1CAN128707, Forwards Info Re Elevated Reactor Bldg Temp & Schedule of Completion of All Committed Actions,Per NRC 871015 Request

From kanterella
(Redirected from 1CAN128707)
Jump to navigation Jump to search
Forwards Info Re Elevated Reactor Bldg Temp & Schedule of Completion of All Committed Actions,Per NRC 871015 Request
ML20237C864
Person / Time
Site: Arkansas Nuclear Entergy icon.png
Issue date: 12/11/1987
From: James M. Levine
ARKANSAS POWER & LIGHT CO.
To: Calvo J
NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM), Office of Nuclear Reactor Regulation
Shared Package
ML20237C867 List:
References
1CAN128707, NUDOCS 8712220231
Download: ML20237C864 (34)


Text

t o eg. , .-

ARKANSAS POWER & LIGHT COMPANY POST OFFICE BOX 551 LITTLE ROCK, ARKANSAS 72203 (501)377-4000 December 11, 1987 1CAN128707 U. S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 Attn: Mr. Jose A. Calvo, Director Project Directorate IV Division of Peactor Projects III, IV, V and Special Projects

SUBJECT:

Arkansas Nuclear One - Unit 1 Docket No. 50-313 l License No. DPR-51 Elevated Reactor Building Temperature Safety Evaluation Report for Arkansas Nuclear One, Unit 1 (ANO-1)

Dear Mr. Calvo:

We have received and reviewed the Safety Evaluation (SE) transmitted with your correspondence of October 15, 1987 (1CNA108703). This SE was in response to the submittal by AP&L of a Justification for Continued Operation (JCO) on August 28, 1987, due to the elevated temperature inside the ANO-1 reactor building.

Your aforementioned letter of October 15, 1987, in addition to providing the SE results, also requested that AP&L provide a schedule of completion of all committed actions other than those undertaken as part of the mid-cycle outage (October 16 - November 3, 1987) or the next refueling outage (IR8).

The enclosed information is in response to that request.

If you have any questions regarding this information, please contact my office.

Very truly yours, 8712220231 871211 / // -

PDR ADOCK 05000313 wt F P PDR Ja,mes M. Levine, Executive Director

AND Site Operations J

JML:MWT: lw Enclosure p" o li MEMBEA MiOOLE SOUTM UTILITIES SYSTEM

l

s. .

ATTACHMENT 1 l

A. SE, Section 3.2, Reactor Building Structural Integrity

1. The staff requested that the licensee submit ANO-1 reactor building prestressing tendon surveillance records for review.

AP&L Response: Attachment 2 contains records from the previous l

surveillance of the ANO-1 reactor building tendons which was the ANO-1 ten year physical tendon surveillance report on the tendon surveillance i conducted in February, 1983. The next scheduled tendon surveillance on ANO-1 is tentatively identified to commence the first week of February, 1988. As final negotiations with the contractor are completed and this date becomes firm, we will provide 30 days prior notification to the AN0 NRR Project Manager.

2. The staff requested that the licensee provide additional information on the liner integrity by comparing the calculated l liner strains associated with the high temperature to the design allowable liner strains committed to in the FSAR.

1 AP&L Response: A listing of original design liner strains, as derived from available original design calculations and the ANO-1 Safety Analysis Report (SAR) will be developed. These values will be compared to a listing of similar strain values developed from liner strain calculations utilizing the higher containment temperature. The resultant comparison will be completed and forwarded to you by March 1, 1988.

B. SE, Section 3.4, Piping and Piping Support

1. Provide the following information to clarify the piping evaluations:
a. Hydrogen Purge lines (HBC-1 and HBB-15) were reviewed concurrently. A calculation has been performed but I apparently further review is being undertaken. The licensee has committed to provide the conclusion of this review.

AP&L Response: From the JC0 Sectica II.B.4.b(4), page 7:

"HBC-1 & These lines will be reviewed concurrently, since they HBC-2 & are analyzed under the same analysis. The calculation HBB-15 analyzed the piping for 125 F normally and 289 F for a one time faulted case. Since it could be seeing temperatures of up to 160 F in the building further review is required.

The maximum thermal stress is 10902 psi. Multiplying this times a ratio of delta T's gives:

10902 X 90/55 = 17880 psi. This is acceptable vs. an allowable of 22,500 psi.

Increased support loads are being addressed in Sectica II.B.3 of this report."

The above item was apparently misinterpreted to indicate that "further l review" was required for these lines. This is not the case, as the  !

. intended " review" is contained in the evaluation statement. l The following rephrasing should explain:  !

l l "These lines are analyzed by the same calculations and are therefore  !

evaluated together. The file calculation quantifies the piping for normal operation at 125 F and one-time, faulted, operation at 289 F.

At elevated ambient temperatures, normal operation could have been as  !

high as 160 F. Since this is higher than 125 F, the following check of I'

! thermal stresses is provided:

Max. Thermal Stress @ 125 F = 10902 psi l Ratio of delta T's = 90/55 '

Therefore max thermal stress @ 160 F = 10902 x 90/55 = 17,880. This is i l ok, since 17880 <22500 = allowable  ;

Increased support loads are addressed in section II.B.3." l l

i

b. The piping lines HBD-14 and HBD-20, no thermal analysis was i l performed by the 2A and 2B cooler nozzles. However, a

! commitment was made to perform an inspection of these nozzles I to validate the approach used for estimating the thermal l stress increase in the nozzles of the VCC-2C and 2D coolers.

The licensee committed to provide the conclusion of this inspection to the staff.

AP&L Response: These nozzles were visually inspected during the  ;

l October, 1987, outage. No evidence of damage or elevated temperature  ;

j effects was detected, therefore, the conclusion of acceptable  ;

stresses / loads on these nozzles is validated.

l C. SE, Section 3.5, Environmental Qualification of Equipment l

1. To maintain compliance with 10CFR50.49, the licensee must update l the ANO-1 EQ program to reflect the results of the reanalysis.

AP&L Response: Actions to address this item are in progress. It is  ;

I anticipated that these activities will be completed resulting in the ANO-1 EQ program being updated with the results of the JC0 reanalysis by March 1, 1988.

D. SE, Section 3.6, Instrumentation Inside the Reactor Building

1. The licensee was requested to submit additional details regarding the few instruments where accuracy was affected by the higher temperatures. These details, as a minimum, should include the channel identification, the total amount of accuracy that was affected by the higher temperature, the original temperature drift, the new temperature drift, the original margin, and the new I margin. l l

AP&L Response: Attachment 3 contains supplementary information to that ,

present in Section II..D.4 of the JC0 pertaining to system effects.

This information discusses the methodology for evaluating equipment installed inside the reactor building as well as providing an evaluation of the systems inside containment. At the conclusion of this information a table is provided which compares, for certain instrumentation, previous vs. revised loop errors, total error, and margin to the safety analysis limit.

E. SE, Section 3.7, Non-EQ Electrical Equipment Inside the Reactor Building

1. The licensee should confirm that the insulations in the replacement pumps will be upgraded to Class "F" as well as the insulation in any other replacement motors in the reactor building.

AP&L Response: Future replacements of motors inside the ANO-1 containment will be specified to be Class "F". Changes to motor specifications are being made to assure that any replacements made inside the ANO-1 containment are provided with insulation of Class "F" l or better.

2. The licensee should confirm that the Maximum Ambient temperature referred to in Appendix II.D.2-2 of the JC0 is the Ambient Design Limit temperature. ,

1 AP&L Response: The' term Maximum Ambient temperature which is used in 4 the JCO, Appendix II.D.2-2 " Electrical Con.ponents Non-EQ Analysis" is j equivalent to the term " Ambient Design Limit" temperature. Both terms refer to the design value for the ambient temperature in which these components would normally be expected to operate. In cases where the actual ambient temperature exceeded this value, then the continued operation of that cotaponent, under those conditions, was justified.

l 1

1

)

i

,,--,r--. --,.,-.rr---wm-- - - - - - - ------ - - - . - - , - _ - -

.3 , 0l

' ' e J

l ATTACHMENT 3

II.D.4 SYSTEM EFFECTS II.D.4.a Methodology for Evaluating Equipment Installed in the Reactor ,

Building The approach described in the following paragraphs was used in evaluating the operability of devices subjected to ambient temperatures which may exceed the original assumed reactor building operating temperature.

  • Definition of Operability - For the purposes of this evaluation operability is defined as follows:
1. Normal Operating Condition - The device is capable of  !

detecting or maintaining the parameters of the system in which it is installed within design parameters.

2. Abnormal Operating Conditions - The device is capable of detecting or maintaining the parameters of the system in which it is installed within design parameters.
3. Accident / Emergency Operating Condition - The device is capable of detecting, responding, and performing its assigned function in the system in which it is installed, so that the parameters for the system and the plant are maintained within applicable analysis limits.
  • Ambient Temperature - The ambient temperature used for evaluation of operability was determined as follows:
1. Resistance temperature detectors which are installed for determining containment temperature during integrated leak rate testing are located at various elevations in the reactor containment building.
2. Ambient temperatures at various locations in the reactor building were measured using the installed resistance temperature detectors.
3. The data was L..alyzed and a reactor building temperature profile was developed as described in I.F.
4. The temperatures for instruments were rounded upwards to the next highest multiple of ten (e.g., 114 F to 120 F).
  • Equipment - With the exception of instrumentation, the predominant effect of increased ambient temperature is a potential reduction in useful tifetime of equipment (including cables). The effects of elevated reactor building operating temperature on the qualified life of environmentally qualified equipment is discussed in Section II.C.

The manufacturer's technical documentation was used to determine the design rated temperature of each device. This temperature was then compared to the component's ambient temperature to determine if the )

device was operating within its design rated temperature. In addition, for each instrument the effect of the increased ambient temperature on 1

accuracy was determined from manufacturer's technical data to determine whether the accuracy of each instrument was within the accuracy limits required by the design of the system in which it is installed.

II.D.4.b Evaluation of Systems Inside the Reactor Building Upon completion of the evaluation of the effects of increased reactor building temperatures on individual devices, interfaces and associated cables, the systems were reviewed to determine if any effects alter the operability of the subsystems, or associated systems. The results of this review are presented in the following subsections.

II.D.4.b(1) Safety Related Systems Pressurizer Subsystem The normal operating function of the pressurizer subsystem is to maintain reactor coolant system pressure at a nominal pressure of 2155 psig. This is accomplished by cycling groups of pressurizer heaters in response to reactor coolant system pressure instrumentation. The water level in the pressurizer provides an indication of reactor coolant system water inventory so that the ,

control room operator can take action to add or drain reactor coolant from the system. Pressurizer water level indication is provided through level transmitters connected to the pressurizer. The nominal (full power) operating level in the pressurizer is 220 inches.

The effects of increased reactor building ambient temperature on the level -

instrumentation and pressure instrumentation associated with normal operation have been evaluated in Subsection II.D.3 and were determined to have no significant effect on normal steady state operation.

During transient conditions, the pressurizer subsystem accommodates volumetric expansion and contraction for changes in power level, as well as various anticipated transients described in the SAR. This is accomplished by cycling heater banks to adjust for decreasing pressure and opening the spray line motor operated valve for pressure increasing transients. The equipment used to accomplish these functions are pressure transmitters, heaters, motor operated valves, and associated cables and connectors. The effects of elevated reactor containment building temperature on the equipment has been evaluated. The errors associated with the pressure instrumentation have a negligible effect on the pressurizer heater control circuitry located outside containment. The pressurizer heaters are affected by pressurizer temperature and heater temper ature; reactor building temperature does not affect the heaters. The operator for the spray valve is a Limitorque comparable to environmentally qualified Limitorque actuators which have been evaluated for elevated operating reactor building temperatures. The relief valve position indicating system (TEC Acoustic Monitor) has been evaluated for elevated reactor building temperature in Section II.C.

2

r ,

Some of the cable associated with the pressurizer previously exhibited age related degradation as a result of elevated reactor building operating temperature. The cable was replaced during the IR7 refueling outage and will be upgraded to high temperature silicone rubber insulated cable during an upcoming refueling outage. In the interim, the existing cable will be inspected for degradation at each refueling cutage. Analysis of instrumentation cables indicates a life in excess of 40 years at elevated reactor building operating temperatures.

The function of the pressurizer subsystem under accident conditions is to provide for pressure relief through safety valves to prevent exceeding plant design pressure. The code safety valves and associated piping have been analyzed, and it was established that the ambient temperature has no effect on safety valve performance. There is also an electromatic PORV associated with th? pressurizer which is set to relieve below the safety valve set point. iNe PORV is actuated by a class H insulated solenoid valve which is normally desnergized. The PORV can be isolated by a Limitorque motor operated vaive which is environmentally qualified. The PORV itself is not affected by ambient temperature. The MOV has a qualified life in excess of 40 years at the elevated temperature and the normally deenergized solenoid valve has an expected life in excess of 20 years at the elevated temperature.

It is concluded that the increased reactor building ambient temperature has no significant effect on the performance of the pressurizer subsystem.

Reactor Building Spray System The reactor building spray system is not required for normal operation. It 1 is required for reactor building heat removal in the event of LOCA or MSLB.

Portions of the reactor building spray system located inside the reactor building are only the nozzles, piping and check valves. This equipment is not affected by reactor building ambient temperature. The normally open reactor building sump isolation valves are environmentally qualified Limitorque motor operated valves, which are not adversely affected by reactor building ambient temperatures. Therefore, elevated reactor building ambient temperature has no significant effect on the reactor building spray system.

High Pressure Injection System The High Pressure Injection System is required for core heat removal in the event of a LOCA or MSLB. Portions of the high pressure injection system located inside the reactor building include only the piping, normally open manual isolation valves, and check valves. This equipment is not affected by reactor building ambient temperature. Therefore, the elevated reactor building ambient temperature has no significant effect on the High Pressure Injection System.

1 3

Low Pressure Injection System 1

The Low Pressure Injection System is required for core heat removal in the I event of a LOCA or MSLB. Portions of the low pressure injection system located inside the reactor building other than reactor building sump suction valves (addressed under reactor building spray system) include the piping, normally open manual isolation valves, and check valves. This equipment is j not affected by reactor building ambient temperature. Therefore, the l elevated reactor building ambient temperature has no significant effect on the Low Pressure Injection System.

Core Flooding System The core flooding system is required to function in the event of a LOCA.

The core flood system consists of two flood tanks which are nitrogen pressurized. Level instruments indicate core flood tank level and pressure instruments indicate nitrogen pressure. i The equipment associated with the core flooding system has been evaluated for the effect of elevated reactor building ambient temperature. The volumetric expansion of the liquid in the flood tanks due to increased ambient temperature has an insignificant effect on liquid level. Similarly the effect of increased ambient temperature on the pressure and level instrumentation has been analyzed and determined to be negligible.

The effect of increased temperature of core flood liquid has been analyzed with respect to emergency core cooling. It was determined that increased liquid temperature has no adverse effect on core cooling and an insignificant effect on reactor building temperature and pressure. It is therefore concluded that the effect of increased reactor building ambient temperature on the core flooding system is insignificant.

High Point Vent Subsystems The high point vent subsystem circuits are not used during normal operation but are used for startup and shutdown. The high point vents are provided for severe accident conditions (i.e. beyond design basis) in accordance with NUREG 0737. The high point vent subsystems inside the reactor building consist of piping, solenoid isolation valves, pressure transmitters used for leak detection, and connecting cable for the solenoid valve and pressure transmitters. The equipment has been analyzed for the effects of increased reactor building ambient temperatures. There is no effect on operability of the solenoid valves due to increased ambient temperatures other than qualified life reduction which is addressed in Section II.C. The pressure transmitters have been evaluated for the effects of increated ambient temperature and the error is less than 1%. This error is considered to be insignificant. Therefore, the elevated reactor building ambient temperature has no significant effect on the high point vent subsystems.

Air Particulate Monitoring System The air particulate monitoring system is used during normal operation to detect the presence of airborne radioactivity and is not required for accident conditions. The only portion of the air particulate monitoring system located inside the reactor building consists of a Limitorque motor operated containment isolation valve and connecting piping anc cabling.

This valve with its associated Limitorque actuator and the cabling to the actuator have been evaluated for increased reactor building ambient temperature. There is no significant effect of increased ambient temperature on the valve, the actuator, or the cabling. It is therefore concluded that the air particulate monitoring system is not significantly affected by increased ambient temperature.

Reactor Building Air Purge System The reactor building air purge system is not used during normal operation.

The isolation valves are locked closed except during cold shutdown. The only portion of the reactor building air purge system which is located inside the reactor building are the motor operated reactor building isolation valves and connected piping. The motor operators are Limitorque actuators and the valves are Henry Pratt Co. 54" butterfly valves.

Evaluation of the Limitorque operators for increased ambient temperature indicates that the operators have a useful life in excess of 40 years. The butterfly valve seat material is rated for 300 F. Aging data is not available for the propr'ietary seat material. However, technical specification leakage tests determine whethar seat material degradation is occurring. It is concluded that the increased reactor building ambient temperature has no significant effect on the reactor building air purge system.

Hydrogen Purge System The hydrogen purge system is not used for normal operation. It can be used for post LOCA removal of hydrogen although it is not credited as such. The only equipment in the hydrogen purge system located inside the reactor building are the Limitorque operated containment isolation valves and the connected cabling and piping. The Limitorque operated valves are  ;

environmentally qualified and increased reactor building temperature has 1 been evaluated and has no effect on operation of the valves. The qualified .

life is in excess of 40 years. The connected cable has been evaluated for l elevated reactor reactor building temperature and the qualified life is in excess of 40 years. It is concluded that elevated reactor building temperature has no significant adverse effect on the hydrogen purge system.

1 j

5 I I

i<

Hydrogen Recombination System The hydrogen recombination system is not used for normal operation. It is-used to recombine hydrogen in the post LOCA period. The system consists of Westinghouse Electric recombination units, a non-safety grade thermocouple which is not used for control or operation, and interconnecting cabling.

The recombiner consists of a metal structure and metal oxide resistance heaters. The recombiner is not significantly affected by reactor building ambient temperature. The non-safety grade thermocouple is not significantly affected by ambient temperatures. If for any reason it is not functional, it has no significant effect on operation because operation is controlled by current / power to the heating elements.

It is therefore concluded that reactor building ambient temperature has no significant effect on the electric hydrogen recombiner system.

Reactor Building Cooling System The reactor building HVAC system is required during normal operation to maintain the building temperature within design limits. It is required to provide heat to the builoing for cold shutdown periods during winter months.

The HVAC system is also required for reactor building cooling in the event of LOCA or MSLB. The system consists of vane-axial fans and motors, Limitorque motor operated dampers, Namco limit switches, interconnecting ducting, piping, and interconnecting cabling. The Reliance Class H insulated motors are environmentally qualified units which have been evaluated for increased ambient temperature and the qualified life is in excess of 40 years. Testing of the motors at temperatures in excess of 200 F for one year after simulated 40 year aging showed operability. The environmentally qualified Limitorque motor operator and dampers have been evaluated for increased ambient temperature, and are operable for a qualified life in excess of 40 years. The environmentally qualified Namco limit switches have been evaluated and are presently within their qualified life as discussed in the section on environmentally qualification. With the ,

exception of the power cable, the environmentally qualified cable has been '

evaluated as having a life in excess of 40 years. The qualified life of the power cable is 20 years. The interconnecting ducting including chillers and piping has been evaluated and is not affected by increased reactor building ambient temperature.

It is concluded that the elevated reactor building operating temperature l will not significantly affect performance of the reactor building cooling system. Subsection I.F provides a discussion of reactor building cooling system performance as this relates to the elevated reactor building operating temperature.

[

r 6

L ._ .___ _ _ -

Reactor Protection System Setpoints High Reactor Coolant Pressure Trip (PT-1021, 1023, 1038, 1039)

The high reactor coolant pressure trip establishes a margin between the maximum steady state pressure and the high reactor coolant pressure of 2750 psig (110% of design pressure). It is the primary trip for pressure increasing transients resulting from slow reactivity insertion such as small rod group withdrawal or moderator dilution events, or from a decrease in secondary system heat removal such as loss of feedwater or turbine trip.

The high reactor coolant pressure trip setpoint used in the safety analysis is 2448 psig at the core outlet. The current Technical Specification setpoint is 2355 psig. There is a -58 psi process error resultant from the maximum dynamic pressure difference between the core outlet and the pressure measurement tap for any permitted RC pump operating configuration.

The previous instrument loop error was 30.45 psi utilizing linear addition of individual device errors. The revised instrument loop error is 25.23 psi utilizing the square root sum of squares (SRSS) methodology for combining individual device errors. This revised instrument loop error assumes the component ambient temperature is 150 F for conservatism. The actual maximum component ambient temperature should be less than 130 F.

With the revised total error (process error plus instrument loop error) of

-83.23 psi using SRSS Methodology, a margin to the safety analysis limit of 9.77 psi exists. Therefore the higher ambient reactor building temperature is acceptable for this trip. Furthermore, using the actual maximum component ambient temperature of 130 F with linear addition of device errors yields an-acceptable margin to the safety analysis limit.

Low Reactor Coolant Pressure Trip (PT 1021, 1023, 1038, 1039)

The low reactor coolant pressure trip prcvides transient DNB protection and is the primary trip for pressure decreasing transients such as steam line break, steam generator tube rupture, etc.

The low reactor coolant pressure trip setpoint used in the safety analysis is 1785 psig at the core outlet. The current Technical Specification setpoint is 1800 psig. There is a -19 psi process error resultant from the minimum dynamic pressure difference between the core outlet and the pressure q measurement trip, for any permitted RC pump operating status. j The previous instrument loop error was 30.45 psi utilizing linear addition of individual device errors. The revised instrument loop error is 25.23 psi utilizing the SRSS methodology. This revised instrument loop error assumes the component ambient temperature is 150 F for conservatism. The actual  ;

maximum component ambient temperature should be less than 130 F.

7 i

_ . _ _ _ _ J

i With the revised total error of +6.23 psi using SRSS Methodology, a margin to the safety analysis limit of 8.77 psi exists. Therefore, higher ambient reactor building temperature is acceptable for this trip. Furthermore, using the actual maximum component ambient temperature of 130 F with linear addition of device errors yields an acceptable margin to the safety analysis limit.  !

, Variable Low Pressure Trip (PT 1021, 1023, 1038, 1039) l (TE 1012, 1013, 1040, 1041)

The variable low pressure trip provides primary steady-state protection for DNBR and varies as a function of reactor outlet temperature.

The variable low pressure trip setpoint used in the safety analysis is the set of pressure-temperature limits shown in Technical Specification Figure 2.3-1 which constitutes the Technical Specification setpoint.

This figure already takes into consideration the 19 psi process error discussed previously with the Low Reactor Coolant Pressure Trip. In addition, this figure takes into account an assumed worst case instrument loop error of 50 psi.

The previous instrument loop error was 40.05 psi utilizing linear addition methodology. The revised instrument loop error is 46.43 psi utilizing the same methodology. Once again the revised loop error assumes the component ambient temperature is 150 F for conservatism. The actual maximum component ambient temperature should be less than 130 F.

With the revised instrument loop error of 46.43 psi a margin of 3.57 psi exists; therefore, higher ambient reactor building temperature is acceptable for this trip.

High Reactor Building Pressure Trip (PT 2400, 2401, 2402, 2403)

The high reactor building pressure trip initiates a reactor trip whenever the reactor building pressure increases beyond the pressure setpoint, i.e, when conditions indicative of high energy line breaks are present. This instrument provides a trip for small, high energy line breaks inside containment, although credit has not been taken for this trip function in the safety analysis. Both LOCA and non-LOCA analyses use the low reactor coolant system pressure trip for transients where this trip is expected to provide the first trip. The Tech. Spec. value of this trip setpoint is taken as a nominal 4 psig.

The elevated reactor building ambient temperature increases the error from the previous 0.301 psi to 0.396 psi. The revised error assumes the component ambient temperature is 170 F for conservatism. The actual maximum component ambient temperature should be less than 150 F.

The increase in measurement error due to the elevated ambient temperature has an insignificant impact on the trip function as no credit was taken for this trip in the safety analysis. This trip serves as an additional protective function of the RPS beyond that credited in the safety analysis.

- )

Power / Imbalance / Flow Trip (PDT 1028 thru 1031, PDT-1034 thru 1037, NE-0509 thru 0512)

The power imbalance flow trip provides primary protection for DNB and linear heat rate (kW/ft) limits during steady' state operation. The power flow setpoint provides overpower protection for partial reactor coolant pump conditions, reactor coolant pump coastdown and locked rotor events.

The calculation of the power / imbalance / flow trip is broken down into two main tasks, the calculation of the flux / flow setpoint and the calculation of the power-imbalance setpoint envelope. The flux / flow setpoint is determined from the'limitiag RC pump coastdown while the power-imbalance envelope is determined from steady-state DNBR and linear heat rate power-offset limits.

For this evaluation, the Technical Specification safety limit envelope was error adjusted with the increased errors and then compared to the Technical Specification setpoint. Adjustments in the flow, power and imbalance parameters were considered.

A comparison of the total error for this trip for four, three, and two pump operation is shown below. The neutron power measurement has no increased error due to elevated temperature. The flow transmitters do, however, exhibit an increased error, and the revised error is shown below based on 150 F component ambient temperature for conservatism. The actual ambient temperature at the installed location should not exceed 130 F.

Previous Error New Error

%FP %FP Heat balance error 2 2 Neutron Measurement 4 4 Bistable 0.84 0.84 Deadband on Function Generator 0.16 0.16 Flow error 2/2 2.12 2.45 2/1 3.12 3.82 1/1 3.55 4.44 Total error 2/2 9.12 9.45 2/1 10.12 10.82 -

1/1 10.55 11.44 9

L _ ------- -- -

l The peak power levels for three and two pump operations were calculated as 90.75 %FP and 64.08 %FP res'pectively. (These were calculated by multiplying the flux / flow ratio times the flow and adding the appropriate instrument errors.) These can be compared to the allowed peak power levels from reload core DNBR analysis of 91.2 %FP and 64.5 %FP. Since the maximum allowed partial pump power levels are above those calculated with the new errors, the Technical Specification setpoint envelope remains bounding.

The adjustment showed that there exists sufficient margin in the imbalance portion of the trip envelope to accept an increased error due to elevated reactor buildir.g ambient temperatures.

High Reactor Coolant Outlet Temperature Trip (TE 1012, 1013, 1040, 1041)

The high reactor coolant outlet temperature trip establishes an absolute upper limit on reactor coolant outlet temperature and limits the range over which the variable low pressure trip setpoint must provide protection.

This trip is not used as the primary trip function for any accident analysis and exists only as a backup to other RPS trips. The current Technical Specification setpoint is 618 F.

The elevated reactor building temperature has no measurable effect on the instrument string error because the RTD measurements are insensitive to changes in reactor building temperature. .

1 High Flux Trip (NE 0509, 0510, 0511, 0512)

The high flux trip provides the primary overpower protection. This trip limits power during steady-state operation to less than 112% FP. The high flux trip also provides transient protection for fast reactivity transients where the power changes faster than other system parameters. The power is  ;

measured by out-of-core uncompensated ion chambers that detect fast neutron leakage.

The high flux trip setpoint used in the safety analysis is 112% FP. The current Technical Specifications high flux trip setpoint is 104.9% FP. .

l The neutron measurement error is bounded by a constant 2% FP error assumption in the safety analysis calculations. This 2% FP error accounts for changes in the process following calibration. Compensation for the actual neutron detector error is provided when the out of core neutron detectors are calibrated to the plant heat balance. The error here is bounded by the assumed 2%. Thus, any ambient temperature changes are accounted for during normal neutron detector calibrations. Also, the instrument string was calculated to be less than the assumed 4% error.

Since the assumed error for the high flux trip setpoint adjustment is unchanged, the Technical Specification setpoint is unchanged and therefore, higher ambient temperature is acceptable for this trip.

10 l l

Power / Pumps Trip (NE 0509, 0510, 0511, 0512)

The' power-to pump monitor trip ensures a reactor trip when no reactor coolant pumps are operating in one steam generator loop. This trip also provides the primary protection for multiple reactor coolant pump coastdowns, single reactor coolant pump coastdown from partial pump operation, and reactor coolant pump coastdowns resulting in the loss of both pumps in either loop.

This trip is designed to operate in a nearly digital manner (i.e. , power operation allowed, or power operation not allowed). The determination is based on the comparison of the measured neutron power to the allowed power for the pump combination operating. The contact monitor in the trip string rapidly determines allowed power level for the pump combination, and the bistable then determines the need for trip.

4 Due to the nature of this trip (essentially digital), the equipment accuracy (neutron flux) is accounted for in the response time of the instrument loop.

If the response time can be met, the loop is performing adequately.

Therefore, a ' relationship between the analysis setpoint and the Technical Specification retpoint does not really exist and a calculation of the setpoint is not necessary.

Since the response time of the pump monitor is not impacted by the temperature, the effect of higher ambient temperatures on this trip are inconsequential. No effect on the neutron flux measurement exists either as shown in the previous discussion. >

ESAS Setpoints Low RC Pressure Actuation (PT-1020, 1022, 1040)

The ESAS low reactor coolant pressure actuation setpoint initiates high pressure and low pressure injection (HPI and LPI). The low reactor coolant pressure actuation setpoint used in the safety analysis is 1480 psig. The current Technical Specification setpoint is 1526 psig.

The previous instrument loop error calculation utilized in the setpoint analysis assumes an accident temperature of 180 F. Sufficient conservatism exists in the input assumptions which establish the basis for the 180 F to account for the elevated ambient temperature inside the reactor building.

Hence there are no changes in the revised instrument loop error calculation; therefore, the higher ambient reactor building temperature is acceptable for l this actuation.

High Reactor Buildino Pressure Actuations (PT 2405, 2406, 2407)

Within ESAS there are two setpoints (4 psig and 30 psig) associated with the high reactor building pressure transmitters. The 4 psig actuation setpoint initiates high pressure injection and low pressure injection (HPI and LPI).

In addition the 4 psig actuation setpoint initiates containment isolation.

The 30 psig actuation setpoint initiates reactor building spray with chemical addition.

11 l

r :-

The 4-psig actuation for HPI and LPI serves as a backup to the ESAS low reactor coolant pressure actuation. No credit is taken for this actuation in the safety analysis; therefore, no analysis setpoint has been established for this function.

The 4 psig actuation for containment isolation was selected arbitrarily.

The SAR only takes credit for the fact that containment isolation takes place. The isolation should take place within the first minute or so during a LOCA. The isolation valve stroke times are approximately 30 seconds. As can be seen from SAR Figure 6.2-8, the pressure rises to approximately 8 psig within the first second after a large LOCA and to 40 psig within the first 10 seconds. Therefore, small variations in the pressure trip will not significantly affect the total isolation time.

The 30 psig actuation for containment spray with chemical addition was also chosen arbitrarily. Containment spray is primarily useful in long term cooling and dose reduction due to iodine scavenging. Additionally, this setpoint will be reached during the initial pressure spike of a large LOCA as shown in SAR Figure 6.2-8. Therefore, small variations in the containment spray actuation point do not significantly alter the plant response to a LOCA.

' EFIC Trip Setpoints The only EFIC trip setpoint that could be affected by the increased reactor building temperature is the low OTSG level initiate. The other setpoints such as EFW level control setpoint (at least one RCP running), natural circulation control level setpoint (no RCP's running), and ECCS level i

setpoint are control levels, not initiate setpoints. Their setpoints are based on accident errors. The main steam isolation setpoint is unaffected because the steam generator pressure transmitters are located outside the reactor building.

Low OTSG Level Initiate (LT-2617, 2618, 2621, 2622, 2667, 26G8, 2671, 2672)

The low OTSG Level Initiate function provides for Emergency Feedwater Actuation upon loss of OTSG inventory for loss of feed and feedwater/ power mis-match events, j

The safety analysis setpoint for this initiate function is 6 inches above the lower tube sheet. Although no setpoint is established in the technical specifications, the SAR establishes a 13.5 inch setpoint.

12

- - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - )

The present instrument loop error analysis assumes 100 F for calibration temperature for the transmitter including reference leg and 140 F at time of initiation. Calibration procedures assume 120 F for reference leg temperature. New error analy.iis has shown that using the more realistic assumptions of 80 F for transmitter calibration and the actual reference leg temperature assumed during calibration of 120 F, yields acceptable results for an assumed 140 F temperature at the time of initiation. The actual temperature should be less than 120 F. The reference leg heatup error is not as large as previously calculated. There is a slight increase in instrument error due to elevated ambient temperatures, but the margin to the safety analysis limit is larger than previously calculated because the increase in instrument error is more than offset by the reduction in reference leg heatup.

II.D.4b(2) Non-Safety Related Systems Control Rod Drive System The only portions of the contrcl rod drive system which are located in the reactor building are the rod drive mechanisms, position indicators, and connected cabling. Calculations as well as actual measurements of the reactor cavity ambient temperature demonstrate that the temperature does not exceed 154 F. The rod drive motors are independently water cooled and therefore are not affected by the ambient temperature. The connectors and cables are all high temperature materials having a continuous rating of 200 F or more. The supplier of the rod position indicators (B&W) has stated that the insulation system is class H rated for 393 F except for the circuit board which must be maintained below 300 F. The rod drive mechanism pressure boundary is designed for 2500 psi at 650 F and is not significantly affected by.the ambient temperature. It is therefore concluded that elevated reactor building ambient temperature has no significant effect on the portions of the rod drive control system located inside the reactor containment building.

Reactor Coolant Quench Tank Subsystem The reactor coolant quench tank system is required for startup, normal operation, and shutdown. It is not required for accident conditions. The equipment, piping, and instrumentation associated with this system are located at the lowest elevation of the reactor building. This area has not been subjected to elevated temperatures and therefore there are no questions of operability associated with the system.

Reactor Coolant Pump and Motor Assembly This subsection summarizes evaluations completed to demonstrate that the Reactor Coolant Pump (RCP), its motor and all associated supporting equipment are acceptable for the existing ambient temperatures. The major sources for identifying these systems and components have been the ANO-1 SAR (Figure 7-1), the Byron-Jackson Pump Manual, and the Allis-Chalmers Motor Manual.

I i

13 {

l l t

\ - _ _ _ _ i

General Instrumentation, electrical, mechanical and structural aspects were examined on a component level. The performance of the RCP during operation is measured by the seal cartridge parameters, motor bearing temperatures, motor oil sensing devices, and the vibration monitoring equipment. None of these devices are used alone to take action which might challenge any safety system.

The evaluation for reactor coolant pump instrumentation was conducted by considering specific errors associated with elevated temperature as well as any general effects on equipment reliability, operability, etc. as they might affect the plant operators. Thi: evaluation showed that a significant portion of the varices instrumentation- k qualified for the actual, reactor building temperatuhn and the remainder are justified for elevated temperature use band on the following'; considerations: ,

Pesomance of mechanical devices is not significar.tly affected by ambient temperature variatio n In-line flow eleme.rts (e.g. lube oil flow switches) are affecte3 Dy process temper 3h re only.

Local pressure, level'and flow gauges are designed for process temperature conditions md are used ir. frequently for performar.c {

monitoring, and n.;t for post-accidst monitoring. ,

/

r

(, / ,\

I RCP vibration switchei are u, sed for qualitative ayesmit of s nump .(

performance (j e., trend monitoring) and are bac y r; of a

~

t 1

complete and fadependent analog vibration system wnIcb is l

qualified for the Axisting temperature conditions and which provides mulb hare speciftz,f9 fermativa.

t , ,

e RCP seal cavity pretsure and seal bleedoff flow are,ged for 5 trending indication. Alarm functlons are used tc:tritjger increasedoperatormonitoringofvariousperjormanceicdicator j y trends. ,

k ej e

)

The RCP motor and auxiliary motors were teviewed for insuleMon rating, ,

h ambient temperature rating, and operationa) mode (i.e. , cc.titv'cA's, intermittent). The motors were then euluated for,* he effect o f the specific ambient temperature or eace n.che. .The RCP motors and most of the auxiliary motors are rated for 50 V antilit knd crovided with Class F insulatinn. From the data reviewed,9.he' auhi iary (W. ors were Vetermined to be within equipment ratigs even thrv@y. In st ne case! (i.e. , auxiliaries) the actual ambient tegoerature; excee6 <tne des'ign a@ umpdan. This is based ontheinsulationtemMratureratingrdcteingexc/jJedonacontinuous basis.

~

l , e , ,

/ / ,

The power and control cables we}e rev?dwed for te,rgaturi consider 3tions.

The cables are rated for 9N 4:d, is adequate A r the actual reactor '

j building amt lent temperatures. The effect of the ambienti'smperature nn N cable ampacity would be minimal >'since the cables ar2,nized for a a nimum of 125% of mq p full load cur pat by design. The RCP p>4cr surge capacitors 'l J

were reviesea Tor potential /anbv?rt temperature effect. Since the actual i temperaturesarewithintheidesh). lin:its, the capacitors are adequate.

i i c

$ /

14

,f ,

i - i 1'

v l1

)'J l

a  ; s.

't ' {\ . '

+q) ,, ,

1

, j t

9'?b 1

' l' -

The R(?1[,~

, notor iubricant elastomers and RCP mechanical componeats were

{

's 3 i evaluctid to identify peak effects of increased.r'eactor bui?cing I temperature. The evaluations were based on the actual tduperature profiles 4 '

and vendor desi'gn information. In general,the mechanical;compunents are effected more oy the cooling medium than ny the ambient temperatures. The

! motor lubricant and pump' cartridge seal cre both cooled by the Intermediate Cooling Water (ICW) systen. The motor lubrication was evaluated for'

, performance in the system and through.the oil collection system. Individual pumps in the oil system and the sea' fleak off system were reviewed for l

design ratings and elastomer / seal r;crformance. Although the seal cartridge i f &, elastomers are not significantly S fected f by ambient temperatures, they were N

g evaluated. Mechanical component / issociated with the reactor coolant p pg pump / motor assembly were qualified or ent.iuated satisfactorily for the elevated anbient temperatures. Specifie component. evaluations are discussed LT w in Subsection 11.0J1 H L s The piping systems (pipe anu supperts) attached to the pump and motor were L

' evaluated for the elevated temperature ef#ects. Subsebcion II.B.4 describes the specific components. All iines atttdried to the assemoly were acceptable. The snubbers attached tc the RCP were alco 'edaluated for the higher ambient conditiens. A review of Jhe vendor data'and test information indicated tnat the snubbers were acceptes"e for higher ambieret temperatures.

, The failure mode of these snubbers is a weeping seal which would be ditected through the rigorous snubber inspection program in place at ANO-1. This weeping would not effect the accident perfor;Mance of the snubbert. The '

s specific evaluation is9 contained in II.B.4. Gf this report.

N l '

From the above discussion, all areas related to the Reactor Coolant Purp ar.d-motnr atsembly were evaluated. The condition of components was found to Se accept-able for,the higher temperature based on the design ratings or through evaluallan that' the performance would not cause the operators to taCa-unneed(d

  • actions.

Integrated Cintrol Systen and Non-Nuclear Instrumentation .

The Integitted Control System (ICS) provides the prcper coordination of the reactor,1Leam generator feedwater control and turbine under all operating conditions. Proper' coordination consists of producing the best load response to the unit load demand while recogni:.ir.y the capeilities and limitations of the reactor, steam generator, teedwater systes,' and turbine.

When any single portion of the plant is at an operating Hmit or a certain s control section is in manual, the ICS design uses the ilmited or manual section as a load reference.

The ICS maintains const nt average reactor coolant temperature between 15 and 100 percent rated poker and constant steam pressure at all loads.

Optimum unit performance is maintained (1) by limiting stets pressure variaticos, (2) by limiting the imbalance between the steam generator',

turbine, and the reactor, and (3) by limiting the total unit load demand upon loss of capability of the steam generator feed system, the reactor, or the turbine generator. The control system provides ?imiting actions to l assure proper relationship: between the generated load, steam and feedvater flow, and reactor power.

15

e l'

Non-Nuclear Instrumentation The non-nuclear process instrumentation (NNI) provides the required input signals of process variables for the reactor protection, regulating and auxiliary systems. It performs the required process control functions in response to those systems and provides instrumentation for startup, operation and shutdown of the reactor system under normal and emergency conditions.

1. Reactor Inlet Temperature (T-cold)

T-cold is measured with resistance temperature detectors (RTDs) in the reactor inlet piping connected by leadwires to 4-wire resistance bridges outside the reactor building. RTDs with 4-wire bridges are totally insensitive to changes in ambient temperature. Therefore, the elevated reactor building temperature induced error in T-cold is zero, and the effect of the error in T-cold on ICS operation is zero. The effect of the error on the assumptions for reactor coolant temperature in the safety analysis is, likewise zero.

2. Pressurizer Level Pressurizer level is measured with differential pressure sensors inside the reactor building connected to the pressurizer with a reference leg at reactor building temperature. Pressurizer level is controlled by NNI at a setpoint of 220 inches.

The accuracy of the level signal and control is affected by high reactor building temperature induced (HRBTI) reference leg error, which is non-random, and HRBTI sensor error, which is random. The HRBTI reference leg error stems from fluid density changes in the reference leg. Automatic control response or operator manual response to this non random error would lower the actual pressurizer level below the setpoint value of 220 inches.

The HRBTI sensor error would, in the worst case, add to the HRBTI reference leg error and produce a maximum combined HRBTI error of +1.75 percent or 5.6 inches below the 220 inch setpoint. An HRBTI error of this magnitude would nave no adverse effect on normal pressurizer operation.

If, however, during abnormal transients the pressurizer level falls to the 40 inch setpoint for the pressurizer heater low level interlock, the possibility did exist that some of the elements in the upper bank of pressurizer heaters could be uncovered. A modification to raise the setpoint has been implemented. This is an equipment protection concern and not a safety concern.

16

3. Pressurizer Temperature f Pressurizer temperature is measured with RTDs in the pressurizer connected by leadwires to 4-wire resistance bridges outside the reactor building. Since RTDs with 4-wire bridges are totally insensitive to changes in ambient temperature, the HRBTI error in pressurizer temperature is zero. The effect of HRBTI error on temperature compensation of pressurizer level is zero.
4. Reactor Coolant Pump (RCP) Seal Bleed Off Temperature RCP seal cavity bleed off temperature is' measured with an RTD in each RCP seal bleedoff lines inside the reactor building connected by leadwires to 4-wire resistance bridges outside the reactor building.

Since RTDs with 4-wire bridges are totally insensitive to changes in ambient temperature, the HRBTI error in RCP seal bleedoff temperature is tero.

5. RCP Seal Cavity Pressure RCP second and third stage seal cavity pressure is measured with pressure sensors in the reactor building. RCP seal cavity pressure is a measure of satisfactory seal staging. Variations of 400 psig on second stage seal cavity pressure of 1200 psig and 200 psig on third stage seal cavity pressure of 770 psig are acceptable in the operating procedures.

The HRBTI error in both the second ard third stage RCP seal cavity pressure is .45% or 11.25 psig. Aa HRBTI error of this magnitude is six percent or less of the allowable variation in seal cavity pressure and, therefore, has no significance in operator interpretation of RCP seal cavity pressure information.

6. Steam Generator (SG) Lower Downcomer Temperature SG lower downcomer temperature is measured with RTDs in the SG connected to 4-wire resistance bridges outside the reactor building.

Since RTDs with 4-wire bridges are totally insensitive to changes in ambient temperature, the HRBTI error in SG lower downcomer temperature is zero.

7. Steam Generator (Operate Range) Level Steam generator (SG) operate range level is measured with differential pressure sensors inside the reactor building connected to the SG with a reference leg at reactor building temperature. SG level is normally an uncontrolled variable above 15 percent power, but the SG operate range level is used to limit high SG 1evel at 92.5 percent of operating range.

17

4 The accuracy of the level signal and high limiting action is affected by HRBTI reference leg error, which is non-random, and HRBTI sensor errur, which is random. The HRBTI sensor error would, in the worst case, add to the HRBTI reference leg error and produce a maximum combined HRBTI error of 1.2 percent. This would reduce the high level limiting action to 91.3 percent of operating range and reduce unit power output by one to two per cent, if either steam generator were on the upper level limit. This potential restriction in unit power has operational significance, but no safety significance.

8. Steam Generator (Startup Range) Level SG startup ranga level is measured with differential pressure sensors inside the reactor building connected to the SG with a reference leg at reactor building temperature. SG level is normally an uncontrolled variable above 15 percent power, but the SG startup range level is used to limit low SG level at 27.5 inches. The accuracy of the level signal and low level limiting control is affected by HRBTI reference leg error, which is non-random, and HRBTI sensor error, which is random.

The HRBTI sensor error would, in the worst case, add to the HRBTI reference leg error, produce a maximum combined HRBTI error of 1.68 percent and lower the low level limiting control by 4.2 inches. This would have no significant adverse ef fect on ICS or plant operation, nor would it represent an increase in challenges to Emergency Feedwater Initiation and Control (EFIC) during low load operation when steam generator level is on the low level limit, because ICS low level control does not overlap EFIC actuation with abnormal and HRBTI errors applied to both systems.

9. Steam Generator Full Range Level SG full range level is measured with a differential pressure sensor inside the reactor building connected to the SG with a reference leg at reactor building temperature. SG full range level is used only for filling and draining the steam generator during plant shutdown and is not needed during plant operation.

The accuracy of the level signal is not affected by HRBTI reference leg error or HRBTI sensor error, because the level indications are only used during plant shutdown.

10. Reactor Coolant Pressure Low Range Low range reactor coolant pressure is measured with a pressure sensor inside the reactor building connected to the reactor coolant hot leg.

The signal is not used for control or protection actions. The HRBTI sensor error is 0.03 percent or 1.5 psig. An HRBTI error of this magnitude has no significant impact upon the operator interpretation of the low range reactor coolant pressure indication.

18

_ - - , . _ - , - - - - , - - - - - - ' ' ' " ' ~

11. Steam Generator Outlet Steam Pressure SG outlet steam pressure is measured by pressure sensors in the reactor building connected to each SG outlet steam header. ICS opens the

. turbine bypass valves when the SG outlet steam pressure exceeds 1050 psig. The ICS BTU limits have recently been modified to perform only alarm functions rather than control functions. Therefore, the SG outlet pressure input to the BTU limits should not have any effect on plant control.

Although the HRBTI sensor error is 0.3 percent (3.6 psi) on one of the four sensors, ICS turbine bypass control of high SG outlet steam

pressure is not adversely affected by HR8TI sensor errors.

L

12. Reactor Coolant (RC) Flow Reactor coolant flow is measured by differential pressure sensors connected to flow nozzles in the reactor coolant hot leg. This buffered output signal from the reactor protection system (RPS) has been removed from the BTU limiting function of'the ICS. Consequently it is no longer used in the feedwater flow ratio circuits of the ICS unless the plant is operating with less than 4 RC pumps. 1 HRBTI errors on the RC flow signal do not adversely affect the feedwater flow ratio circuits in ICS, because the RC flow is only used as a feedforward signal with final ratio control from the difference between Tcold in each loop. In this control configuration RC flow errors up to ten per cent or more could easily be accommodated.

Challenges to the RPS are not increased because RC flow is an uncontrolled variable.

13. Reactor Coolant Pressure - Narrow Range Narrow range Reactor Coolant System pressure is measured with pressure sensors inside the reactor building connected to the reactor coolant hot leg. Reactor coolant pressure (narrow range) is used by the NNI to control operation of the' pressurizer heater banks, the pressurizer spray valve and the PORV. The signal originates from, but is isolated from safety related circuits.

The accuracy of the reactor coolant pressure signals and controls is affected by the HRBTI sensor error (0.38 percent or 3 psig increased error). HRBTI errors of this magnitude have no significant adverse effects on reactor coolant pressure control. ,

Challenges to the RPS are not increased because NNI control of reactor coolant pressure does not overlap RPS trips on high and low reactor coolant pressure with abnormal and HRBTI errors applied to both systems.

l 19

14. Reactor Coolant Pressure - Wide Range Wide range reactor coolant pressure is measured with pressure sensors inside the reactor building connected to the reactor coolant hot leg.

Reactor coolant pressure (wide range) is used by the NNI to interlock (close) the decay heat isolation valves and to interlock (close) the PORV for LTOF. These signals are isolated from the ESAS safety related circuits.

The accuracy of the reactor coolant pressure signals are affected by the HRBTI sensor error (0.14 percent or 3.6 psi increased error).

HRBTI errors of this magnitude have no significant adverse effects on the NNI interlock functions.

15. Reactor Outlet Temperature (T-hot)

T-hot is measured with RTDs in the reactor outlet piping connected by leadwires to 4-wire resistance bridges outside the reactor building. Since RTDs with 4-wire bridges are totally insensitive to changes in ambient temperature, the HRBTI error in T-hot is zero, the effect of HRBTI error in T-hot on ICS operation is zero, the effect of HRBTI error in T-hot on the assumptions for reactor coolant temperature in the safety analysis is zero and the increase in challenges to the RPS is zero.

16. Core Flood Tank Level Core flood tank level is measured in each tank with differential pressure sensors inside the reactor building connected to the core flood tank with a dry reference leg at reactor building temperature.

The fluid and gas in the core flood tank are at the same temperature as the gas in the dry reference leg. The core flood tank level signals are used only for monitoring and do not have automatic control or interlock func; ions.

The additional error associated with the tank level instrument due to HRBTI was evaluated and determined to not significantly affect the safety analyses assumptions.

17. Core Flood Tank Pressure Core flood tank pressure is measured in each tank with pressure sensors inside the reactor building connected to the core flood tank. The core flood tank pressure signals are used only for monitoring and not for any automatic control or interlock functions.

The additional error associated with tank pressure instruments due to HRBTI was also evaluated and determined to not significantly affect the safety analyses assumptions.

20

. a .

18. Neutron Flux-Power Range Power range neutron flux is measured with an uncompensated ion chamber in the reactor building around the reactor vessel. Reactor power signals from the RPS are used for reactor control in the ICS. These signals originate from, and are isolated from the RPS safety related circuits.

The accuracy of the reactor power signals is not significantly affected j by changes.in ambient temperatures. Therefore, the HRBTI error in reactor power is zero, the affect of HR8TI error in reactor power on ICS operator is zero, the increase in challenge to the RPS is zero, and, the effect on the assumptions for reactor power in the safety analysis is zero.

19. Conclusion Regarding ICS and NNI An error survey of the following ICS/NNI input measurements in the-reactor building for the impact of high reactor building induced (HRBTI) errors was completed:

Reactor Inlet Temperature Pressurizer Level Pressurizer Temperature Reactor Coolant Pump Seal Bleed Off Temperature Reactor Coolant Pump Seal Cavity Pressure Steam Generator Lower Downcomer Temperature Steam Generator (Operate Range) Level Steam Generator (Startup Range) Level Steam Generator (Full Range) Level Reactor Coolant Pressure - Low Range Steam Generator Outlet Steam Pressure Reactor Coolant Flow Reactor Coolant-Pressure - Narrow Range Reactor Coolant Pressure - Wide Range Reactor Outlet Temperature Core Flood Tank Level Core Flood Tank Pressure Neutron Flux-Power Range The results of the error survey showed that HBRTI errors in ICS input measurements did not adversely affect ICS/NNI performance in full automatic or operator manual mode, and that challenges to EFIC, ESAS, and RPS were not increased by the application of combined abnormal and HRBTI errors to both ICS control and EFIC and RPS actions.

II.D.4.c Conclusion of Systems Evaluation The evaluation of the effect of increased reactor building ambient temperature at both the device level and at the systems level demonstrate that the safety related and non-safety related systems are operable. It is i therefore concluded that the systems located inside containment will perform properly during normal and abnormal operating conditions and maintain safety margins during accident conditions.

21 L_ _ _ _ _ _ _ _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ _ _ _ ______ _ _ _ __ J

e ** '4

>- e

& a W> e mm e (Ze Mm e de O e M M wM e ~ M~ ~ M~ ~M-O 2M

= e e.

OM M

s 3a -

M OM wa M

M 5"' a OM8M wa M

a- N M O

ww3w Me~e waSa Mw-w OJ e =N >m =N -De =N>m W W -4>4 -MDM

> e w e-M4 e >N we >N we >mw@ r H Dewe w og e w eg e

<2 e w e g e w e g e w og e Q O E<* m@ A4 EC AM mMa@ 2 2 mO4C mehe U C O g w e - =

w 4 e M M M

MMe A g e O M ps - 6a EJe e O w M w w Ud e m Q & L w @ 4 W4 e M W O O e C

>> e N e 2 4 2 @ e

= w M M && AA

== wm m == 4 4 aww aww MM MM -M MM = = WMM wMM Af6MAh M&Maa MMm MMM MN4MHNm M M O 3 O& 3 0430 0A34 OmWW3Nem Ow3m OADA e WMMOmm W MOmm w =O e w O we e eDe *

  • WNON wwOm JE o MNN=dd MMN=44 M$=O M4-w M eCe= eOO MM-M MM=M 40* ~
  • e> e * =N e> * * ~4>d =@>O =@ew>@ee we>C -4>@

kme >mMweg > eWwe@ De we >MwM W eg a De o e we a e w%%%g%%%

> ow e su e= g e

>e we wNg%

OM e WNCmMw W@dWed W%g%

WW e g+ e A+ e g+ e &+ e m+4+ MO&O m+++a+++ ue A s g+&+

= m m w w -

M M M M M M M M MW e b & b 4 4 4 & & A&& 441 20e + e www www w w wee M Me M Mm MMm @Me MMMMMMM N N M Emo ON 34 ON QW C430 O@30 OmNv3NNm OM M D13a Dwa w e O e w e U e w eO e wMOM wvevoeem we C wwow u e @@ ~O Mc ~O M4mo M em e M e o em e e e M e~ e MMmM

> e =N >M =N >M ~4>v -ODO ~NWe>NOW -e>e ~W>4 M e De we De we >e we > e we > e e e y e e e De we D e we w%g%

2 e W% MN w% g% w%E% wNm% w%%%r.%%% w%mN ad e m+ 4+ W+ 4+ K+4+ m+A+ M+ ++A+++ E+h+ m+4+

M e M e = = O O e M e M M e e

>- WW e A A & A w e e

UOs ww ww m C @ < < > < >

A e

e e

'ame Ous AW e e

e c

o 2

2 NO N2 2

MO N2 e e M e

> M e e >e m e Omo E e eMue woe -e se e =ma e Q O O>we Me = =

e Zw e n e M M A 4g e WM s A & e N M e w

e >> e w ww M w e wse O m W w M > > > OW N QJ e e wd e 4 4 N O O O NO e 42 e 42 e 2

g e Md e N e 2 2 M e

- e

< e -

2 2 2 2 2 2 A e H e 2 m w w e 2 e w w w w e we O Q O e E , _O _ _ _C _O _O _

2 . U U u U U U U Oe U U U U U U U e < < < < < < <

_e s:

we 5-2 5

2 5

2 5

2 5

2 5

2 5

2 5:

_. a a a a a a a me - - - -

E & E

( e E E M N Ds w w w > > > >

> e M M M U S X X X M

( s M M M E E m E b

D M e

A 4 K M N O e

a e e W 8 e t

$ e WMm WMm Mame me > Ma CA Q= EN@ EN@ > N>ee 30^ w Q ee eN

> DOM QOM 2 OMO4 M4M U L 04 me U Meo weO 4 eCeQ MNO 2

  • eo Om M e e 3J e e e e gwe V d em M e e e O 2KMe e DOw>e www e O > e we Om>N a MN O WW e we OMMW M&&
  • 4 s e W MMW 2W QMbb d M =Z w> U4m &.M 4 e- A 4 we + 4.N om U => a3 C eA U 3 **

>N = 3@ =

  • N 2 e E4 e
  • U e = w = (OOA E N e O@ wwdeo so-OC wwwNm <>NO sWmNmec w20 = -

m2mCV wee 2% =04* MMZOM w2OM mO3OMO4 m-VN eo M M e M <ec mdeo (>AeDec QNO M e e dweO Om e e se e =UM e e e e Zae 4 3po Es4 e e Z e O 00 U s >Z O=>N OO e COZW e Owe

@% 2 e MO= > a 3 >e m(wk e we mOwmW m2W MN De g> Q AM (wMa>>w =3AW M4W wA Emw4 mw2 Zumwr Ew2 aw w e e EUw&

9 a >ghwaw>

g -s e ,

> e

>= a W Da e h me e

<Ee M. e O .J ee M

>- M e e M ==

== 8 O M D En O 3 2M e W in an w O

> e W ll6 .O CL an M3 == 3 0 =.J e w == N D N w uk == 4 >e=

ag eC e >- >QwO >= >= >N wQ

<2 e O w e er e O O w

  • ot e E<e 2 cc o EL O 2 2 as N C6 N U G W 4 =

g i C6 e in O N i

fn M e 4 CL O == "

We == M S 23e W 4 in C6 art W e>-

WJ e e- N CL W et

  • O in O MO

>> e 2 e- 4 M ei" E e.= e to en == *i=

gh C6 M M *= ==

EL C6 M M M M to CL M C6 M O Ok O' O @ D *# O 3 O 3 e d w e a w4ON w Grt O in wsaOss a.J & e an te == to to e e e en = = = = M =# -t - Ch an 4O6 w == v > v ==*=>e= == 4 > @ == N @ > =# 4

> e =w e e e- as e ei- >a we > e we > a wa C as e O w % as % saa % or N w N ec % w @ M esen M

>= w e 2 CK + CL + GC + C6 + Es + EL + CK + e C6 + e

== em en in == me

>* & 6 CL CL M M *= *=e 2Oe Ch CL M to W oc e M M M CL M CL 3 2 E CK e O Ch O @ D =# O 3 O in Dwe 4 Les a C*= wVON w 4 gg a ga gn == grg M e w ane==

in Q in. M e e

>- e W == .s > *a = = =e> = = = = = == @ > @ == M ==>M in a w > e wa > e we > e we >e we 2 e O w % cc % w % cc % w%m% w% ac N eg +

===J e 2 sg e C6 + oc + CL + cs + CL + C6 +

N en e 3 e M w O e in e O D >=

==

  • 44 e ws O3 0

> e w& s .En.e O >O

-s 2 w e uOe ac e O or e et et et et > e w aw g6 e og oc e  %  %  %  % W *= tg N w 6 db w e 2 2 2 X En + C6 + in e e M e

> M e e

t.i=.

. e e

Qg e 6 teJ Q

  • E8 e
    • sn at e ==.Je e . og e e d O>we en e ==

Zw e *= a an aC sg e > En e 4 EL in an 80 a >->e w e w .J e w O w W W

>=

Oae h4 e e.- e3 >- >=

  • st 2 e O =# O O E O g e M< a 2 e 2 2 @2 e

.M a CK= e

( e >= e- e=

CL e >- e 2 2 2 e 2 e w w taJ e We Q Q Q e Ea == >- == >= ==

2 e U 2 U 2 U O e U w 06 u w 06 U

< Q et O <

.cc

= ee -W e.= w

>a 2 U >= 2 U ** 2 O

2 e O UO O UO we 2 et 2 2 eC 2 2

!E E U > m. O_5 -

w U #= 2 >= w

-s w < == eC > >=

4 2 ~s .4 .J et

====2 Q O os ~3 in M CL Q w == U == M w QC laJ W D e- >=>= ba h M 2 22 2 2 e in W ww W >

Oe WME EE E O

a. e CL 22 2 2

>= 4 == cc og (L w 2.=

a = = = = *= w e fg b eC 4C eC 4 0 e >= Li e- >->= >= @ oc

>= a 03 2 2 w U e M =* Q Q Q 14J K eE e et 2 =J U U >= w O

2 M U eC ==

sn w tak W H H e 2 m w = w = 2 =

O^ O wN oc @ os @ O E0 * =^

e* N == CK N DC DO == +- N N N en *= e= 3O M *# M4 >= W@N N O in U M e* MN in N U >= N @ @

e O 2 M e W uJ e Ewe 2 et

  • NNN M LtJ e D & w >= O oc >= O cm >= D ==>=aeo C62 MJ b O QC Qe D- Cao (L

>= (L C6*

6

  • >=J>=>=>=
  • 2 emCL
  • U
  • U * =J =.3 aJ
  • N 2 e 06
  • EU O^ et 0 in a eC 0 an a E O 2" O to Oe O e- w ec >- N O w2ON w2ON w en e Ne- *== N e-=*

2% == e  % en e= e- w 2 O s2 2 ~ st O CC~4O >=> @N (V O M >=

  • U e Of O to e O En GC eC ra O

> e e=

O N wf 2 .e e N Q N =J 2 a.e N inO=JN@Qk

>= we NNN v@

D% 2 e We 44 3 >= e O == e== e O == e.e-e M 3 > >= e e e SN De 2 LaJ Q & >= *= 3 1 e'* =* 3 4 em Q idJ d >* ** e" C6 ** b o w2 e .4 U w dL 2 CD w (L 2 CD w C6 e J =J w =J =J =J e e

4 .. ,

NOTES FOR LOOP ERROR TABLE l 1. The Tech Spec setpoint for the variable low pressure trip results from l ' the formula 11.75 Tout - 5103 psig. The safety analysis limit for variable low pressure trip is defined by Figure 2.1-3 of the Technical Specifications.

2. The High Reactor Building Pressure trip is not taken credit for in the safety analysis. As such, no safety analysis limit has been established.
3. The Power / Imbalance / Flow trip is based on a set of curves in the Tech Specs. There is still a 0.45% FP margin for the 3 pump cperation and a 0.42% FP margin for the two pump operation.
4. The Power / Pumps trip is essentially digital. As the temperature does not affect the power signals, the trip function has no associated error.
5. The ESAS High RB pressure and High High RB pressure Tech Spec limits do not have an associated analysis limit. As such, no margin is specified.
6. The EFIC high range transmitters are not included because the functions they control and associated setpoints are based on accident errors. Additionally the EFIC steam generator pressure transmitters are not included because they are located outside the reactor building.
7. This error is due to reference leg heatup. The previous process error was overly conservative.
8. Level is referenced above the lower tube sheet. This level is required so that initiation takes place within the range of the instrument.

(Not a safety analysis limit).

9. Accident instrument errors are not affected by small char.ges in operating temperatures.
10. This is based on the heat balance equation calibration.
11. This is the assumed safety analysis limit in the computer simulations. However, in all accident analyses, this setpoint was never reached as the RX trip mechanism.
12. Level is referenced above the lower tube sheet. This setpoint is established in the SAR, not the technical specifications and is based on B&W's position on R.G. 1.105, which involves not setting setpoints within the upper or lower 5% of instrument ranges, i Changed to SRSS methodology.

r- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --- n-l The JC0 listed seven LERs involving instrument drift submitted in the last 13 years. Of the seven, one was a mechanical problem incorrectly listed as related to instrument drift, another involved equipment located in a cantrolled environment outside containment, and a third involved loss of fluid .in the reference legs of level transmitters. The remaining four events involved pressure transmitters that, according to the temperature profile in the JCO, were operating in ambient temperature environments that were well within manufacturers' design ratings. These transmitters have since been. replaced with late model transmitters that are operating successfully in an environment that is well within their design temperature limits. A description of each event is attached, t

4 . . . . .

Description of Events LER 74-08 Following a reactor / turbine trip test from 40% power an investigation was made to determine the cause of the discrepancies between readings of the core flood tank level instrumentation. Instrumentation was giving an increasing level indication for several days indicating reactor coolant in leakage and coolant had been removed from both core flood tanks to stay within Technical Specification limits (13 ft t 0.4 ft.). Level instrumentation on both tanks were found to have lost liquid from the reference leg causing instrumentation to drift high giving a false level increase indication. When reference legs were properly established, core flood tank "A" level was found to be 12.4 f t, which was 0.2 ft. lower than the Technical Specification limit.

This problem was unrelated to elevated ambient temperatures. At core flood tank pressures, the observed magnitude of ambient tensperature increase above design temperatures could not have caused fluid evaporation in the reference legs.

LER 76-03 While making preparation for removing surveillance specimens for inspection of holder tubes due to wear at other B&W plants, it was found that spring cartridge was missing from two tubes and one of those two tubes was partially missing along with push rod and journal bearing. Search for missing pieces initiated (50-313/76-3).

This LER was incorrectly listed in the JC0 as being instrument drift related.

LER 83-03 On 1/24/83, while in refueling shutdown, a Reactor Building pressure transmitter (PT-2405) was found to be out of specified tolerance.

Engineered Safeguards Actuation Systems (ESAS) Analog Channel 1 Calibration indicated that PT-2405 had zero-shifted approximately 1.125% low. On 1/26/83, while performing ESAS Analog Channel 2 calibration, another Reactor Building pressure transmitter (PT-2406) was found to have zero-shifted by approximately 1.25% low. These shifts in transmitted outputs would have resulted in late ESAS actuation of approximately .45 psi for PT-2405 and .50 psi for PT-2406. Part of the followup actions to the events described above included a recheck of the pressure transmitter setpoints during the next cold shutdown of sufficient duration. This recheck was accomplished. on 7/12/83, at which time it was found that the output of ESAS pressure transmitters PT-2406 and PT-2407 had drifted such that actuation setpoints would have occurred late. The containment isolation actuation signal would '

have occurred at approximately 19.06 psia for PT-2406 and 18.84 psia for 1

__ __ - - - _ _ _ . . _ _ _ ._ _. _ -__--_________-_a

PT-2407 (Technical Specification (T.S.) limit 18.7 psia). The reactor building spray actuation signal would have occurred at 45.24 psia for PT-2406 (T.S. 6.12.3.2.a.). No similar occurrences have been reported regarding these particular transmitters. On 7/25/83, the actuation '

setpoints were reduced from 18.6 psia to 18.03 psia and from 44.6 psia to 44.03 psia to provide additional allowance for drift.

These pressure transmitters were calibrated during a mid-cycle maintenance outage that occurred in March, 1984. None of the errors found during calibration would have resulted in a late Engineered Safeguards Actuation on {

reactor building pressure. These pressure transmitters have since been i replaced.

LER 83-04 On.2/14/83, while performing engineered safeguards actuation system (ESAS) calibrations during the refueling outage, reactor coolant system (RCS) pressure transmitters PT-1040 (Channel 3) and PT-1020 (Channel 1) were found to be out of tolerance. The technical specifications (TS) trip setpoint is 1500 psi. A trip from the output of PT-1020 would have occurred at 1486.25 psi. A trip from the output of PT-1040 would have occurred at 1491.25 psi.

This occurrence is reportable per T.S. 6.12.3.2.a. The cause of the ocurrence was transmitter drift. The transmitters were recalibrates and left within the allowable tolerance. These transmitters have since been replaced with a different make and model. There have been no problems with the new transmitters.

LER 83-06 On 3/21/83, during refueling shutdown, Reactor Coolant (RC) Pressure Transmitter PT-1023, was found to be out of tolerance during the Reactor Protective System (RPS) Channel "B" calibration. PT-1023 was found to be 2.3 psi low at 0% (1700 psi) and 6.6 psi low at 100% (2500 psi). If linear regression is used to calculate the error at the RPS trip setpoints, the low pressure trip would have occurred at 1802.84 psi, instead of 1800 psi, and the high pressure trip would have occurred at 2305.5 psi instead of 2300 psi. This would result in a conservative actuation for the low pressure trip, but a nonconservative actuation for the high pressure trip. This occurrence is reportable per Technical Specification 6.12.3.2.a. The other three pressure transmitter feeding the RPS were found to be within tolerance. PT-1023 was adjusted to within the specifications of the RPS Channel calibration procedure. All RPS instrumentation was calibrated and tested during the refueling. These transmitters have subsequently been replaced with a different make and model. There have been no problems with the new transmitters.

-4, . .

LER 83-07 On 3/8/83, while in refueling shutdown, Reactor Building (RB) pressure switch PS-2403 was found to be out of tolerance while performing the Reactor Protection System (RPS) Channel "D" calibration. Also, on 3/21/83, RB pressure switch PS-2401 was found to be out of tolerance while performing the RPS Channel "B" calibration. PS-2401 and PS-2403 were found to open at 18.9 psia, above the maximum technical Specification Value of 18.7 psia.

Since the switches would have actuated at a pressure higher than required, this -occurrence is reportable per Technical Specification (TS) 6.12.3.2.a.

RB pressure switch PS-2400 for RPS Channel "A" and RB pressure switch PS-2402 for RPS Channel "C" were found to be set at less than or equal to the value specified in T.S. The pressure switches were reset to 18.5 psia per the RPS calibration procedure and returned to service. These pressure switches have since been replaced. They are located in a controlled environment outside the reactor building.

LER 84-03 While shutdown for a planned mid-cycle outage, Instrumentation and Control (I&C) technicians were performing Engineered Safeguards Actuation System (ESAS) calibrations. On March 27,1984, at 1530, PT-1040 which provides Reactor Coolant System (RCS) pressure signal to ESAS Channel 3, was found to be out-of-tolerance such that the Channel 3 low primary system pressure actuation would have occurred at 1489 psig rather than > 1500 psig as required by Technical Specifications. On March 28, 1984, at 1430, PT-1022 which provides RCS pressure signal to ESAS Channel 2, was found out-of-tolerance such that Channel 2 actuation would have occurred at 1478.25 psig. On March 28, 1984, at 1500, PT-1020 which provides RCS pressure signal to ESAS Channel 1, was found out-of-tolerance such that the Channel 1 actuation would have occurred at 1492.5 psig. The cause of these occurrences was a nonconservative high drift of all three pressure transmitters. These transmitters have since been replaced.

I w-- _ _ -_ -