ML20024H754

Forwards Response to NRC 910402 Request for Addl Info Re Digital Channel Operational Test Definition.Radiation Monitoring Sys,Mfg by Ga Co,Consists of Detection Sys & Centralized Display & Control Sys for Radiation Monitors

ML20024H754
Person / Time
Site:	Seabrook
Issue date:	06/04/1991
From:	Feigenbaum T PUBLIC SERVICE CO. OF NEW HAMPSHIRE
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NYN-91091, TAC-79742, NUDOCS 9106100388
Download: ML20024H754 (9)

Text

V

, a

~

New Hampshire Y khh Ted C. Feigenbovm President and Chief Executive Of fim N YN- 91091 June 4,1991 United States Nuclear Regulatory Commission Washington, D.C. 20555 Attention: Document Control Desk

References:

(a) Facility Operating 1.icense No. NPF-86, Docket No. 50 443 (b) NHY Letter NYN 91010 dated Jar.uary 24,1941, " Request for License Amendment; Definitions of Digital Channel Operational Test", T. C.

Feigenbaum to USNRC (c) USNRC Letter dated April 2,1991, "Scabrook. Technical Specification Digital Channel Operability Test Definition: Request for Additional  ;

Information (TAC No. 79742) "G. E. Edison to T. C. Fei,enbaum

Subject:

Request for AdditionalInformation: Digital Channel Operational Test Definitio :

Gentlemen:

The additional information regarding the performance of a Digital Channel Operational Test and digital equipment used at Seabrook Station as requested in Peference (c) is provided in the Enclosure. This information does not change the request for a license amendment nor does it affect or change the no significant hazards consideration provided by Reference (b).

Should you have any questions regarding this matter please contact hi t. James N1.

Peschel, Regulatory Compliance hianager, ai (603) 474 9521, extension 3772.

Very truly yours, f/fY/

Ted C. Feigenbh 1

Enclosure TCF:J h1P/ssi/act //

9:106100388 910604 PDR ADOCK 05000443 P PDR New Hampshire Yankee Division of Public Service Company of New Hampshire P.O. Box 300

• Seabrook, NH 03874

• Telephone (603) 474-9521

.c v u.v' u- n _(l /

,,m,

United Stqtes Nuclear Regulatory Commission J une 4,1991 A tt<:ntion: Document Control Desk Page two cc: Mr. Thomas T. Martin Regional Administrator United States Nuclear Regulatory Commission Region 1 475 Allendale Road King of Prussia, PA 19406 Mr. Gordon E. Edison, Sr, Project Manager Project Directorate 13 Division of Reactor Projects U.S. Nuclear Regulatory Commission Washington, DC 20555 Mr. Noel Dudley NRC Senior Resident inspector P.O. !!ox 1149 Scubrook, Nil 03874 Mr. George L. 'verson, Director Office of Emergency Management State Office Park South 107 Pleasant Street Concord, Nil 03301

. . _ _ . _ . _ . - . _ _ . - _ . _ ~ . _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ . - - . _ . . _ _ _ . _ _ _ _ _ _ _ - -

9 4

New llampshire Yankee J une:. 4,1991 I!NCLOSURE TO NYN-91091 d-r t

h

'tWef*Ng T-4 m+f uyp-g=y--gwwy.p,,, gag,,mwmp.p.p.g,,qq,,p,, , , g,,, _,,, ,,_ _ , , , , _ , __ ,_ _ _ _ _ _

h Response to Request for Additional Information Digital Channel Operational Test Definition Request No. _1 Please provide a description of the digital equipment that will be affected by the proposed definition and proposed footnote changes, include a hardware configuration description and reasons, if any, that the injection of a simulated signal is not practical.

Response

New flampshire Yankee utilizes 22 radiation monitors that are governed by the Technical Specifications. Of these 22 radiation monitors,10 are Class IE and 12 are non-safety class, The radiation monitors and their classifications are provided in Figure 1. The monitors are part of the Radiation Monitoring (RM) System, manufactured by General Atomic, which contains equipment that organizes radiation monitoring data and presents a centralized display and alarms to the operator. - The system consists of three sub systems:

. Detection system.

-The detection system includes the radiation detectors and the field monitors (RM-80s hereafter called RM 80 monitors) which operate the detector. The RM 80 monitor is a microcomputer controlled electronics system which can control and process up to four indi.idual detectors. The RM 80 monitor can provide local radiation and status indication for four detectors, l

. Centralized display and control system for Class IE radiation monitors.

The centralized display and control system for Class 1E radiation monitors consists of two control room ' cabinets (CP 180A and CP 180B) and wiring between those cabinets and the Class IE monitors, Por each Class IE monitor there is one control / display module (called an RM 23) which is located in l either CP 180A or CP 180B,-depending on the train A or train H classification of the monitor. 'These cabinets provide centralized indication and control (in

~ the control room via the RM 23s) for all the Class IE radiation monitors.

. Centralized display and control system for non safety class radiation monitors.

The centralized display and control system for non-safety-class radiation monitors consists of redundant computers (called RM 11 No.1 and RM-11

, No. 2). These computers are connected to all Class 1E and non safety class

RM 80 monitors through communication wiring and provide display on cathode ray tubes (CRTF in the Control Room and at the Health Physics Checkpoint and they provide an input to the Main Plant Computer. Qualified isolators are provided in the communications loop to ensure separation between Class IE and non safety class devices, it must be noted however, that control of the Class 1E RM 80 monitors is only available from CP-180A or CP 1088, The RM 11 computers.are redundant; if one RM 11 fails, the other RM 11 assumes full system communication load with no loss of capability. The RM 11s are located in the Administration Building computer room. Every two seconds 1

. . - . _ . - _ - - _ _ _ . _ - ~ - _ . - . _ _ _ _ , _ - _ . - - --

l cach RM 80 monitor is polled by the RM 11 computers. The RM 80 monitors respond with existing radiation levels for each detector and codes that describe overall RM 80 monitor and detector status, in this way, the operator is provided with convenient, centralized display for all RM-80 monitors. The display is by way of custom consoles with CRTs, located in the Control Room and at the Health Physics Checkpoint, from which the operator may request I various organized displays of the data gathered by the RM 11s. I Eight of the Class 1E RM 80 monitors utilize Geiger-Mueller (GM) detectors which contain signal shaping and preamplifier circuitry within the detector housing. The detected signal is then routed to the RM 80 monitor and into a digital counter chip which decrements by one count for each pulse that it receives. The remaining Class 1E RM-80 monitors, the Post LOCA monitors, do not contain the digital counter chip. These RM 80 monitors convert the extremely small ion chamber current to a digital value directly within the RM 80 monitor by means of an analog to digital converter circuit.

The non-safety class RM 80 monitors receive signals from GM tubes, ion chambers and scintillation detectors. The scintillation detectors require preamplifier / discriminator circuitry

- within the RM-80 monitor. The preamplifier /discriminators are set up in a gross counting mode with the upper window discriminator inhibited and the low energy window set to meet the low energy sensitivity per the FSAR requirements for that particular RM-80 monitor.

The RM-80 monitor operates in the same manner as those connected to GM tubes after the signal passes through the preamplifier / discriminator.

The RM 80 monitor provides signal processing and control for up to four radiation detectors.

Besides signal processing, the RM 80 monitor has a wide range of communication, computation and control capabilities. These capabilities are derived f:om the design of the RM-80 monitor as a microprocessor-based computer (microcomputer).

Each RM 80 monitor is equipped with three communication ports. Two of these communication ports are used in a redundant mode to communicate information between the RM-80 monitor and the RM 11 computers. The third communication port is used on Class IE RM-80 monitors to communicate with a remote control and display module (RM-23) in the Control Room.

The microcomputer'in the RM 80 monitor is controlled by an 8-bit Intel-8085 microprocessor which is supported by a number of specially designed and compatible integrated circuits.

The operating system and application programs are all stored in read-only memory (ROM),

which is loaded at the factory. Existing detector counts per minute, conversion factors, setpoints, radiation history, and other operator-loaded information, such as RM 80 monitor and detector data base, are held in low-power random access memory (RAM). Information in RAM will be retained by battery backup even if the 120 V AC supply- to the RM 80 mcnitor is lost.

The RM-80 monitor calculates the output of each radiation detector every 600 milliseconds by means of a 16 bit digital counter which counts down from a full to empty register condition. The RM 80 monitor calculates the number of counts received since the last 600 millisecond reading and then performs a calculation to determine the counts per minute (CPM), As the counter nears a zeio count it resets and continues to decrement to ensure that the counter is not saturated and that accurate counts are obtained. The CPM which 2

l t

l t

, . - _ . - - . _ - , _ _ _ . _ ~ _ _ - . - _ _ . . , . . _ _ __ _ __- . , . _ ._.

is displayed and used for alarm purposes is a calculated value which is the counter reading processed through a software smoothing algorithm. The algorithm prevents an unstable display which would occur if only counter contents were used to provide the Rhi-80 monitor display.

As cach ' smoothed

• radiation value is processed the Rht-80 monitor performs a series of checks to verify the value is below the high and alert alarm setpoints. The alarm setpoints are set by Chemistry Department personnel or llealth Physics Department personnel depending on the Rhi-80 monitor's function. These setpoints reside within the Rhi 80 monitor in a fixed memory location. These locations each have a unique ' address' in R Ahi (Random Access hie mory). The R ht.80 monitor performs a comparison every 600 milliseconds of the new radiation value to the high or alert alarm setpoints. If the alarm setpoint is exceeded the Rht-80 monitor executes the actions required for the alarm.

The New Hampshire Yankee radiation monitoring system, since it is a digital system, does not experience the phenomenon of setpoint drift that is common to analog systems. The repeatability of alarm setpoints is discussed in the answer to question 3.

Since the setpoints do not drift, and because values located in microcomputer R Ah1 locations are verified on a periodic basis during surveillance testing, the injection of a signal is not required to verify equipment operation. There are three major reasons for performing testing by manipulating the setpoints. One is to minimire the potential damage to the detectors and Rht-80 monitors that may occur when the test equipment is connected, such as broken leads.

The signal injection method requires either a coaxial or lugged connection to be removed to allow connection of the signal generator which creates the potential for equipment damage. It must also be noted that the utilization of a signal generator does not maintain i.no ndence matching of the circuit. Secondly, some of the Rht-80 moni:ars are located in

'ur radiologically controlled area (RCA). The practice of periodic testing by signal injection is contrary to NHY's As-Low- As-Reasonably-Achievable program for radiation dose control as it requires technicians to spend more time in the RCA. Third, the data base manipulation method does not require the detector to be disconnected from the Rhi-80 monitor and to be physically removed from service so that the system could not detect radiation. Additionally, since a smoothing algorithm is used in the RN1-80 monitor it is exceedingly difficult to inject a signal at the snne rate every time such that exact repeatability of a test cannot be assured by the signal injection method.

Request No.2 Please provide a description of the surveillance test and a description of the usage of the " user command" that changes the setpoints, include specifics on the independent software verifications and the computer printout check of setooints and system functions. If the surveillance procedure n 'f is provided please highlight and explain the procedure sections that pertain to the NRC questions.

Response

During the performance of a surveillance test the alarm setpoint of the RN1-80 monitor ;s ,

decreased to a value less than background radiation levels by an 1&C technician stationed at a Control Room console. As the setpoint is entered the desired value is displayed to ensure the 1&C technician has typed the correct setpoint value. Once the I&C technician has keyed in the correct slue he depresses the enter key, the display blanks out and then returns again showing the same value. What actually happens is the keyboard entered value 3

1 on the display screen is sent to the RM-80 monitor and the display blanks out while the value is being transmitted to the RM 80 monitor. The Itht 80 monitor receives the value, evaluates the format, and stores it in the R AM location reserved for the setpoint. The itM-80 monitor then ' reads' the newly stored value and sends it back to the 1&C technician screen. The 1&C technician actually sees the value stored in R AM, not just the value he typed. If the storage area in RAM was defective, the value would have been corrupted and the l&C technician would have seen a value return other than that which he had transmitted.

Upon completion of this data entry the I&C technician ve ifies that the RM-80 monitor's alarm saftware functioned properly by confirmation of system alarms, illumination of lights and the energi7ation of relays controlled by the alarm function.

After verification of the alarm function the 1&C technician restores the alarm setpoint value to its original value utilizing the same process as described above. The I&C technician will again verify that the appropriate value is stored in the RAM. A second individual will independently verify that the correct value is stored in the R AM. Additionally, the I&C technician will verify via a computer printout that only the alarm setpoint was manipulated during the sur cillance test and that the original value has been appropriately re entered in the R AM.

Request No, 3 Please provide a basis of the adequacy of ihe current test as it pcriains to testing as much of the system as possible, including the channel input circuitry and any signal conditioning circuitry, include specifics on the check source and self diagnostics.

Response

Each GM detector contains a Cl-36 source assembly which automatically energizes every 24 to 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. The source tests the detector and associated amplifier and counter circuitry of the microprocessor in each RM-80 monitor, if the check source test results in no counts or low counts the RM-80 monitor fails and an annunciation occurs in the control room.

The Post LOCA monitors are provided with a continuous ' keep alive' current from a uranium source contained within each detector. The checksource used by these detectors injects a current automatically every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The checksource tests all of the circuitry f om the detector input to the RM-80 monitor and the associated amplifiers and analog to digital conversion circuitry of the RM-80 monitor.

The checksource circuitry for the scintillation detectors consists of a Cl-36 or Cs-137 source which actuates every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The source tests the detector and associated amplifier ano counter circuitry of the microprocessor of the affected RM 80 monitor. If the checksource test results in no counts or low counts the RM-80 monitor fails and an annunciation occurs in the Control Room. ,

During startup testing of the radiation monitoring system each RM-80 monitor was thoroughly tested by a combination of signal injcetion, source testing, and data manipulation.

Signal injection was performed on each RM-80 munitor during startup testing. The test results are available for review at % brook Station During these tests, pulses or currents were injected into the RMC monitors to verify the software algorithms. The software for GM detectors automucally corrects for nonlinearity of the GM tube. In this type of RM-80 4

monitor the calculation results were verified over the full range of each detector. For lon chambers, currents were injected over the full range and response was verified to expected response data. For scintillation detectors, pulses were injected and a linear response up to the maximum design specifications was verified.

The startup testing performed at Seabrook Station verified the proper operation of the Rht-80 monitor software and provided the baseline data for their operation. Once the proper operation of the software was verified it responded exactly the same every time.

No drifting of setpoints or repeatability problems were experienced since the circuits are digital.

Each type of detector was " type" tested using sources up to 1000R to verif y the proper operation of detector over-range saturation circuitry. Design flaws on one type of detector were discovered which were corrected by the vendor.

Primary calibrations were performed at Seabrook Station during startup teving of all Class IE Rht-80 monitors and non safety class R ht-80 monitors required by .be Technical Specifications. For these RM 80 monitors, sample geometries were designed which were identical to the monitor samplers for liquid monitors. The calibration geometries cere sent off to be filled with a National Bureau of Standards (NBS) traceable source m a erial.

Airborne monitors were calibrated using an NBS traceable gas. Particulate monitors w ere calibrated with a filter paper embedded with an NBS traceable source. Gh1 and ion chamber detectors were calibrated on-site with an NBS traceable Shepherd Beam Irradiator. These cabbrations established and verified detector conversion factors.

The decision to perform Digital Channel Operational Test (DCOT) testing by setting the alarm setpoint below the background radiation value was based on baseline test data obtained during startup testing and the demonstrated operation of the radiation detectors and Rht-80 monitors. The injection of a signal does not prove energy response since the preamplifiers are nonlinear devices and impedance matching of the actual detector to the circuit is not maintained when a sio,nal generator is connected. Since the alarm setpoint is in a tixed memory location it is not necessary to test the actual value of the setpoint. What must be tested is the abilly of the RM-80 monitor to enter the alarm software task once the actual radiation value is above the setpoint. By signal injection an actual radiation value is not used and the risk of damaging equipment by lifting leads or injecting the wrong voltage is increased. Additionally, the injection process exposes personnel and equipment to the RCA environment unnecessarily and makes the task more complex than required. Finally, a signal injection test does not provide any additional usable data other than that obtained by lowering the alarm setpoint.

Our present DCOT testing verifies Rht 80 monitor operation by setting the alarm setpoint to a value which is less than the actual background radiation value. The RM-80 monitor operation is verified without disabling the RM-80 monitor's ability to respond to any radiation which actually exists since the detector is not physically removed from service.

On a refueling outage basis cach channel is calibrated utilizing an NBS traceable source positioned by fixed geometry test apparatus. This type of calibration ensures accuracy and repeatability and provides a much more accurate calibration and check of the system than the signal injection method.

5

2 Request No. 4 Please provide specifics on the manufacturer recommendations and current known industry testing practices on this equipment.

Response

l The manufacturer, General Atomic, has reviewed the NilY surveillance testing methodology and agreed that the setpoint change method to test RM-80 monitor response is acceptable.

They also stated that the injection method does not provide any additional test data on a

-periodic basis, but must be performed as an initial-check to verify software response. NHY perfoi.ned the sigi.nl injection testing during startup testing.

New llampshire Yankee contacted several other utilities to determine what types of tests  ;

were performed for surveillance and calibration testing of digital equipment. There is no established industry standard and the practice ranges from signal injection all of the time l to never using signal injection. The individual plant decisions appear to be based upon the preference of the plant's staff.

6 l

l 4

., - , ,- . , ~ . ,. - . - _ _ _ _ _ - _ _ . _ - _

FIGURE 1 R ADI ATION MONITORS TECHNICAL MONITOR NOMENCLATURE SPECIFICATION Cl. ASS 1R ,

1 RM RM 6506A Control Room East Air intake, Train A 4.3.3.1 YES 1 RM RM 6506B Control Room East Air Intake, Train B 4.3.3.1 YES 1 RM-RM-6507A Control Room West Air Intake, Train A 4,3.3.1 YES 1 RM-RM-6507B Control Room West Air intake, Train B 4.3.3.1 YES 1 RM RM 6527A Containment On Line Purge, Train A 4.3.3.1 YES 1 RM-RM 6527B Containment On Line Purge, Train B 4.3.3.1 YES 1-RM RM 6535A Fuel Manipulator Crane, Train A 4.3.3.1 YES 1-R M-R M-653513 Fuel Manipulator Crane, Train B 4.3.3.1 YES

- 1 RM-RM 6576A Containment Post LOCA Monitor, Train A 4.3.3.1 & 4.3.3.6 YES 1-RM RM-6576B Containment Post LOCA Monitor, Train B 4.3.3.1 & 4.3.3.6 YES 1 RM-RM 6528 Plant Vent Stack Wide Range Gas 4.3.3.10 NO l RM RM 6503 Carbon Delay Beds Outlet 4.3.3,10 NO-1-RM RM-6504 Waste Gas Compressors Discharge 4.3.3.10 NO

- 1 RM-RM-6515 Primary Component Cooling Water Loop B 4.3.3.1 & 4.3.3.9 NO 1-R'M 'RM-6516 Primary Component Cooling Water Loop i . 4.3.3.1 & 4.3.3.9 NO 1 RM-RM-6519 Steam Generator Blowdown Flash Tank 4.3.3.1 & 4.3.3.9 NO Discharge 1-RM RM 6509 Waste Liquid Test Tanks Discharge 4.3.3.9 NO 1 RM-RM-6526 . Containment Atmosphere 4.4.6.la -NO 1-RM RM-6562 Fuel Storage Building Ventilation Exhaust 4.3.3.9 NO RM RM-6481 Main Steam Line Loops 1 & 4 4.3.3.1 NO l'- 1 RM-RM-6482 Main Steam Line Loops 2 & 3 4.3.3.1 NO 1-RM-RM-6521 Turbine Building Sump Pumps Discharge 4.3.3.9 NO l