vere turned off by the operator, but no fuel damage is believed to have occurred d u ri r. g this time interval since calculations have shown that the circulating fluid from loop A provided adequate heat removal.

4 Time 0540 - The RC pumps in loop A vere turned off by the operator. As a result, the flow decreased rapidly, vich a corresponding separation of steam and water. The calculated water level uns at the bottom of the core inlet pipes, which are 3 feet above the top of the active core. The calculation was based

on coolant quality j us t prior to trip and was inferred to consist of 30 to 50" voids. This inference is consistent with

{

gentile tube flow measurements and source range data.

5. Tice 0540 to C615 - The water level in the core gradually decreased between points B and D. The change in slope of the SR detector level at point C was interpreted to indicate the start of detector uncovering. This is supported by the reflux boiler calculation and the coolant loss through the PORV. During this ti=e, the RCS was acting as a reflux boiler; that is, steam was being created in the core region, condensing in the steam generators, and returning to the core by the cold legs.

The return of cold water to the reactor vessel was verified by the subcooled temperatures observed in the cold legs during this period. Reactor coolant continued to be lost from the system through the POR7.

6. Time 0615 to 0654 - The block valve upstream of the PORV was closed at 0615, preventing furthet loss of reactor coolant.

The core was approximacely 50% uncovered at this point and re-mained near this level until 0654. During this time interval, system pressure increased rapidly from 620 to 2150 psig. System pressure was then manually regulated using the block valve.

7. Time 0654 - Based on alarm messages, it was concluded that RC pump 2-3 was started and ran either intermittently or continuously for approximately 19 minutes. The core coolant level increased with at most 2 to 3 feet of the core remaining uncovered.

This inferred level is supported by some incore thermocouple ,

readings which came on scale and read below 700F. In addition, the SR levels from E to F indicate a rapid increase in coolant j

density.

8. Time 0654 to 0724 - he open PORY block valve (0713),

I core boil off, and the turning cff of RC pump 2-B dropped the reactor coolant level so that approxi=ately 4 to 5 feet of the core were uncovered during this time period.

9. Time 0724 - The alarm printer indicated that high-pressure injection of about 1000 gpn was started and continued for about 2 minutus before the operator took control. After this time, EFI flow is uncertain but apparently was at least reduced in flow.

During this period, the core was partially refilled until only 2 to 3 feet of the core was uncovered. The temperature in e.he peripheral incore thermocouples decreased rapidly to the 600-700F range.

10. After 0724 - The water level in the core gradually increased with minor perturbations. This was determined from some i peripheral thermocouples that came back on scale, indicating the temperature was below the saturation level of 600F. Core covering was further substantiated by the return of the SR detector readings to corrected normal shutdown levels. At about 0730 the PORV was closed as determined by the RC pressure increase.

e

2.3 Behavior of Looo Flow Ind: cation Gentile flow tubes are used te measure mass flow in each loop. For solid water conditions, the reactor coolant pumps will act as constant volume pumps with the mass flow changing as the density of the water varies. If steam voids begin to for= in the loop, the reactor coolant pumps vill still act as constant volume pumps with some degraded performance.

The formation of steam voids in the loop reduces the fluid density and consequently the mass flow in the loop. For this two-phase flow condition, the indicated flow will no longer accurately represent the mass flow. However, the indicated flow will follow the trend of a decreasing measured flow with an increasing void fraction. ' Figure 11

is the measured loop flow during the IMI-2 accident. This curve illust: tes the expected behavior of the measured loop fiev during two phase flow conditions with a gradually increasing void fraction.

(

FIGURE 1. LOCA LIMITS POWER SHAPE - 6 FT PEAK 1.8 -

1.6 -

1.4 -

3 1.2 -

2 u.

y 1.0 -

E a- 0. 8 -

E 0.6 -

0.4 0.2 -

0. 0 e i i i e i e i i 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Core Height, Ft

FIGURE 2. SADDLE SHAPE POWER CURVE-UNEQUAL PEAKS 1.6 --

1.4 -

1.2 -

~

% l.0 -

u.

t24 C

I 0.8 -

a.

0.6 -

0.4 0.2 -

0.0 i e i I e I e e i 1 0 1 2 3 4 5 6 7 8 9 10 11 12 Core Height, Ft

FIGJRE 3. SADDLE SHAPE POWER CURVE-EQUAL PEAKS l.6 -

1.4 -

- 1.2 -

0 2

1.0 -

E 0.8 -

n.

0.6 -

0.4 .

0.2 _

0.0 e i i e , , , , ,

O I 2 3 4 5 6 7 8 9 10 11 12 Core Height, Ft

FIGURE 4. LOCA LIMITS POWER SHAPE - 10 FT PEAK l.8 1.6 -

l.4 _

g 1. 2 -

U" g 1. 0 -

a.

-- 0.8 -

"c 0.6 -

0,4

0. 2 -

0.0 i e i e i I i e i 0 1 2 3 4 5 6 7 8 9 10 11 12 tore Height, Ft

n n FIGURE 5. SMALL BREAK POWER SHAPE 1.6 1.4 -

\

1. 2 -

a j 1.0 -

u.

N 30.8 -

?

3 0.6 -

0.4 -

0.2 -

0. 0 e i i i e i e i i i 0 I 2 3 4 5 6 7 8 9 10 11 12 Core Height, Ft

Figure 6 RCS PRESSURE VS CORE EXIT FLUl0 t TEMPERATURE FOR 1400*F CLAD TEMPERATURE LINIT 1300 --

REF. FIGURE 1 1200 -

REF. FIGURE 2 #

u /

1100g REF. FIGURE 3 o" /

6 1000 ,

B 2

$ 900 A j '

t 2' 2

w 800 -

5 700<> REF. FIGURE 5 600 -

500

(

400 ' ' ' ' '

( 730 600 1000 1400 1800 2200 Pressure, psia

Figurt 7 RCS PRESSURE VS CORE EXIT FLUID TEMPERATURE FOR 1800*F CLAD TEMPERATURE LIMIT 1600 REF, Fggggg y

' FIGURE 2 1500 _

Q REF. FIGURE 3 1400 -

b E

2 (1 2 REF. FIGUR5 4

8. I300

,y

.E 2

1200 -

=

su 2

!!00 -

g 1000 -

REF. Fggygg g

()

900 800 i

'700 , ' ' '

( 200 600 1000 1400 1800 gggy Pressure, pgg, 1

Figure B CORE EXIT FLUID TEMPERATURE INDICATION TO Lilili CLAD TEN?ERATL'RE

(

1200 _

1100 -

e T

, CLAD LESS THAN 1800*F

{ 1000 E

[ 900 -

E C

O 800' T

a T CLAD LESS THAN 1400*F 700 -

600 -

500 400 ' , , , p 200 600 1000 1400 1800 2200 Pressure, psia

FIGURE 9. SOURCE R44GE TRACE FOLLOWING f REACTOR TRIP (TYPICAL) 4 10 I

103 D _

a E -

E c.a 2

i ."

102 x

1 ' ' ' ' ' '

10 ,

0 2 4 6 8 10 12 14 Time From Reactor Trip, Hours

(

! 1 2 I

( .

y caec: w::: sm EG ;

, -a gI E aR Er -

1

=

5E r5

~s W5 E e*

1 35

- . cm

ama :# eue -

)~

c}Gdo Aacs ETr. ,, -

w

2 S.N w-Y *

L -

cama u a no _.

I e

b I 5 Io w te 3

E 5W og

- - - - r. . 9'

=(p$ ( sil

~$

w 7 h5 u ed4. TZ wa Eg

.d4. vi atk T990 5

~

. a . sZ

.sa. e, a me gy ea

\:

5

-s~~

89 bf i

9 m acem ,ene us -

~

! I ,

% % o

'.o. -

t e

o

+

==

am S e

M g

,, C ==

= a

( -

w a w

em E E

. w C.:

o p- a-

, e' f/5 6 b

m

== =C

= .M.

E 6 S

~

em b

M e

6 M

iI

~

M b m

e i5 m

M M

D m e am W

== ip e

en W

EG W

D O

e o aus ee e iB S 9 ens

" e

= 99 e.

= at ee s

" Z

{ e ammer O e me

W e g ea so 1 _

e =, , >=

M M

N ,,

e N

e e k k

" I C C

0 0

== ww e o

k %

E E I to es g me m

-. se G S OO em 'b O O'

- e w w .

6 33

1 I i l l E E em

.a es up 6

m ea e O

g m s

- I

- 71 e a 11

~~

L W O g, M y

= a ,i 9 m- -

e q i

i

==

.i .

q W

9

( i _

i i i i i i i i i i a ; i e

EddM .

t 1

Table 1. Su=-arv of FOAM Inout f

Inout paranoter Descriotion Core power 1.2 X ANS at 200 see for 2772 5't Core hydraulics 177 FA ccre with 15 I 15 fuel Axial power shapes 5 shapes (Ref. Figures 1 through 5)

Initial core water level 2 through 10 feet Core pressures 600, 1000, 1400, 1800, 2200 psia Core inlet enthalpy h,,

Radial peaking factor Based on LOCA li=it for m d--

local power f

REFERENCES

(

l. B. M. Dunn, C. D. Morgan, and L. R. Cartin, Multinode Analysis of Core Flooding Line Break for B&W's 2568 MWe Internals Vent Valve Plants. BAW-10064, Babcock & Wilcox, Lynchburg, Virginia, Ocecber 1975.

2. Babcock & Wilcox Revisions to THETAl-B, a Computer Code for Nuclear Reactor Core Therma' Analysis (!N-1445), BAU-10094, Babcock & Wilcox, Lynchburg, Virginia, April 1975.

3. Operating Guidelines for Small Breaks for Oconee 1, 2, 3:

Three Mile Island-1, 2; Crystal River-3; and Rancho Seco, Euergency Operating Specification 69-1106001-00.

4. Operating Guidelines for Small Breaks for Arkansas Nuclear One-1, Emergency Operating Specification 69-1106002-00.

5. Operating Guidelines for Small Breaks for Davis Besse-1, Emergency Operating Specification 69-1106003-00.

e

ATTACIDfENT TO OUESTION 52 SUPPLLvD;T 1, PERT 2 t

Y v -

o ,s s '

7\ w '

t M

{ - N~ I h E, 3 Ia fr; G j/

+

N h 1 l

0 v! o 3 - ..

sg

., r 4 i

/

s e

- - %W ,

e C

\\

i g e s v :l Co.'*. :*

/ g a <

\ s9 SE o #.: k \

Y. I! C[

/ .s .9 i

' - s _

1 e : ss a, s,.

N 3 .

s 3 -

ry A^; '+.- :~

/

\ '_ s

':

e=s ~

.;~s i' c

./ \\asN'x '

i 6

g / > y e r 3

m

! v o

7 3 P/

c j K

i z ri M, ~e/

f,e /g/

j i s i

g ii t.; $

s W'/ %y, A o; 2 \

j N/

Mi .e a l'Q;t,

! -i !!'! 1l !E

.r c~ 2 a q q!;, !r : s.4 30

.v ,, r~ w

ce

/ w ..

li idu,. 3m ..

C .

. . . . . . . . i a '& ~; h;j 3 x /\.s

\

/ .e,

,y

. - - ' \ ,.,,

A

~

c,3

\'

. ). .

E-a s e -I wrt - iari y a

f y,,7 .,,,1,,, *T

.\ g h

/

. . ' w* ff g , 8, *C,

. s 't

' a !**l

j n'l

l Mfl G 77 vm 3%vapn3 3 1 ///. er 3 ., . Op CPT 30838 4491%

% 3tA 0?9avw1 C7:07

,l . ?,>

._ m.I -; .rt . i mi o h

/ . ,. / [

M-"'I y

M, a. ; .,, ,

, 1., . "7,.,',.7..".,W.U,Wd.

p /.4 7/\ \\.

t

,, ,. ,e .b. .

i\ i - ~

., / r$2 s '

.2 -q. .-sa. i -

~' * ,

\ d / c :

/ <

j. /q :

-s

[\ ,/ / ~2

[

i.

\ e :

_' mi a ;

' /\ -

/ wet 3,ns g\ \

.!- .!,'f I 3 2 ag s vcunau 7 1 ,,j// ..

N\ d j E

~

2:? g 7=~--~~."-

y i

stl ? n o .,

j$ '

-f '

2 51 '

.l 3 e 3 8 g }";;;f i - ,:::, -

f sN O 3

/e v

su _ ,

2 y -

/*i

\N\ :

'\/ =

,, <i k Eg,..

s i

hg

. cN U M

.,;e si Y3

~

- - J w -

\ O % 3N e mi 3

/ It a wo s o g N :. _ _ - df $g7e < dd O

\

. . . . . o -

eg - -

J u

O "

b b b e w

O g, -

9 8.j "

5 w

( Tn 2 R ',, -~

E g e c dd d 2u t 3

A I

Am 13

SUPPLEMENT 1, PART 2 f

QUESTION

95. Paragraph 2.1.3.b of NUREG-0578 requires a description of further measures and supporting analyses that will yield more direct in-dication of low reactor coolant level and inadequate core cooling such as reactor vessel water level instrumentation. Section 2.1.1.6.

of the Restart Report does not address further measures (to be im-plemented by January 1,1931), nor does it address the question of ,

reactor vessel water level instrumentation. Provide a conceptual description of what additional measures will be taken to detect inadequate core cooling. Provide an implementation schedule for these changes.

RESPONSE

Babcock 6 W:.lcox is currently evaluating the need for further measures beyond those described in Section 2.1.1.6. that will yield more direct indication of low reactor coolant level and inadequate core cooling.

This evaluation will cover all the inadequate core cooling evaulation cases and is scheduled for completion on December 14, 1979. Babcock 6 Wilcox is now scheduled to provide recommendations to Met-Ed for additional instrumentation (if any) by March 1,1980. It is our intent to review this generic B6W evaluation for applicability on TMI-1. The conceptual descrip-tion of additional measures to detect inadequate core cooling and an imple-mentation schedule for any required changes will not be available until March 31, 1980, at the earliest.

Although the recommendations are not complete at this time, there are several conclusions that can be drawn from the work which has been per-formed to date. First, reactor vessel level instrumentation may be of value in anticipating loss of adequate core cooling when the plant is in the decay heat mode. Second, there is no need for reactor vessel water level to determine inadequate core cooling. The basis for this statement is that reactor vessel level is not used or needed in order for the opera-tor to take necessary corrective action to assure adequate core cooling.

In particular, there is no additional corrective action that can be taken that would not already have been taken in response to a more direct measure of inadequate core cooling; namely, core exit temperature.

(

Am. 13

SUPPLEk!ENT 1. PART 3. QUESTION 1 (Cont'd.)

Calculations to quantita:1vely define what reduction in flow rate was caused by the inclurion of the cavitating venturi are not available.

.However, from the calculations that have been made some qualitative assessments can be made. For one pu=p operations, the cross-connect design with venturis appears to result in a slight improvement in HFI perf ormance. The venturis appear to take the place of the resistance provided by the partially opened high pressure injection valves while providing a more even distribution of flow to each of the four EPI legs.

For, two pu=p operations, the venturis do restric HFI flev slightly with the amoun: of this redue:1on increasing as RCS pressure decreases.

For example, a: an RCS pressure of 600 psig, the venturis will restrict flow to abou: 84% of that achievable without the venturis installed.

However, :his still resul:s in almost 87: more flow than required by the B&W LOCA analysis.

The tes ab stract for the testing to be performed on the HFI sys:e=

prior to return of IMI-l to power operations is contained in See:ics 5.0 of Report GED 0005. The inf ormation concerning test description is contained in this cection of the report except for the basis for the 550 rpm " upper lteit" acceptance criteria. Section 3.4 of the subjec:

report contains the basis for the 550 gym upper limit. When available, the test procedure f or the subjec: testing will be provided.

The testing method described in Section 5.0 of Report GED 0005 is based on g a two point test (i.e. , one test point in the cavitating condition and one test poin: in the non-cavitating condition).

With only one HPI pu=p running, the venturis are predicted to be in cavita-tion when RCS pressure is below 600 psig. Therefore, the first test point will be established at an RCS pressure of less than 600 psig. The acceptance criterir, for the test will be that flow in any oae leg shall not exceed 30%

of the cotal flow to the four HPI legs, and that the total four leg flow is 500 to 550 gpm. It should be noted that it is not important to demonstrate the exact point where cavitation occurs. The acceptance criteria of 500 rpm mini =um flow assures that delivered flow will be greater than that assumed in the ECCS analysis even if cavitation occurs at an RCS pressure greater than 600 psig.

The second test point will be conducted at a pressure of 1200 psig to 1800 psig. This range of pressure ensures that non-cavitating flow will occur through the venturis and that the non-recoverable pressure losses through the venturis will be large. In this manner, a significant discrepancy be-tween assumed and actual losses will be observed by a significant deviation in the observed versus predicted flow races. The acceptance criteria is that the measured flow shall be greater than that assu=ed in the B&W ECCS analysis and that the flow split is no worse than the ECCS 70/30 assumed flow rate.

(

Am. 13

{

SUFFLEMDiT 1, PART 3, OUESTION 1 (Cont'd.)

In addition, the data obtained fron this second test vill be used to verify the convervatism of the GAI computer model of the system. If significantly lower than predicted flows are measured, then an engineering evaluation of the test data qill be performed. The reason for the deviation vill be iden-tified and a determination vbether and for what corrective action is required vill be made.

l C

Am. 13

. > c-

-c-W.[.-q" 1,

.r _ ATTACFF.E;T TO SUPPLDd.DiT 1, PAET 3. OhESTION 1

[_k_, Dh['hhhh SDD 211A ,

t 9

SYSTEM DES!GN DESCFd? TION FOR HIGH PRESCURE INSECTION CROSS-CONNECT THR* T MILE ISLAND UNIT #1 i

cEn m e o S

* = =

a m- --e..- e, .w., ..-

Am. 13 1

ACOO OO27 G

e

TABLE OP CONTOTTS ,

I section Page 2.0 In trod uct ion 1 2.0 Summary 2 3.0 Discussion 3 3.1 Background 3 3.2 Calculational Techniques 3 3.3 Systen Performance Under Cold Leg Break Conditions 3 3.4 Core Flood Line Break 4 3.5 High Pressure Injection Line Break 4 3.6 Normal Plant Operatiora 5 3.7 Transient System operations 5 3.8 Core Cooling Using Only High Pressure Injection 6 4.0 Operator Action 7 5'. 0 Post Modification Testing ;

5.1 Required Test Equipment 8 5.2 Procedure Abstract 8 5.3 Acceptance Criteria 9 6.0 References 10 O

Am. 13

LIST OF FIGURIS .

f Figure No. Title 1 Flow Diagram for Pump "A" 2 Flow Diagram for Pump "C" 3 Simplified System Schematic of HPI Cross Connect LIST OF TABLES Table No. Title 1 High Pressure Injection Flow Requirements 2 High Pressure Injection System Performance With Valves MU-V16 A & B Open and MU-PlA Operating 3 High Pressure Injection System Performance With Valves MU-16 C & D Open and MU-Plc Operating

4 High Pressure Injection System Performance With Two Pumps and All Valves Open 5 High Pressure Injection System Perfor=ance Under High Pressure Injection Line Break Conditions e

Am. 13

1.0 Introduction To Leprove the ability of TMI-1 to withstand the consequences of a small break Loss-of-Coolant Accident (LOCA), a change in the design of the high pressure injection system was developed. The subject change involves cross connecting the "A" and "C" high pressure injettion (RPI) legs and the "B" and "D" high pressure injection legs. The design is shown on CAI flow diagram C-302-661 Rev. IA-3 and detailed on GAI drawing E-304-666 Rev. IA-3. This report describes and presents the results of the flow calculations performed to verify the adequacy of the cross connect design.

As a result this report also represents the system description for the TMI-1 high pressure injection system.

O e

S

(

1-Am. 13

2.0 Su==arv f The criterion established by B&W for the small break analysis requires that 70% of the total flow for one HPI pump be injected into the unbreken legs of the reactor coolant system. This criteria applies to a 277: Mw thermal 177 fuel-assembly plant. For TMI-l with a licensed core power of 2535 Mwt, reference (a) indicates the 70: - 30 criterion can be relaxed in direct proportion to the amount that licensed power is less than 2772 MWt. Therefore for THI-1, the acceptable flow split can be relaxed to 64: - 36%. The above criteria are applicable to all breaks except for those reall breaks which occur as a result of a break in the high pressure injection line between the RCS and the first isolation check valve in the injection leg.

The analysis presented in this report demonstrate that the proposed cross connect at TMI-I c,n meet the B&W ECCS acceptance criteria.

The analysis of this report further demonstrate that if the r=all break LOCA results from a break in the HPI line, sufficient high pressure injection flow will occur through the unbroken HPI lines to satisfy small break ECCS criteria.

Am. 13 G

3.0 Discussion ,

3.1 Background

(

A nechanical piping cross connect has been accepted as a viable long-term solution for additional corrective action on the small break LOCA. The basic. concept is derived from the BSAR 205 plants. The issue and scenario to arrive at the solution are extensive, well documented by references (d) thru (p), and will not be repeated in this report. The proposed design changes to be imple=ented at TMI-1 are detailed by CAI drawings C-302-661 Rev. IA-3 and E-304-666 Rev. IA-3. The following discussion presents the system flow calculations performed in support of the subject design changes.

3.2 Calculational Technioues A simplified sketen of the proposed IMI-I cross connect is shown in Tigure 1 and Figure 2. based on these layouts, a task was initiated to deter =ine the performance of the proposed design under small break LOCA conditions.

Included in this consideration were those small break LOCA's which could result from a HPI line break. .

CAI's "PIPF" computer code was used to model the system. The FIPF Code is described in Topical Report GAI-TR-105NP-A. The subject report has been submitted to the NRC and accepted.

73 System Performance Under Cold Leg Break Conditions The initial performance of the TMI-1 high pressure injection system was evaluated with all RC loops assumed to be at a preasure of 600 psig.

For this analysis the design did not include installation of cavitating venturis. The worst case flow slit occurs with only the "C" cake-up pump running and under these conditions a 68: - 32T flow split was predicted to occur. B&W was asked to evaluate these conditionns and to investigate the possible relaxation of the 70 - 30: criterion based on the lower power level for TMI-1. In reference (a), B&W documented their conclusion that the 68: - 32% flow split was acceptable.

As further assurance of the acceptability of the TMI-I design, B&W was requested to provide the HPI flow assumptions utiliced in their small break analysis code. Reference (b) and (o) transmitted this information.

The criteria of reference (b) and (o) are contained in Table 1 and are based on undegraded pump performance. For the small break ECCS calculations, a factor of 0.9 was applied to subject flows during the first 10 minutes of the transient to account for pump degradation due to wear. The resulting degraded pump performance is also presented in Table 1 and forms the basis for Figure 6.2.59 of reference (c).

As explained in reference (i), the operator action was assumed at 10 minutes to increase the HPI flows to the intact cold legs. This was te be accom-plished by balancing one HPI pu=p to the four delivering lines, thus reducing the total effective resistance to the flow. Therefore, pump flow increased and pump degradation was implicitly taken into account. The resulting flows are presented in Table 1 of reference (o), and are those specified in Table 1 for undegraded pu=p flow.

B&W was also asked to specify the maximum pressure difference which would

( exist between the unbroken cold legs and the cold leg containing the break.

It was reported this value would be less than 4 psig.___

~~ '

Am. 13

Based on the above, GAI has performed calculations for each of the RC pressures specified in reference (b). For conservatism, the shortest leg was assu=ed broken and at a pressure 4 psig lower than the other

( legs. To account for the worst case single failure, only one high pressure injection peap was assumed operating. Undegraded pump perfor-mance was assumed. For reasons discussed in Sections 3.5 cavitating venturi were incorporated into the original cross-connect design.

The sizing of the cavitating venturi was based on pump run-out considera-tions. Run-out flow for the TMI-1 high pressure injection pumps is con-sidered to be slightly greater than 550 gpm since higher flows have not yet been demonstrated by test. The venturi were, therefore, sized to limit flow to 137.5 gym (i.e. one-fourth of 550 gpm) when only a single pump is operating. Specifically, they limit flow to 137.5 gym when the venturi inlet pressure is 813.6 psia. At this flow, the vendor indicates that the non-recoverable pressure losses are 15* of the inlet pressure.

The GAI calculations, therefore, assumed that the navi~"m achievable flow varied as the square root of the absolute inlet pressure and that the non-recoverable pressure losses varied as the square of the flow.

Table 2 and 3 presents the calculated results of the TMI-l high pressure injection system perfor=ance under the above assu=ptions. Table 4 pre-sents the system perfor=ance when both pumps are running. The acceptance criteria was obtained by multiplying the values of reference (b) by 0.7 to account for the 70: -

30 flow split criteria of a 2772 MW plant and by the power ratio of TMI-1 to the generic plant design (i.e. 2535/2772).

As indicated by Tables 2 and 3, the TMI-l cross-connect design with cavi-tating venturi not only meets the above acceptance criteria but also meets the 70/30 flow split criteria for a 2772 Mut plant.

i 3.4 Core Flood Line Break The core flood line break establishes the maximum size acceptable for a cavitating venturi. Under these accident conditions, low RCS pressure occurs and high pressure injection is required. The cavitating venturi should, therefore, limit flow to less than run-out flow of a single pump.

Run-out flow for the DdI-l high pressure injection pumps is considered to be slightly greater 550 gpm since high flows have not yet been demon-strated by test.

As indicated in Section 3.4, the venturi has been sizad based on it=1 ting flow to 137.5 gpm when the inlet pressure is 813.6 psia. Based on this site venturi, calculations performed by GAI indicate that below 600 psig RC system pressure, the venturi will be in cavitation and flow will be limited to 551.5 gpm assuming the highest head pump is in operation.

3.5 High Pressure Injection Line Break One of the small breaks considered in this evaluation was that associated with the break of a high pressure injection line break. Such a break if it were to occur between the RCS cold leg and the first isolation check valve would result in both a small break LOCA and a 32I line break.

Calculations performed on the original design (i.e. the design without cavitating venturi) indicated that the Hyl line break required the operator

(

Am. 13

~ ~ '

to isolate the leg containing the high flow. Such action goes against

~

the operators normal tendancy and therefore presents the opportunity for operator error. As a result , the system design was modified as discussed

, above by the addition of cavitating venturi in each of the HPI legs.

These devices will eliminate the need for the operator to take action which goes against his normal judgement.

Based on the venturi para =eters presented in Section 3.3, the performance of the HPI systcm under high pressure line break conditions were analyzed.

The results of the analysis are presented in Table 5. These results were then transmitted to B&W. Their review and conclusions are documented in reference (c) and reference (q), and indicate that the design performance of the system will be acceptable. Table 5 and Figure 4 su==arize the ac-ceptance criteria contained in B&W references fer this break and demon-strates the acceptable performance predicted for the syste=.

3.6 Normal Plant Operations Due to space limitations at TMI-1, the high pressure injection line cross connect can only be accomplished inside of the reactor buildin;.

The cross connect will, therefore, be installed downstream of the location where under normal operations reactor coolant system make-up was being supplied. This resulted in an unacceptable system design since due to the cross connect, oscillating make-up flow between the "B" and "D" legs could occur and result in high cycle ther=al f atigue failure of the injection nozzles.

To correct this problem the nor=al make-up injection point to the "B" HPI line was relocated as shown in Figure 3. Specifica11*,, the line is to be extended into the reactor building through spare peneiration 323 and g connected into the "B" EPI line downstream of the cross cornect. A check valve will be added to the "B" EPI line to prevent back flow of normal make-up water to the cross connect. The existing containment isolation valve, MU-V18, will provide outside containment isolation for the line.

Inside containment isolation will be acco=plished with a check valve.

With the above changes, the normal system operations of the TMI-I make-up and purification system will remain unchanged.

3.7 Transient System Operations During transients which result in overcooling of the reactor coolant system, pressurizer level decreases and operator action is taken. This action currently consists of opening MU-V16B and startin; a second make-up pump as necessary to restore pressurizer level. This action is requiret because normal make-up control valve, HU-V17, significantly restricts the maximum deliverable flow. The above operator action provides high flow rates without any thermal shocking of the HPI nozzles.

With the installation of the cross connects, continued operator action in L 's manner would result in additional thermal shoc' king of the "D" EPI nozzle. Operations in this matter is not recommended since the stress calculations for the HPI line nozzle allow only 80 cycles of cold water injection to a hot nozzle (a nozzle without continuous flow).

(

Am. 13

As a result, an alternate means for initiating high make-up flow to the "B" HPI leg has been provided by the design. Specifically, bypass valve, MU-V217, is to be installed around MU-V17. Valv.e, MU-V217, which is a

( high pressure injection valve similar to the MU-V16 valves, will be a normally closed, motor operated valve capable of being opened by the operator from the control room. As such it will provide the sane function as the current MU-V16B and allow the operator to deliver a maximue of *50 gpm o f acke-up.-at 1800 psig RC pressure backpressure. Flow will only be through the "B" HPI nozcle and therefore thermal shocking vill not occur.

To provide indication of flow, a flow meter is being installed on the bypass line. The flow meter is a strap-on sonic flow meter manufactured by Controltron Corporation. This flow meter will have a range of 0 cc 500 gpm. The output of the flow devices will be transmitted to the control roem where a meter will be installed to read flow directly.

3.8 Core" Cooling Using Only High Pressure Injections The high pressure injection system provides a back-up means of core cooling in the highly unlikely situation where all secondary system cooling, including auxiliary feedwater is lost. To provide this back-up cooling capability, B&W analysis indicates that one HPI pump capable of injecting 216 gpm at an RCS pressure of 2500 psig is required.

This assumes decay heat following reactor trip are at levels given by ANS 5.1 with a safety factor of 1.0.

The THI-l cross-connect design has been analyced to ensure that the cavitating venturi due not restrict flow b~elow 216 gpm. The results of the GAI analysis indicate that the system will be capable of injecting a minimum of 253.9 gpm at an RCS pressure of 2500 psig. Therefore the back-up method of core cooling using the HPI system will be available. '

Am. 13

4.0 Operator Action .

Following actuation of High Pressure Injection, no specific operator action should be necessary to achieve proper flow and flow split condi-t tions. The on,1y immediate action required by the operator is to verify that at least the minimum level of actuation has been achieved. This can be ac'complished using flow transmitters M"23-DPT 1 thru 4. The criteria for successful actuation is as follows:

1. Total flow must be greater than or equal to that shown in column A of Table 1.

~

2. If both HPI nu=ps are operating, two or more HPI valves must have full open indica

n except that an indication that only valves MU-V16A and 16C or MU-VAud and 16D are open represents an unacceptable two valve combination. The installed flow indicators should be used to confirm the valves have indeed open.

3. If only one pump is actuated, both MU-V16 valves in the train with the operating EPI pump must be full open. The -installed flow indi-cation should be used to confirm the valves have indeed opened.

If the above criteria are met then the operator is assured of compliance with the ECCS acceptance criteria. If the above criteria is not met then multiple failures or other reasons have prevented proper automatic actu-ation. In this event the operator must diagnosis the problem and take corrective action (i.e. start idle HPI pumps open closed EU-V16 valves, etc.).

s h a Oe e

(

Am. 1:3

5.0 Post Modification Testing The following tests of the high pressure cross-connect design are reco=menced

( prior to the return of TMI-1 to power operations.

5.1 Reauired Test Equipment

1. Wide range RC Pressure - RC3A-PT3, RC3A-PT4 and RC3B-PT3; range O te 2500 psig.

2. HPI Flow - MU23 - DPT1, 2, 3 & 4; range O to 500 gpm.

3. Temporary HPI Injection Leg Flow Meters - FI - 1, 2, 3, 4 (See Figure 3) consisting of:

a. Flow Display Computer: Controltron P/N 24IN - 2.5SS.375; Range O to 330 gym.

b. Multiplexer - Manual Selection 4 channels Controltron P/N 242-10.

c. Transducer (4): Controltron P/N 240N - 2.5SS.375.

d. Cables (4): 25 ft each Controltron P/N 242 25.

4. Make-up Pump Discharge Pressure: MU22 PII, MU22 PI2, MU22 PI3; range O to 5000 psig.

5. Make-up Pump Suction Pressure: PX-412, PI-413, PX-414 t

5.2 Procedure Abstract d

1. Establish RC pressure at less than 600 psig and take adequate precautions for overpressure protection.

2. Ensure minimum recire, line is open on pump to be operated.

3. Start MU-PIA or MU-PIS and slowly open MU-V16A and MU-V16B. Maintain balanced flow.

4. Shut MU-V36 or MU-V37 to secure pump recire. prior to reaching 400 gpm total flow. Open MU-V16A/B to full open position.

NOTE The cavitating venturis were designed to prevent inadvertent pump run-out and cavitation without the need for throttling MU-V16 A thru D.,

Maximum expected flow below 600 psig RC pressure is 552 gpm. If necessary set MU-V16A/B to limit flow to less than 550 gpm. Reduce RC pressure to verify venturies are in cavitation as indicated by flow remaining constant.

CAUTION Extended operations of HPI pumps under cavitating conditions should be avoided. Terminate test as soon as cavitation is observed or unstable conditions are observed. Do not allow pump discharge head to

( drop below 450 ft. Re:ord finv at which unstable conditions are re ach ed .

Am. 13

CAUTION Do not operate any other make-up pump while minimum flow recire, f

~ line is secured. Do not close MU-V16A and B while recire. line is isolated and pump is running.

5. Record flow through MU23-DPTl & 2 and through each sonic flow indicator.

Record pump suction and discharge pressure. Observe venturis and cross-connects for signs of any unacceptable vibration.

6. Reduce flow to less than 400 gpm and open MU-V36 and MU-V37.

7. Start MU-PIC. Open MU-V16C and MU-V16D and close MU-V16A and MU-V163 as necessary to maintain flow in each of the four legs.

Stop MU-Plc.

8. Re pea t steps 3 and 4 for valves MU-V16C and D and MU-PIC. If necessary use MU-V64C for recire. isolation in lieu of MU-V36 and MU-V37.

9. Record flow through MU23-DPT3 & 4 and through each of the sonic flow meters. Record pu=p suction and discharge pressures.

10. Open or verify open MU-V36, MU-V37, and MU-V64C.

11. Secure testing uniel RCS pressure is raised to 1200, 1500, 1600 or 1800 psig. Testing at 1200 psig is preferred, however, testing at one of the other pressures is acceptable.

12. At 1200, 1500, 1600, or 1800 RCS pressure, start MU-PlA and or MU-PlB if not already started and open MU-V16A and MU-V16B.

13. Shut MU-V36 or MU-V37. Observe caution of step 4.

14 Record flow through MU23-DFT 1 & 2 and through each sonic flow indi-cator. Record pump suction and discharge pressure.

15. Open or verify open MU-V36 and MU-737. Close MU-Vl6A and MU-716B.

5.3 Acceptance Criteria Step 5 - Flow in any leg shall not exceed 30% of total flow. Total flow of all four legs shall be greater than 500 gpm and less than 550 gpm. No unacceptable pipe vibration is allowed.

Step 9 - Flow in any leg shall not exceed 30% of total flow. Total flow of all four legs shall be greater than 500 gpm and less than 550 gpm. No unacceptable pipe vibration is allowed.

I Step 14 - Flow in any leg shall not exceed 30% of total flow. Total flow of all four legs shall be greater than that specified in Table 1 Colu=n A. In addition, if total flow of all four legs is less than 95% of the predicted value based on Table 2, an engineering evaluation of the data shall be requested to verify proper system performance.

(

_9_ .

Am. 13

6.0 References (a) B&W (G. T. Fairburn) letter TMI-79-17 dated February 5,1979 letter

, to R. M. Klingaman, RE: Three Mile Island Nuclear Station - Unit 1 Small Break Analysis.

(b)' B&W (C. T. Fairburn) letter TMI-79-74 dated May 21, 1979 to J. F.

Fritzen, RE: HPI System Flow.

(c) B&W Report " Evaluation of Transient Behavior and Scall Reactor Coolar.t System Breaks in the 177 Fuel Assembly Plant", dated May 7, 1979, Volume I.

(d) Met-Ed letter GQL #0714 dated April 17, 1978 from J. G. Herbein to R. W. Reed / S. A. Varga (NRC).

(e) GPUSC memorandum NF-279 dated May 2, 1978 from G. R. Bond to V. E.

Potts.

(f) B&W letter dated May 1,1978 from J. T. Janis to W. E. Potts, Re:

TMI-1 Small Break Analysis.

(g) Met-Ed letter GQL #0854 dated May 5, 1978 fre= J. C. Herbein to S. A.

Varga (NRC).

(h) Met-Ed letter GQL #0907 dated May 11, 1978 from J. G. Herbein to S. A. Varga (NRC).

(i) B&W letter dated July 18, 1978 from James H. Taylor to S. A. Varga (NRC).

(j) Met-Ed letter GQL #1254 dated July 24, 1978 from J. G. Herbein te R. W. Reid/ S. A. Varga (NRC).

(k) Met-Ed letter GQL #1619 dated November 21,1978 fro: J. G. Herbein to R. W. Reid (NRC).

(1) Meeting Report dated December 12, 1978, Re: Licensee's Revised Proposed Modification to Eliminate Reliance on Prompt Operator Action Following A Small Break LOCA.

(m) Met-Ed letter GQL #2031 dated December 21, 1978 from J. G. Herbein to R. W. Reid (NRC) .

(n) Met-Ed letter GQL #2072 dated December 29, 1978 from J. G. Herbein to R. W. Reid (NRC).

(o) B&W (G. T. Fairburn) letter TMI-79-208 - dated December 21, 1979 to D. G. Slear, RE: Ac ces sement of HPI System to Mitigate Small LOCA's.

(p) Met-Ed'/GPUSC Report In Response to NRC Staff, " Recommended Require-ments for Restart of "hree Mile Island Nuclear Station Unit 1",

dated September 7,1979 as sucolemented and amended (q) B&W (G. T. Fairburn) letter TMI-80-036 dated February 25, 1980 to I

( D. G. Sleak, Re: Draft Input for SDD-211A, Rev. 1.

Am. 13

FIGURE 1

~

Flow Diagra: for Pi=; "A" .

\ ' ~

d -

< ogBg Leg "A" Leg "C" "D" i lea"B" Le'h A s .

L21=301.3 ft. L22=237.L5 ft.' L23-196.19'ft.' L2h=120.87 ft. -

FZ-38h / TI-386 )( E-385 ) ( FZ-3ST )(

i L11*=159 98 ft. eq. -.U .M m-vi6A m-v.163 MM L12*=148.71 ft. eq.

d = 2.125 in. a '"

d= 2.125 in. ~

(

\ [ m23 m23 ) (

s i \ FE1 FE2 J

--r ~

> RCP 3eal Injection L1=93.6 ft. eq. -

d1=3.62h in. ,

W -V7hA L1A=121 98 ft. eq.

c.1A=2.62h in.

m,.v73A -

8 -

W-P1A . .. ,

(

cxclusive of valve enttivalent longth Am. 13

e

FIGUE2 2 < .

s

. . . s Flow Dias n= for Pu=p "C" g : .

.o -

t' D -

k g "A" k g "C" Le "B" Leg"D"

~

g g - '. ' . ' . .?

A n s I \

.h ' . . . . . :.

L21=353 75 ft: L22=185 ft.* L23=2L9.69 ft'. Idh=67.37 ft.' ' ' :. - j[ .

F2-3% ) n,-386 }(

FE-385 }( FE-387) _'

4 L11*=233.6 ft. eq.

"2 1*9

~ .. a e-W-vlsc- m -vl6D M *

'9" d = 2.125 in. ^ , d = 2.125 in.

)(MU23 MU23)

I R3 Ek I

--r %

L146.52 ft. eq.

d1=3.624 in. "

MU-V7kC .

L1A=121.98 ft. eq. . .-

d1A=2.62h iii. . .

W-V73v .-ff2 N ".lY.. ..

. y.;;f-

. i.

e'

, MU-Plc. ,, ,

~

t .

"CXClu31Ve Of valve equivalent length Am. 13 ' '"

, ','?

y %

ef".'.

g

^

"B "C

y > > >

9 u _

4 I N M x M X X X

'l f[ 3 F s

( , A f t

c e

Q I

n \

vi n J o

C s k N I M N .

s o .

r .

C , i B.

I B.

B x E n g. R t c P

I A

B P3 e2 - C D I

f @w6 1 6

1 g4 61 glis 6 1

o . V V

. V . V c S. '

D. S.

i E E E t

i v

' 31 32 33 27 I 3h

'c uP uP 2T P

2I S HD HD HD uF HD m g e l C t

s r y I P

y 8 -

S

u E ." H d n e : o i

f i i

t

l c (

/ i I

p l e n P aj i

C en x

i S BSI s'

7 7 B x

1 1 I V 2 P V

_ H u

A \

) h

- g x

6 A

x l P

u H

-. A g- C

Table 1 High Pressure Injection Flow Requirements Column A Column B RC Pressure, psig Pump Flow Pump Flow Undegraded, gn Degraded , ge 0 550 495.0 600 500 450.0 1200 437.1 393.4 1500 404.3 363.9 1600 391.4 352.3 1800 -

364.3 327.9 2400 260 234.0 2500 216.0 194.4 (1) Only for small break conditions other than a HPI line break. See reference (o) for flow assumed under HPI line break conditions.

Am. 13

- -~

e TABLE 2 IIPI System Performance with Valves HU-V16 A & B Open and HU-PIA Operating RC Pressure in Psig ilPI Flow in gpm Flow in Unbroken Leg g gpm Calculated Seal Calculated Required (IT Total Flow Loop "A" Loop "B" Loop "C" Loop "D" Leg "A" Leg "B" Leg "C" Leg "D" Injection flinimum ( 647.) gpm gpm (Three Leg Total) 0 0 0 0 127.8 127.0 126.8 126.0 34.3 381.6 352.1 541.9 500 500 500 496 130.3 130.6 128.5 128.8 21.0 389.4 539.2 600 600 600 596 125.8 131.9 129.1 130.7 18.3 386.8 320.1 535.8 1200 1200 I200 1l96 110.7 114.8 113.6 121.0 16.3 339.1 279.8 476.4 1500 1500 1500 1496 101.9 105.6 104.5 111.8 15.1 312.0 258.8 438.9 1600 1600 1600 1596 98.5 102.2 101.2 108.3 14.7 301.9 250.6 424.9 1800 1800 1800 1796 91.5 94.8 93.9 100.9 13.7 280.2 233.2 394.8 2400 2400 2400 2396 64.9 67.1 66.8 73.4 10.2 198.9 166.4 282.4 2500 2500 2500 2496 59.1 60.9 60.7 67.3 9.5 180.7 138.3 257.5 (1) Refer to Section 2.0 for basis. Am. 13

R

TABI.E 3 IIPI Systeri Performance with Valves HU-V16 C & D Open and HU-Plc Operating s

RC Pressure in Psig IIPI Flow in gpm Flow in Unbroken Legen{pm Calculated Seal Calculated Required (l) Total Flow loop "A" Loop "B" loop "C" Loop "D" Leg "A" Leg "B" Leg "C" Leg "D" Injection Hinimum (64%) gpm gpm 0 0 0 0 132.8 132.9 132.8 132.7 N/A 398.5 352.1 531.2 600 600 600 596 132.8 132.9 132.8 132.7 N/A 398.5 320.1 531.2 1200 1200 1200 1196 Ii1.4 114.6 119.4 126.6 N/A 34 5. 4 279.8 4 72 1500 1500 1500 1496 102.5 105.4 109.9 116.9 N/A 317.8 258.8 434.7 1600 1600 1600 1596, 99.2 102.0 106.4 113.2 N/A 307.6 250.6 420.8 1800 1800 1800 1796 92.1 94.6 98.7 105.5 N/A 285.4 233.2 390.9 2400 2400 2400 2396 65.4 66.8 70.3 76.7 N/A 202.5 166.4 279.2 2500 2500 2500 2500 59.9 61.9 64.5 67.6 N/A 186.3 138.3 253.9

\

(1) Refer to Section 2.0 for basis. Am. 13

m m

TABLE 4 IIPI System Performance With Two Pumps Operating and All valves Open RC Pressure in Psig ilPI Flow in gpm Flow in Unbroken I.e6s, gpm Calculated Seal Calculated Required (l[ Yotal Flow Loop "A" Loop "B" Loop "C" loop "D" Leg "A" Leg "3" Leg "C" Leg "D" Injection Hinimum (64%) gpm gym 0 0 0 0 199.1 199.0 199.1 199.0 51.8 597.1 352.1 848 600 600 600 600 199.4 199.7 199.4 199.7 42.3 598.5 320.1 640.5 1200 1200 1200 1200 201.2 201.0 200.9 201.7 29.8 603.1 279.8 834.6 1500 1500 1500 1500 188.4 196.3 197.4 204.9 23.2 582.1 258.8 810.2 1600 1600 1600 1600 181.9 189.3 190.5 200.3 22.5 561.7 250.7 784.5

-1800 1800 1800 1800 169 175.9 177.1 186.2 20.9 522 233.2 729.1 2400 2400 2400 2400 119.7 124.7 125.6 132.1 15.2 370 166.4 517.3 2500 2500 2500 2500 108.6 113.2 114.0 120.1 13.9 335.8 138.2 469.8 (1) Refer to Section 2.0 for basis. Am. 13

.~

TABLE 5 *

, ilPI System Performance Under IIPI Line Break Conditions s

Operating RC Pressure in Psig IIPI Flow in gpm Hakeup Flow in Unbroken LegsA6pm Calculated Pump Loop "A" Loop "B" Loop "C" Loop "D" Calculated Required (l),T2) Total Flow Leg "A" Leg "B" Leg "C" Leg "D" Hinimum (64%)(3) gpm C 0 0 0 0 132.8 132.9 132.8 132.7 398.5 320/336.5 531.2 C 600 600 600 0 132.8 132.9 132.8 132.7 398.5 268.6/305 531.2 C 1200 1200 1200 0 103.1 89.8 110.5 172.0 303.4 195.4/264.7 475.4 C 1500 1500 1500 0 90.1 64.4 96.6 188.9 251.1 154.6/N/A 440.0 C 1600 1600 1600 0 85.4 54.4 91.6 195.3 231.4 137.2/N/A 426.7 C 1800 1800 1800 0 76.2 33.0 81.9 206.4 191.1 102.9/N/A 397.5 C 2400 2400 2400 0 28.4 0.0 30.7 233.9 59.1 N/A/N/A 293.0 (1) Refer to Section 2.0 for basis of 64% flow split.

(2) First value is required flow during first 20 minutes. Secon<1 value is required flow a f ter 20 minutes.

(3) Acceptance criteria are documented in reference (o).

Am. 13

.~. -

SUPPLE'ENT 1, PART 3 ATTACHMENT TO QUESTION 1 i

I

Babcock &Wilcox p ,e. oen.,, son c,cu j P.O. Box: 250. Lyn:nourg. u 24505 Teiepnene: (5o4)354 5111 February 25, 1980 TMI-80-036 Mr. D. G. Slear (3)

TMI-1 Project Engineering tianager GPU Service Corporation 100 Interpace Parkway Parsippany, NJ 07054

Subject:

Draft Input for SDD-211A Rev. 1

Dear Mr. Slear:

Per the request of Mr. J. F. Fritzen, we have prepared the attached draft input for the Met Ed System Description No. SDD-211A Rev. 1. Our inputs include a rewrite of Section 3.3 and more details on the HPI line break analysis discussed in our letter TMI-79-208 dated 12/21/79.

If you have any questions or require more information, please advise.

Very truly yours, AJ&

G. T. Fairburn Service M6 nager GTF/cw cc: J. G. Herbein -

L. L. Lawyer G. P. Miller J. J. Colitz J. F. Fritzen R. W. Keaten - GPUSC F. R. Faist J. C. Lewis - Phila. Sales -

(

The Babcock & Wilcox Company / Established 1867

( HPI LINE BPIAE ANALYSIS An analysis of the HPI line break has been performed for the cross-connected HPI systen of the Midland plants, using the bloudoun code CRAFT 2. The accident

. vas analy:cd following the s=all break model described in the letter report J.H. Taylor (35L') to S.A. Varga (NRC), dated July 1S, 197S(i) . .However, the HPI flows utili:ed in this analysis were modified, relative to that in the letter report, to reflect the HPI syste= perfor=ance for an HPI line break.

The single failure assu=ed in the analysis is a failure of one of the e=ergency diesels. his results in the loss of one train of all ECCS equipment and thus minimizes the injectica of ECCS fluid. Because the syste= is cross-connected, the location of the EPI line break is not that i=portant. However, in order to further =ini=ize the injection flow, the break is assu=ed to occur in the active EPI train. This minimizes the line losses to the break, thereby in-creasing the EPI flow that bypasses the RCS and =ini=i:es the delivered HPI flow.

In modeling 'the EPI line break, the break was assu=ed to occur in the active HPI line between the last check valve and the HPI nostle. Due to the ther=al sleeve in the HPI nos:le, the break area is limited to 0.0225 ft2 The HPI flows utilized reflect the effect of the asy==etric discharge pressures on the amount of ficw injected into the RCS. Furthermore, the EFI flows were reduced by 10:: for conservatis=. At 1260 seconds, the operator was assu=ed to isolate the affected HPI line. This time is based on a 20 minute, time delay for op- ,

erator action after receipt of the ESFAS signal (60 seconds).

The results of.the calculation are su==arized on Table 1. The core pressure transient is depicted on Figure 1. During this accident, the core never un-covers; the mini =u= inner vessel mixture height reached is 19.23 feet above the bottom of the core. Since the core never uncovers, the cladding te=pera-ture re=ains within a few degrees of the fluid te=perature, no cladding rup-tures and no metal-water reaction occurs. At 3310 seconds, the HPI flow into the RCS is in excess of the boil-off and long term cooling is established.

Thus, all the criteria of 10 CFR 50.46 are met. .

( .

g.8 O

I e

e

[

APPLIC/GILITY OF TIE MIDLAND HPI LINE ET2AK ANALYSIS to nrT-1 Figure 1, discussed in the previous section, shows that the RCS depressuri:es, initially af ter the opening of the break, and then stabilizes at 200 seconds and at approxi=ately 20 psi above the SG secondary pressure. The slepe of the curve does not change appreciably with the isolatien of the broken HPI line at 1260 sec, which results in an increase of the flow delivered to the RCS. Thus, the pressure in the primary side is controlled r ainly by the energy removed by the steam genera, tor, and, were the analysis performed for the D'I-l plant, a si=ilar transient pressure history would result.

In the TMI-1 plant, cavitating venturis have been installed in the HFI lines to restrict the flow through the broken HPI line, thus increasing the arount ,

fluid available for core cooling and thus eli=inating operator action at 20 minutes to isolate the EPI broken line.

Figure 2 shows that the high pressure injection flows delivered to the intact cold legs versus time available in the TMI-I plant are greater throughout the transient than the correspending Midland flows used in the analysis. This will allow the syste= to attain long tern cooling at approximately 1000 seconds before the ti=e calculated by 3&*J for the Midland HPI line break (see Table 1)e Therefore, confor=ance with the criteria of 10 CFR 50.46 is ensured.

O e

o e

e q 9

e

9 e

e G

e

(

.- . o. .

e m am m -

e-I Tcbic 1. Sequence of Events for Midland HPI Line Break Ansivsis ,

Item Tisc, s Break occurs 0

.s Reactor trip on low RCS pressure 20 ESF/S trip 60 Operator isolates broken HPI line 1260 Long tem cooling established 3310 e

9 9

e a

4 e

4 G, . %. . .

c.-@. . .. . - .

4.* p:

'$?[h}f.-.'

9 ; .. . . . . .- ., 1.. -

ennr.+

Q'g r4., . .::m, = ,.. ._

[ n -n . ~ *. w.

z_ "gD. - . h*,5.+ 4[,;7*g, ...,.".' . . -

~ CGFMT# p d_- ** ,- *

^51 A *'*:s *s d.'

f -Q4. .ki-.,:.

r.

w.r., x ._. ~ . ."

n :. .

Ap=>. 4.u . .: ~- W.a t.

2a.

'.~**ti'. .* .. . ..

.cw* . . . . . .

(g.rQ n- .'e g.i 'k 't R42" ".*, ' . *.

y;hMJ.Z I.%.y Tri.@-

b .M ,

~

y=o *b7Vihbt y a . - . .

f ~.tx - --

-- l - -- - - * *

.. t

...,.........i.. . .' F.iGu.8E -

1- -

7 _. . ..._.; - . . . . . - . .

. . .~. ~_.. . .

.._.e. w .

..._L._-_-- :

_.._. R.6_5 . .P.R u sa , u. .Rz

. ...cv ..... ___._

.. I . j M ,c -...

p..

__. H .P.L..-

- - c. . _.. .2.0.S.5_._.G

_ . O AIU: c. .C ..

-p g:3 p y _ . , y 9 ...-

l . _;

. . _ _ . +- '

^

. 19D0 ' ' ,

t 0

y. , ,

-..-..9 j

e ,

o r

--] p J. --.

a e --

( g w , ,

. j l i , ,

i 1730- -

, . . a -

, , . J l , . . .

g?$ . s n- ,

4 i ,

i

g. --

, 4, . ._ _.-

1500 -

n'

. V ._

? N .. y y '

b.

s'. 7 . . . . i ... _ .

=r - .. h_ . .

'I N -[

. l l N

d

.x. ,

3 ,__

-s sw vos . .' i , ,

i i

g i

I . ,

<g -

, . i a f

_ Q) i i

'y 4 ... ' ' ' '

~

d r.20- 0- , l ,' l --

_l f '

. .. . .l

.- ..----4

~

, u, ,. . . .

q,

.f.._ ' '

, , , , , f.f.o. .g

~. ~. . .. -

i

- I

,c.-._. _

t04D .

.......... a. .

3-. .

-.,..%.... 4 . ,,_ .._., -+_-

. .y- -- .

i. .IC ..g.,ta . . s3 4 . .g

.o g

.o s

1, j az, .on . . - s. 7 -: :-

(

~-

n,- r;-.,.

_ Cay Q, y jag .. r ... : .

..l - '

W8'"-W'. . W .D-uD .

T I

n G.v R'E. L .

y.. . . _ . . - _ _ . _ _ .

H P.T. L \ W F BR.'E. A %

9 M M_% 15 t_

i .

H 9_I. 'r tow .

To ...NTR c;T Ca_b LT_= crs

/ _ . . . . . . ._

. . n o. .

..j.

.a_ ...._.._._.. . . . . . _ .. . = _ . . . . _ _ . . . _ . . . _ _ . _ _ _ _ _ , . . , _ _ _ _ . . _ _ _ . . . . .

n, . . .

n . . . _ .. _. ...._.._ _._ _ .. .__ _ _ . _ . .__ _ _ . . . _ _ . . . ... _.

r.._._.__ _ . . _ . _ . . _ _ _ _ _ _ _ . _ . . ._ _ . _ _ _ . _.._ __.. .. _ _ .. _ .._.

,1 ... _ _ _ . . . ._ _ _ . _ . . _ . _ _ . . . . _ _ . . _ _ _ _ _ . _._...

. m _ __.. .. . _

._j. .... s . . g.

i .. . . _ . .

.r e . . . .. . .

._,.___....._......-M3.

, t.. .r.

. . _ . i

4. . . _ . . . . . . . -I

.i__ _

. c., . . _ . . . . . . _ . _ _ _ . _ _ . . _ _ . _ _ . _ . . . _ . . . .

_,..Q . . _ _ _ . . . _ . _ . . _ . . . _ . . . . . .

= ._

_ . _ . . _ . . _ = . _ . _ . . . . . _ _ ._._... ._.

r_.. .._.o.. ..........,

_u__..___...r...._......_... _ ._

__ .. .. 1

_ a __. _.

, ). . ; .__r ._

= . . . . . . . _ _ . . _ .

. -r .

=3 _.

7.. . _ . . . .

a . __ _- _

._ , _. . ._ ._=__ _ _ ,

. . . . _ _ 1 .

......__t.__.- _:

_ _,._.._.a ._ .

g .

. . .. ... . . . . . ... g .. . .. . ..

....y *.

4 o 50.3 loco 15 0 o .2o00 :: - .3

.g . ...

('e' D. 2 h

SUPPLEMENT 1, PART 3

{

QUESTION

11. The long-term requirement of IE Bulletin 79-05C requires the B&W licensees to submit a design which will assure automatic tripping of the operating reactor coolant pumps (RCPs) under all circumstances in which this action may be needed. It has been shown through analy-sis that this trip is needed for a certain spectrum of small break LOCAs.

Prior to final design acceptability, the following conditions must be satisfied:

a. Characteristic curves for RCP current / power versus void fraction must be fully demonstrated and documented based upon existing test data and supplemented as necessary with confirmatory data obtained from future tests such as LOFT, full scale testing, etc.;

b. Justification for the RCP current / power setpoint must be shown; and,

c. Satisfactory responses to the following must be received.

RESPONSE

2ne complexity and schedule of responding to items (a) and (b) above and the desire to complete this modification in a timely u nner have lead us to re-consider the selection of RCP power as a trip parameter. We are currently considering the following options:

1. RCP trip on EPI with coincident low saturation margin.

2. RCP trip on low saturation =argin only.

3. RCP trip on HPI with coincident RCP power just below nor=al RCP power fluctuation.

When one of the above options has been selected, appropriate infor=ation will be provided in response to items (a) through (c) of this question.

(

Am. 13 4

SUPPLEMENT 1, PART 3

(

OUESTION

12. By letters dated August 31, 1979, each B&W operating plant licessee indicated a general endorsement of B&W's generic report BAW-1564,

" Integrated Control System Reliability Analysis."

Our joint review of this reoort with Dak Ridge National Laboratory has progressed sufficiently to assure ourselves that the recom-menchtions that the report offers with regard to potential areas of improvement in ICS reliability are reasonable. Therefore, we request that you address these recorTnendations and discuss your followup action plans in this matter. Responses to the following items must be provided.

As part of the continuing review of this report, additional areas may be highlighted as requiring improvement. In that event, we will provide additional requests in these specific areas as necessary.

RESPONSE TO 12.a.1 See Supplement 1, Part 2 response to Questicq 38.

RESPONSE TO 12.a.2 To be submitted at a later date (March 7,1980)

RESPONSE TO 12.a.3(a)

In the past, feedwater oscillations have occurred during two modes of operation:

1. Transition from startup to main feedwater flow control.

2. At power levels less than full power (60 - 75%) due to coupling with thg heater drain system.

The problems during feedwater control transition occur due to leakage through the main feedwater control valve while the valve is closed (nct abnormal for a modulating valve). As the startup valve reaches a predetemined percentage open, the main block valve opens to pemit transition to main valve control.

Leakage results in excessive flow causing the startup valve to close, resulting in reclosure of the main block valve. This problem has been corrected by rechecking startup valve limit switches each refueling outage.

Also, some operators make the transition with the startup valves under manual control, which is acceptable.

Oscillations due to coupling with the heater drain system have been minimized by system tuning. This tuning has eliminated oscillations at full power.

However, changes in system dynamics at reduced power result in some oscillations in the range of 60 - 75: power. These oscillations are not a problem during

(

12.a.3(a), Succlement 1. Part 3 Continued

' power reductions due to the short time that the plant is kept in the affected range. During startup with power holds in the range of 60 -

75t, it may be necessary to place reactor controls in Hand to prevent un-necessary cycling of control rod drives. Other actions taken to reduce oscillations have included installation of hydraulic snuboers on some heater drain system control valves, and tuning of the level controllers on the heaters and on the 6th stage drain collection tank.

Other recurring actions taken to reduce feedwater system problems include precycling of ma1n, startuw and block valves prior to plant startup to prevent sticking, and overhaul of one main and one startup feecwater valve each refueling outage.

RESPONSE TO 12.a.3(b)

The general operating philosophy for the ICS is to maintain all control stations in the automatic mode during steady state and transient operation.

The operator may intervene whenever he juoges that system opera

ion is aonormal, or is inacequate to prevent exceeding reactor trip setpoints.

RESPONSE TO 12.a.3(c)

Procedures used by the operator to perform the operation described above are as follows:

1. Plant emergency procedures address symptoms of abnormal system behavior and . instruct the operator to intervene to achieve the desired result.

2. Operating procedures for specific evolutions, such as plant startup and shutdown, address manipulation of ICS controls. These procedures allow operator intervention if conditions appear abnormal.

3. The ICS operating procedure gives details of how and when to manipulate control s. Specific guidance is given for operating subsystems under manual control. , Again, operator intervention is pemitted if conditions appear abnonnal .

RESPONSE TO 12.a.3(d)

Training covering the operation of the Integrated Control System was covered during the " Operator Accelerated Retraining Program (OARP)". Additional training covering the new procedure guidelines for the operation of the ICS in hand (Procedure 1105-4, Appendix II) has been scheduled.

RESPONSE TO 12.b.1 To be submitted at a later date (March 7,1980)

RESPONSE TO 12.b.2 To be submitted at a later date (March 7,1980)

RESPONSE TO 12.b.3

(

To be submitted at a later date (March 7,1980)

SUPPLEMENT 1, PART 3 I

QUESTION

13. On page 6 of the Commission Order of August 9,1979, Item 4 states:

"The licensee shall demonstrate that decontamination and/or restoration operations at TMI-2 will not af fect saf e oper -

tions at IMI."

In addition to information already provided, you should address this requirenent with respect to systems and operations other than those directly connected with vaste management and affluent monitoring.

RESPONSE

Sections 7.2 and 7.4 and Supplement 1, Part 2, Responses to Questions 52, 54, and 55 describe the features which prevent adverse interaction between TMI-1 and THI-2 for all areas except industrial security, management, auxilliary steam / boiler,and fire protection. Management separation is presented in Section 5.0, industrial security has been separated as des-cribed in the NRC approved security plan, and fire protection, except for the yard loop, is separate for each unit. The yard loop supplies fire water to either IMI-l or TMI-2 and presents no potential for adverse in-teraction. The auxilliary steam / boiler interface is currently under re-view to determine what action, if any, is necessary to prevent adverse interaction.

(

Am. 13

SUPPLEMENT 1, PART 3

(

AUTOMATIC OPERATION OF PORV BLOCK VALVE EMCLOSURE 2 OUESTION 1 PAGE 2 RESPONSE (Cont'd.)

It should also be noted that the Small Break Operating Guidelines specify the opening of the PORV. In many instances, therefore, the operator would be instructed to override this feature. Furthermore, NUREG-0565 provides for a reasoned approach to determining the need for this proposed modifi-cation. First, the probability of small breaks via the PORV and the pro-bability of PORV actuation as a result of Anticipated Operational Occur-rence ( A00's) is determined in light of the modificatior, required by NUREG-0578. Then based on the results of these studies, an appropriate decision can be made as to whether the PORV block valve automatic closure is needed.

Am. 13

. . _ - _ - ~ _ . - , , _ , , . - ,- -- ~.

10.4 Transient Analysis and Procedures for Management of Small Breaks

' As part of the NRC Order of Aug it 9,1979, long term actions recem-mended by the Director of Nuclear Reactor Regulation were included.

Item two of that list recommended that the Licensee:

"... give continued attention to transient analysis and procedures for management of small breaks by a formal program set up to assure timely action of these matters;"

As part of the development of the TMI Generation Group, a new section was established as part of the Systems Engineering Department. This section, known as the Plant Analysis Section, is charged with responsibility for a continuing review of plant performance. This I will include an on-going technical evaluation of the overall plant (

performance as well as the performance of major systems and components.

The Plant Analysis Section is also charged with the responsibilities of reviewing key information from other nuclear plants. This infor- ;

I mation will be obtained from the review of Licensee Event Reports,

'~

general industry survey information, industry contacts, regulatory i body documents, owner's group activities, and standards committees. i Several of these sources of infonnation will be provided to the l Plant Analysis Section by other functional sections within the TMI J

- Generation Group. The overall industry approach to the interchange ,

of such information is currently under development and is expected j to utilize organi:ations such as the Nuclear Safety Analysis Center and the Institute for Nuclear Power Operations. GPU fully intends to participate in these information interchange activities, when developed.

Recommendations resulting from these reviews may be used to modify equipment design, operating and maintenance methods, operator training :

programs, procedures, or other aspects of plant operation.

Additionally, two other sections with the Systems Engineering Department l provide more speciali:ed analytical functions that, on a continuing ,

basis, improve understanding of transient analysis and procedures for j the management of small breaks.

The Control and Safety Analysis Section has the in-house capability to perform analysis of the type shown in Section 8 of this report.

+. m 2 They also contract for and review the results of such work perfonned ,

'by the'NSS supplier or other consultants. The Control and Safety 7" .

.,g,

- Analysis Section has been deeply involved in analysis of programs 7 ,e 9 ee :for small break LOCA operator guidelines and anticipated transient

, , - operator guidelines. This work is closely associated with and has had a direct input to operator training programs and procedures.

51 h s' w 4;.

] N -.;The Nuclear Analysis Section performs or reviews fuel and coreg . re w

evaluations. This work is interfaced with the activities in the

- f.

J,Q Control and Safety Analysis Section .

Mn ,

mw. 7*y*

e ~

, - g , p +- w. ,.

[.. ].Gy%jk -

Am. 13 S;p.f . . - c_ '

-. ...# 10-8 ,

W & h@c.humy" ~

, s -+

ML19290E013

Contents

Text

5.6 REFERENCES

SUMMARY

RESPONSE

RESPONSE

RESPONSE

RESPONSE

RESPONSE

Reference:

SUMMARY

1.0 INTRODUCTION

SUMMARY

SUMMARY

RESPONSE

Subject:

Dear Mr. Slear:

RESPONSE

RESPONSE

Navigation menu

ML19290E013

Text

5.6 REFERENCES

SUMMARY

RESPONSE

RESPONSE

RESPONSE

RESPONSE

RESPONSE

Reference:

SUMMARY

1.0 INTRODUCTION

SUMMARY

SUMMARY

RESPONSE

Subject:

Dear Mr. Slear:

RESPONSE

RESPONSE

Navigation menu

Search