ML20140H173

Rev 1 to Davis-Besse Course of Action

ML20140H173
Person / Time
Site:	Davis Besse
Issue date:	10/01/1985
From:	Williams J TOLEDO EDISON CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
1188, PROC-851001, NUDOCS 8510090009
Download: ML20140H173 (96)

Text

/

• DOCUMENT CONTROL RECEIPT ATTACHED DOCUMENT TITLE: DAVIS-BESSE COURSE OF ACTION (COA) COPY NO. RECIPIENT 8 R. SCHUERGER 15 US NRC DATE SENT: OCTOBER 1, 1985 16 US NRC 17 US NRC ATTACHED REFERENCE, IF APPLICABLE: 18 US NRC 19 US NRC REMOVE & DESTROY INSERT 20 US NRC, REGION III 21 US NRC, REGION III 34 M. FERTE TABLE C.I.l.1 PAGE 2 TABLE C.I.l.1 PAGE 2 37 J. SILBERT INITIAL ISSUE REV. 1 39 R. SEIBERLING FROM SECTION C.l.1 INTO SECTION C.I.1 47 B. BORSUM 48 J. TAYLOR FINDINGS, CORRECTIVE FINDINGS, CORRECTIVE 52 US NRC, WASH., D.C.

ACTIONS & GENERIC ACTIONS & GENERIC 53 US NRC, WASH., D.C.

IMPLICATIONS REPORTS IMPLICATIONS REPORTS 54 US NRC, WASH., D.C.

FROM SECTION C.1.1 INTO SECTION C.1.1 55 US NRC, WASH., D.C.

PLAN NO. REV. PLAN NO. REV. 56 US NRC, WASH., D.C.

lA & IB/lC 1 1A & IB/lc 2 57 US NRC, WASH., D.C.

58 US NRC, WASH. D.C.

15A 2 59 US NRC, WASH., D.C.

INITIAL ISSUE INSERT 60 US NRC, WASH., D.C.

(BEHIND PLAN NO. 9A/9B) 61 US NRC, WASH., D.C.

15B 2 62 US E C, W H., D.C.

INITIAL ISSUE INSERT 63 US RC, W H., D.C.

64 US NRC, WASH., D.C.

(BEHIND PLAN NO. 1SA) 65 US NRC, WASH., D.C.

26 3 26 4 66 US NRC, WASH., D.C.

67 US NRC, WASH., D.C.

68 US NRC, WASH., D.C.

69 US NRC, WASH., D.C.

70 US NRC, WASH., D.C.

71 US NRC, WASH., D.C.

73 G. DOMALESKI 75 J. FAY I ACKNOWLEDG RECEIPT OF THE INITIAL ISSUE OR REVISION (S) 76 J. ENOS OF THE ABOV" IMBERED DOCUMENT (S) AND ACKNOWLEDGE THAT 77 E. SIMPSON THE SUPE S a.u DOCUMENT (S) HAS/HAVE BEEN DESTROYED OR 78 R. DIETERICK STAMPEP SUPERSEDED. 79 N. RUTHERFORD 80 J. LANGENBACH SIGNiTURE: 81 A. HOSI.ER 8510090009 85 ADOCK 0500 PDR 46 PDR 82 D. CLIFT DA74: s 83 R. GANTHER 87 E. INGE

• THIS FORM MUST BE SIGNED AND RETURNED IN THIRTY (30) WORKING 88 E. INGE DAYS. 89 E. INGE 90 E. INGE RETURN TO: THE TOLEDO EDISON COMPANY 91 E. INGE ENGINEERING SERVICES DEPARTMENT, STOP 3047 92 E. INGE 300 MADISON AVENUE 100 D. SHERMAN TOLEDO, OHIO 43652 101 E. RAY 102 MPR ASSOCIATES, INC.

DOCUMENT CONTROL RECEIPT FORM /26 g g k .

PAGE1 EFFECTIVE PAGES D-B COURSE OF ACTION REV. O 01-Oct-85 A

f VOLUME 1 EFFECTIVE PAGES REVISION

\~ 2 ======== =============== ========

TABLE OF CONTENTS PAGES i - iv INITIAL ISSUE INTRODUCTION PAGES 01 - 08 INITIAL ISSUE APPENDIX A PAGES 09 - 11 INITIAL ISSUE APPENDIX B.1 PAGES 12 - 35 INITIAL ISSUE APPENDIX B.2 PAGES 36 - 37 INITIAL ISSUE APiENDIX B.3 PAGES 38 - 46 INITIAL ISSUE APPENDIX C.1 PAGES 47 - 56 INITIAL ISSUE APPENDIX C.2 PAGES 57 - 63 INITIAL ISSUE APPENDIX C.3 PAGES 64 - 72 INITIAL ISSUE APPENDIX C.4 PAGES 73 - 74 INITIAL ISSUE APPENDIX C.5 PAGES 75 - 77 INITIAL ISSUE APPENDIX C.6 PAGES 78 - 80 INITIAL ISSUE APPENDIX C.7 PAGES 81 - 91 INITIAL ISSUE CONCLUSIONS PAGES 92 - 93 INITIAL ISSUE REFERENCES 01 PAGE INITIAL ISSUE ACRONYM GLOSSARY PAGES 01 - 02 INITIAL ISSUE VOLUME 2

==

TABLE OF CONTENTS PAGES i - iv INITIAL ISSUE IV APPENDICES 01 PAGE INITIAL ISSUE

[ APPENDIX B.1.1 PAGE 01 INITIAL ISSUE D-B MGT. STRUC. ORG. CHART 01 PAGE INITIAL ISSUE NUC. ENG. DIV. ORG. CHART 01 PAGE INITIAL ISSUE PLANT MAINT. ORG. CHART 01 PAGE INITIAL ISSUE APPENDIX B.2.1 PAGES 01 - 11 INITIAL ISSUE APPENDIX C.1.1 PAGE 01 INITIAL ISSUE APPENDIX C.1.1 PAGE O2 REV. 1 FINDINGS, CORRECTIVE ACTIONS, AND GENERIC IMPLICATIONS REPORTS CURRENT REVISION PLAN NO. 1A L 1B/1C REV. 1 PLAN NO. 1D REV. 2 PLAN NO. 5, 6&7 REV. 2 PLAN NO. 8 REV. 2 PLAN NO. 9A/9P REV. 2 PLAN NO. 15A REV. 2 PLAN NO. ISB REV. 2 PLAN NO. 16 REV. 2 PLAN NO. 18 REV. 2 PLAN NO. 26 REV. 4 PLAN NO. 27 REV. 3 APPENDIX C.1.2 PAGE 01 INITIAL ISSUE FINDINGS, CORRECTIVE ACTIONS AND i

GENERIC IMPLICATIONS REPORT PLAN NO. 12 REV. 2 j -

APPENDIX C.I.3 PAGE 01 INITIAL ISSUE

FINDINGS, CORRECTIVE ACTIONS AND GENERIC IMPLICATIONS REPORT PLAN NO. 10 REV. 2 d

I PAGE2 01-Oct-85 EFFECTIVE PAGES D-B COURSE OF ACTION REV. O O

) APPENDIX C.1.4 APPENDIX C.2.1 PAGES 01 - 02 PAGES 01 - 12 INITIAL INITIAL ISSUE ISSUE APPENDIX C.2.2 PAGES 01 - 04 INITIAL ISSUE APPENDIX C.2.3 PAGE 01 INITIAL ISSUE APPENDIX C.3.1 PAGES 01 - 12 FIGURE C.3.1.1 THRU INITIAL ISSUE FIGURE C.3.1.17 APPENDIX C.3.2 PAGE 01 INITIAL ISSUE COVER LETTER DATED 7-16-85 INITIAL ISSUE PAGES 01-34 & 66 INITIAL ISSUE PAGES 35 - 85 NOT ISSUED

• APPENDIX C.3.3 PAGE 01 INITIAL ISSUE APPENDIX C.4.1 PAGES 01 - 25 INITIAL ISSUE APPENDIX C.4.2 PAGES 01 - 04 INITIAL ISSUE APPENDIX C.5.1 PAGE 01 INITIAL ISSUE APPENDIX C.7.1 PAGE 01 INITIAL ISSUE APPENDIX III.1 PAGE 01 INITIAL ISSUE I.A PAGES 01 - 02 INITIAL ISSUE I.B PAGE 03 INITIAL ISSUE II.A PAGE 04 INITIAL ISSUE II.B PAGE 05 INITIAL ISSUE III.A PAGE 06 INITIAL ISSUE IV.A PAGE 07 INITIAL ISSUE APPENDIX III.2 PAGE 01 INITIAL ISSUE

/'~'

SUMMARY

SCHEDULE FLOWCHART 01 PAGE INITIAL ISSUE WORKING SCHUDULE SHEETS 01 - 03 INITIAL ISSUE

• CONTAINS PROPRIETARY INFORMATION NOTE: INITIAL ISSUE DOES NOT CONTAIN FOOTNOTES c)

TABLE C.1.1.1 INDEX OF FINDINGS, CORRECTIVE ACTIONS AND GENERIC

/ IMPLICATIONS REPORTS Action Plan No. Revision Title i 1A, 1B, IC 2 Auxiliary Feed Pumps Overspeed Trips -

ID 2 AFPT Overspeed Trip Throttle Valve Problem

- Preliminary 5,6,7 2 SFRCS Trip /MSIV Closure - Preliminary 8 2 Davis-Besse Main Feed Pump Turbine and Control System Failures 9A, 9B 2 Davis-Besse Turbine Bypass Valve Actuator Failure 101 2 Pilot Operated Relief Valve (PORV)

Operation 122 2 Auxiliary Feedwater System Valves AF599 and AF608 ISA 2 Source Range Nuclear Instrumentation NI-1, Channel 2 g ISB 2 Source Range Nuclear Instrumentation NI-2, Channel 1 16 2 Main Steam Header Pressure - Preliminary 18 2 Startup Feedwater Valve SP-7A Problem Analysis 26 4 Service Water Transfer l1 27 3 Main Steam System Valve MS-106 1

This report is provided in Appendix C.I.3 2 This report is provided in Appendix C.I.2 l

l Appendix C.1.1 2 Rev. 1

< J

FINDINGS, CORRECTIVE ACTIONS, AND GENERIC IMPLICATIONS REPORT I

TITLE: Overspeed Trips of the Auxiliary Feed Pump Turbines on 6/9/85 at Toledo Edison's Davis-Besse Nuclear Power Station REPORT BY: Dan Wilczynski PLAN NO. lA & IB/1C Chu:k Rupp PAGE 1 of 19 WRITTEN APPROVED REV DATE REASON FOR REVISION BY BY 0 8/31/85 Initial Issue Wilczynski L. A. Grime Rupp 1 9/09/85 General Correction Wilczynski Rupp s ,

2 9/16/85 General Corrections Wilczynski Rupp ,

Ns 4

1 I

TABLE OF CONTENTS I. INTRODUCTION 3 A. Purpose 3 B. Event Description 3 l

! II. BASIC PRINCIPLE OF OPERATION 4 III.

SUMMARY

OF TROUBLESHOOTING AND INVESTIGATIVE ACTIONS 4 l A. Change Analysis 4 i

s B. Data Search 5

1. Investigation of Other Utility Experience With Overspeed Trips 5

2. Vendor Experience 7

3. Toledo Edison Experience 8 l C. Field Actions 8 4

D. Analysis 9

1. Amount of Water Available 9 I

! 2. Transient Flow Analysis 10 I IV. RESULTS/ CONCLUSIONS OF FINDINGS 10 l

l A. Direct and Root Cause 10 t

l B. Disproved Hypotheses 10 l V. TECHNICAL JUSTIFICATION OF FINDINGS 11 f VI. SPECIFIC CORRECTIVE ACTIONS 12 i

A. Required Corrective Action 12 l2 B. Planned Additional Actions 13 VII. GENERIC IMPLICATIONS 14 j VIII. REFERENCES 14 FIGURES

1. Schematic Representation of AFPT Steam Piping System 15

2. Typical Speed Graph of AFPT Start With Cold Piping 16

3. Typical Speed Graph of AFPT Start With Hot Piping 17 1 4. Computer Model of AFPT #1 Parameters on 6/9/85 18

5. Computer Model of AFFT #2 Parameters on 6/9/85 19 J

!  ! I. INTRODUCTION A. Purpose This report documents the troubleshooting and investigative l

actions performed to identify the root cause of the overspeed i trips of the Auxiliary Feed Pump Turbines (AFPTs) at Davis-Besse on 6/9/85. Also presented in this report are the proposed l

corrective actions to eliminate the cause of the overspeed trips 1 and a summary of the proposed confirmatory testing to ensure proper system operation.

Mode 3 testing which attempts to recreate the overspeed conditions i is no longer considered to be necessary or advisable. This I conclusion is based on several factors

1) Induction of water can be' defined as root cause. Discussions  !

with the vendor and the experience at other plants utilizing Terry Turbines provide confidence in the root cause identifica-i tion. Several other plants have experienced similar j problems with AFPT overspeed and have implemented corrective actions to eliminate water induction. '

i

2) Possible equipment daname. Prior to 6/9/85, Toledo Edison i discovered damage to pipe supports on the steam inlet lines 3

! to the AFPTs which was attributed to the water slugs I created by steam condensation in the lines. Recreation of the 6/9/85 overspeed has the potential of damaging components, piping, and/or rupports. Also, overspeed of the AFPTs could j

result in equipment damage or personal injury if the overspeed i

protection were to fail.

i l 3) The uncertainty of event recreation. Analysis by Toledo Edison personnel and consultant engineers has concluded i that the dynamic situation that resulted in AFFT overspeed  ;

is sensitive to many different variables. Exact duplication ,

of the 6/9/85 event will not be possible. Therefore, the '

probability that AFPT overspeed could be achieved during l 3 Mode 3 testing is not certain.

B. Event Description On Sunday, June 9,1985, normal feedwater flow to the steam generators was interrupted. The reactor was automatically

- shutdown and reactor heat was removed via steaming through the 1 sain steam safety valves and the atmospheric vent valves. The water level in the steam generators was decreasing and at j 1:41:03 a Steam and Feedwater Rupture Control System (SFRCS) <

! full trip was initiated on Channel 1 due,to a low water level in Steam Generator #1 (SG #1). This STRCS actuation attempted'to l

initiate auxiliary feedwater flow by opening the' steam supply ,

valve, MS 106, from SG #1 to auxiliary feedwater pump turbine .

(AFPT) #1 (See Figure 1). Five seconds after the initial STRCS  !

actuation (1:41:08) the reactor operator inadvertently initiated

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

' an SFRCS low pressure trip on both channels and both steam generators. This low pressure trip of SFRCS is intended to i respond to a steam line break or other equipment failure resulting in depressurizing a steam generator. The manual low pressure SFRCS trip initiated the following, as designed:

1. Sent a close signal to MS 106 (which had an open signal at the time) and MS 107 (which was closed at the time).

! 2. Sent a close signal to AF 608 and AF 599, containment

isolation valves on auxiliary feedwater path to steam

! generator #1 and #2, respectively.

l i 3. Sent an open signal. to MS 106A (steam supply for AFPT #1 i

from steam generator #2) and MS 107A (steam supply for AFPT l #2 from steam generator #1) in an attempt to operate both ,

i AFPTs on opposite SGs.

i l 4. Sent an open signal to auxiliary feed pump discharge valves

! AF 3869 and AF 3871.

l S. Sent a close signal to auxiliary feed pump discharge valves 4

AF 3E70 and AF 3872.

Each AFPT tripped on overspeed (4500 RPM) approximately 25 seconds

after initial roll.

i II. BASIC OPERATION OF OPERATION f The Auxiliary Feedwater System (AWS) is required to provide feedwater j flow to the steam generators (SGs) upon loss of the main feedwater j pumps or upon loss of all normal or reserve electric power to remove j decay heat and/or promote natural circulation.

1 The AFW system consists of two separate pumps and turbine drivers.

The turbines receive main steam from either steam generator, depending on the emergency actuation signal supplied by STRCS. The normal steam supply paths are; SG #1 supplying steam to AFPT #1 through

.: 106 and SG #2 supplying steam to AFPT #2 through MS 107. In the

case of a low pressure condition in a steam generator, the system is realigned to isolate the bad generator and the good generator will

supply steam to either AFPT through the cross connect valves (MS 106A

! and MS 107A). .

i  !

l The AFPTs at Davis-Besse are Terry Steam Turbines. The #1 AFPT was

! equipped with a Woodward model PG-PL governor and the #2 ATPT was equipped with a Woodward model PGG governor.

! III.

SUMMARY

OF TROUBLESHOOTING AND INVESTIGATIVE ACTIONS 3

A. Channe Analysis j' In order to determine the reasons for the 6/9/85 overspeed trips

a change analysis was performed to compare the operation of the

\

4

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

i

! l AFPTs on 6/9/85 to past quick starts of the system. The change i analysis reviewed five previous plant trips (3/2/84, 1/15/85,

3/21/85, 4/12/853 6/2/85) as well'as several surveillance tests (STs) to attempt to identify a distinguishing factor between j 6/9/85 and other system operations.

The one difference that was found was that the 6/9/85 plant trip i was the first. time that the AFFTs were run solely on the cross

! connect steam supply lines. Also, 6/9/85 was the only actual overspeed trip of the AFPTs at Davis-Besse.

j During the change analysis phase, maintenance and modification

records were checked to determine whether there was any work j performed prior to 6/9/85 that could have caused an overspeed to

! occur. This review did not reveal any evidence that could l support the overspeed trips.-

t B. Data Search

? 1. Investination of Other Utility Experience with Overspeed Trips j

j Problems with overspeed trips of Terry Turbines is not i unique to Davis-Besse. Nuclear Power Experience (NPE)

! includes descriptions of turbine overspeed incidents at i three plants which have been attributed to condensation in

the steam supply lines to the turbines. The major features 1 of these incidents as documented by NPE are discussed in

this section.

i

{ In addition to the incidents which were directly attributed to condensation, there are other instances of actual l turbine overspeed (as contrasted to spurious overspeed

! trips). These are generally attributed to governor valve  !

{ malfunctions or improper adjustment. In most cases, these 1 l incidents appear to have been the result of specific ,

malfunctions of the governor such as clogged or dirty oil, miscalibration, corrosion, or faulty electronics. It cannot be established whether condensate or condensation during startup may have also had some effect on these other l incidents.

I Plant #1 In June of 1977, the turbine drive auxiliary feed pump (EPF-2) tripped on overspeed upon initial start in a surveillance test. A'new governor was installed; however, l

the problem persisted. The cause of the problem was then attribated to condensation in the steam line. Modifications were made to the drain system to increase its capacity.

Subsequent multiple starts from cold conditions showed that '

the problem had apparently been corrected. However, on July 17, 1977, the main feed pumps tripped and upon auto  ;

start of the auxiliary feed pump EFP-2, that pump tripped

_ ,. - - _ _ _ _ _ _ _ .__.______.___,___..m.___.__

i

! on overspeed. The turbine was reset ud the immediate restart was successful. Subsequent first start overspeed trips were experienced on July 22, 23, 25, and August 1.

The cause was determined to be condensation buildup in the steam inlet lines. In addition to the previous modifications of the drain system, a bypass valve was installed to prevent condensate buildup. Satisfactory operation of the

turbine was demonstrated.

Plant #2 i

In July of 1979, the turbine for auxiliary feed pump 2P7A tripped when it was started automatically by an Engineered

Safeguards Features Actuation System (ESFAS) signal. It is stated that the overspeed trip was due to water in the steam supply line. Water was blown from the line and the 3

turbine was successfully started three times. In September of 1979, the turbine for auxiliary feed pump 2P7A was

< manually started and the turbine tripped on overspeed. The j turbine was then restarted three times to verify its l

operability. The apparent cause was stated as water in the steam supply header. Additional drains were installed.

These additional drains were evidently not completely successful since in April, 1980, the turbine for feed pump 2P7A again tripped when it was started in response to an l

. ESFAS actuation. The turbine was successfully restarted on j three successive starts after the overspeed.

However, on October 11, 1981, steam driven emergency 1 feedwater pump 2P7A tripped on overspeed following a manual l start during a unit trip recovery. This overspeed trip was j attributed to condensation in the steam lines which resulted from insulation not being reinstalled and a steam trap remaining isolated after a maintenance activity. The insulation was replaced and the trap was placed back in service. Also, the trap bypasses were opened to ensure that condensation would not occur in the lines. Pump 2P7A was tested, found operable, and returned to service.

Plant #3 i

On August 25, 1982, it was found' that the drive turbine for auxiliary feed pump 2P-140 was tripping on overspeed during starting. It is stated that the tripping was caused by condensate in the steam supply header and throttle valve.

Some drains were modified to improve their effectiveness.

. Although the investigations by other plants into similar l problems with overspeed of Terry Turbines upon startup have not been definitive as to the mechanism by which water causes the turbine to overspeed, they appear to strongly indicate that condensate in the steam lines is detrimental l to turbine operation. Their approaches to corrective action have been to reducc the amount of ' condensate and in l

s some cases to keep the lines hot by bypasses or repeated starting of the pumps.

) ,

2. Vendor Experience Discussion between MPR Associates and Terry Steam Turbine (the turbine supplier) have revealed that water induction to the turbine can indeed cause an overspeed condition.

The relationship to overspeed is based on Terry Turbine's knowledge of turbine design as well as some testing performed in 1969. The testing involved injecting quantities of water into the turbine to verify the turbine's ability to withstand the water. One of the observations during testing was that turbine speed increased after the water had cleared the governor valve. Although no testing data is available, the information below describes the testing based on discussions with Terry Turbine personnel involved with the test.

The initial set of tests involved operating the turbine on steam, injecting cold water into the pipe upstream of the i

turbine, and observing the turbine response. The amount of water was varied between 50 and 600 gallons. During the transient, turbine speed and governor valve position were measured. The turbine behavior is described below:

As the slug came through, the turbine speed began to fall. The turbine control, or governor, began to develop an error signal due to the speed decrease and attempted to further open the governor valve. However, apparently because of the effect of water flowing through the valve, the. valve was not able to respond to the control signal. Terry refers to this as "waterlock" of the valve. Speed continued to fall until the slug was cleared, at which point steam was applied to the turbine and the valve could respond.

Since an error signal was present, the valve opened causing a rapid acceleration of the turbine. Terry said that the turbine speed would increase quickly and overshoot before settling back to the demanded speed.

The overspeed trip in these tests was set higher than normal, so they observed no overspeed tripping of the turbine. But they did see large overshoots in speed as the water slug cleared.

The final set of tests involved quick starting the turbine with 600 gallons of water initially in the inlet lines. A valve was opened to admit high pressure steam behind the

. water slug to force the slug through the turbine. It is not known whether the governor valve started fully open or was opened simultaneously with the steam supply valve.

N Turbine response is described below:

l

l 1

ex The turbine accelerated as the water was forced I through the machine. The governor gained control of the acceleration. When the water was expelled and steam was applied to the turbine, there was an increase in turbine creed but the governor was able to maintain control.

The Terry Turbine Company representative has said that the i governor valve used in the testing was not the same design used at Davis-Besse. The valve in the tests is thought to have been more susceptible to being affected by hydraulic '

forces due to water flow. However, the Davis-Besse governor valve cannot be guaranteed not to lock up due to water flow because it was designed to run with steam. i Terry Turbine has also stated that water could also cause a turbine overspeed by flashing through the turbine nozzles and thereby, due to the increased mass flow, impart more energy to the turbine wheel and potentially cause an overspeed.

3. Toledo Edison Experience To further support the root cause of water causing the overspeed trips, a review of past AFPT quick starts was e performed. This review involved analyzing the plots of turbine speed versus time during the starting transient.

( The review clearly showed a difference in stability of the turbines (especially AFP #1) when run on cold supply lines versus hot supply lines. When the turbines are started with cold steam supply lines, the turbine speed is very erratic during acceleration to rated speed. However, when the turbines are started with hot steam supply lines, the speed graphs are very smooth with a constant acceleration

to rated speed. (See Figures 2 and 3 for typical AFPT speed graphs of cold versus hot lines.)

The erratic turbine speed is attributed to the formation of water slugs when the steam is introduced into the cold supply lines.

C. Field Actions Two " hands on" investigations were performed to eliminate other possible causes. The first was performed on 6/9/85 at approximately 1:00 p.m. (about 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> after the unit was stable). The AFPTs were " quick-started" on the normal steam supply paths via MS 106

and MS 107. The pump discharge path during this testing was through the min-recire line only, just as it was during the overspeed trips of 6/9/85. The piping was still hot at this time and both AFPTs started without any erratic control. The governors functioned properly and accelerated both units to

(')

gJ rated speed within 40 seconds as designed. This testing indicates

i that it wasn't a governor problem that caused the overspeed trips earlier on 6/9/85.

The second investigation was a visual inspection of both governors by a Woodward Governor Representative to verify that the governors were in proper order. The inspections were performed under Maintenance Work Orders (NW0s) 1-85-2131-00 and 1-85-2132-00.

The visual inspection identified that both governors contained no deficiencies that would impair proper operation.

D. Analysis To better define the possibility of condensate in the steam supply causing the overspeed trips, analyses were performed to '

i determine the amount of water available due to steam condensing in the cold steam supply piping and to simulate the transient formation and transport of condensation in these lines.

1. Amount of Water Available  !

Initial calculations were performed by MPR Associates Inc.

to determine the mass of water condensed as a result of heating the AFPT steam piping from 70*F to 535'F. These ,

calculations were intended to give " ball park" numbers and therefore the assumptions were made that the entire pipe heats up from room temperature (70*F) to the nominal SG

, temperature (535'F) and that all the heat given up by the i

condensing steam goes to heating the pipe. The results are

! shown in Table 1, values are younds of water generated.

The actual values will be less due to the real parameters ,

! involved, (i.e., the effect of piping insulation, the

effect of flow on heat transfer, the time required to heat j the pipe, etc.).

Steam From AFPT #1 AFPT #2  !

3 MS 106 790 ---

MS 106A 1490 ---

, MS 107 ---

260 MS 107A ---

900 TABLE 1 These results show that large quantities of water can be generated in the steam supply lines. Also the cross connect lines are capable of generating more water than the normal supply paths. This is due to the increased length of cold horizontal piping when the system is run on the cross connect lines.

When comparing the aaount of water generated from MS 106 to AFPT #1 to the amount of water generated from MS 107A to l AFPT #2 it is not evident why AFPT #1 has never tripped on overspeed when run on MS 106. However, the isometric

___...m_ - - - , . , , . . . ,. ,g_ ,c-,-~,,9,,.ym._.,,.,,_._w._,,_m._ . , .,,w...,, won.,,mm.,,,,w_,w,,,w.,. ..

i

'~

10- l i

drawing of these two routings shows that the MS 107A line

is comprised of long horizontal runs and a vertical rise

which tends to more easily cause a slug formation when compared to the MS 106 piping which is comprised of a series of shorter horizontal runs and vertical drops. This tendency against slug formation is postulated to be the reason why AFPT #1 had never oversped in the past. It also points to the complexity of the phenomenon.

2. Transient Flow Analysis A transient flow analysis was performed by MPR Associates Inc. to simulate the effects of the 6/9/85 transient. The i computer model used solves the mass, momentum, and energy conservation for the piping system including the effects of the heat transfer to the pipe wall.

The results of the transient flow analysis iadicate that a significant amount of water is formed in the steam supply piping to the AFPTs as a result of the sequence of events which occurred on 6/9/85. Within 30 seconds after the initial opening of valve MS 106, the analysis indicates that the pipe immediately upstream of AFPT #1 is completely filled with water. Also, the pipe directly upstream of AFPT #2 is calculated to be about half full of water at the s same time. This is shown in Figures 4 and 5.

This transient analysis substantiates the fact that large

amounts of water can be generated when steam is initiated

into the long cold piping. In addition, it shows that the condensate can be expected to reach the turbines in approximately the time frame required to support the observed overspeed trips of the AFPTs (25-30 seconds).

IV. RESULTS/ CONCLUSIONS OF FINDINGS l A. Direct and Root Cause The direct and root cause is judged to be Hypothesis A of the original troubleshooting report (Reference #4).

l Hypothesis A:

Water sluss in steam piping to the turbine due to residual j' condensation or rapid condensation of steam while heatins

! lona, cold steam supply path to AFPTs caused the overspeed trips.

. B. Disproved Hypotheses As part of the original troubleshooting report, four (4) other possible causes for the AFPT overspeeds were developed. Of those four, two were eliminated in the original report based on s

I

- -. _... ....-4, ._y., . . . - - -__,,y. -- .,,,,,.,,,,,r,-. -,%-, .g_

[ , m , _ [" , _ _ h _ . _ - _ _#, $,,[_.,". , "] ,

I ps data collected during the 6/9/85 plant trip and previous AFPT

(\ initiations. Those two were:

Hypothesis C:

Sudden decrease in pump load due to a sudden flow reduction when pump discharge flow is abruptly stopped at the closed valves AF 599 and AF 608.

Hypothesis E:

Loss of pump suction source resulting in no pump load.

A third hypothesis was disproved by the 6/9/85 testing and the governor inspections performed by a Woodward representative. It was Hypothesis D of the original troubleshooting report.

Hypothesis D:

Governor problems had caused the overspeed trips.

The final hypothesis was Hypothesis B.

Hypothesis B:

,s AFPT #1 rolling on steam from MS 106 prior to receiving

( stean flow from crossover (Double Start).

L This is highly unlikely to have occurred based on an examination of the sequence-of-events log which suggests that steam flow would not have been interrupted. This judgement is based on the timing of the cycling of valves MS 106 and MS 106A. This hypothesis also does not explain the overspeed trip of AFPT #2 because it was run on steam from MS 107A only.

V. TECHNICAL JUSTIFICATION OF FINDINGS The basis for the direct and root cause determination is:

1. Other utilities have experienced overspeed trips of similar turbines due to water in the steam supply.

2. Toledo Edison experience shows that when the steam supply lines are hot (i.e., small amounts of water generated), the AFPTs accelerate to rated speed without any erratic behavior.

3. Terry Turbine has stated that water in the steam supply can lead to an overspeed condition.

4. Analysis has shown that the existing piping configuration will generate significant amounts of water due to condensation and that this water will reach the turbine during the initial

('~'N) starting transient.

i l

l l

Presently the method by which water causes an overspeed is not l precisely known. It could be any combination of the following:

J

1. Governor valve becomes " water-locked" and is open too far when water clears, therefore allowing too much steam to flow into the turbine which leads to an overspeed.

2. Water entering the turbine flashes through the turbine nozzles thereby releasing more energy than required and the turbine accelerates to the overspeed setpoint.

3. Water entering the turbine slows the turbine down. The gover-nor tries to maintain turbine speed by opening the governor valve further. When the water clears, the valve is open too far to be able to control turbine speed below the overspeed setpoint.

Even though the method is uncertain, it is apparent that the water is the cause of the overspeed trips of the AFPTs that occurred at Davis-Besse on 6/9/85.

TI . SPECIFIC CORRECTIVE ACTIONS 12 A. Reacired Corrective Action To eliminate the water which is judged to have caused the

) overspeed trips, Toledo Edison has decided to reconfigure the l j steam supply system to the AFPTs. The new configuration would

-l provide the ability to sustain the steam piping in a hot and

, pressurized condition whenever the AFPTs are required to be operable. By keeping the piping hot, the amount of water delivered to the turbines during a start would be reduced to a negligible quantity. The system change is highlighted below:

o Valves MS 729 and MS 730 (turbine inlet isolation valves located approximately ten (10) feet from the turbines) will be replaced with pneumatically operated control valves.

i o Valve opening time will be adjustable to allow control of steam flow during initial stages of turbine starting.

This flexibility will alleviate the possibility of over-speeding the turbine when opening a supply valve near the turbine as has been experienced at several other power plants.

o Valves MS 106,106A,107, and 107A will be normally open to pressurize the supply lines up to the new control valves. These valves will retain their close signals to act as containment and steam generator (SG) isolation valves.

l o Steam trap capacity of the supply' lines will be reviewed for adequacy. New traps will be added as required.

o Steam supply lines to the AFPTs will be analyzed as high l

fe ~g energy lines. Support, whip restraint, impingment barri-er, and environmental qualification modifications will be performed as required.

o The new control valves will be opened on an SFRCS actua-tion signal.

o In addition to the piping changes, the existing Woodward model PG-PL governor on AFPT #1 will be changed to a Woodward model PGG governor to increase governor reliabil-ity. This change-out was originally planned for the 1986 refueling outage.

These modifications or other alternatives to keep the piping hot and pressurized will be performed prior to restart of the Davis-Besse unit. After the modifications are performed, Toledo Edison will perform a series of tests to confirm that the new system configuration will operate as designed.

B. Planned Additional Actions The new auxiliary feed pump steam supply system modifications will be tested after implementation is complete. The testing will consist of the following:

p t 1. Several quick starts of both AFPTs to verify that the

, \ units accelerate to rated speed without any erratic behavior. This testing will also verify that the turbines reach rated speed in less than 40_ seconds as required by Davis-Besse Technical Specifications. During these quick starts the following data will be collected:

o Pipe movement will be measured to verify that piping and hanger loads are within design limits.

I o Governor valve position will be monitored to verify proper governor operation.

1 o Turbine speed versus time will be recorded to verify a smooth acceleration ramp.

2. AFPT stability will be checked. This will be done by momentarily increasing turbine speed by using the over-speed test device. Turbine speed will be plotted to verify that the governor resumes control after the insta-bility input is removed.

3. All steam traps on the steam supply piping will be checked for operability and adequate removal capacity during the steady state condition (i.e., lines pressurized with l ~'N turbine stopped). This testing will ensure that condensa-tion will not accumulate in the piping during the long g ) periods when the AFPTs are' idle.

. ~ . . _ - . _

fs s 4. Testing will be performed to verify the hypothesis of Action Plan IC. (Hypothesis - Trip throttle valves being only partially open limited steam flow and thus limited RPM. This was perceived as not being able to control AFPT speed from the control room.)

All of these modifications and testing will combine to assure Toledo Edison that the Auxiliary Feed Water System is function-al to perform its intended safety function at any time as required.

VII. GENERIC IMPLICATIONS l2 The design of the Davis-Besse Plant utilizes only two quick start steam driven turbines (AFPT #1 and #2) to supply motive power to equipment. Therefore, there are no generic implications beyond the two AFPTs.

VIII. REFERENCES l2

1. Telephone conversation documentation between MPR and Terry Steam Turbine Company, (7/2/85).

2. Steam condensation calculations by MPR, (4/5/85).

3. Draft transient flow analysis by MPR.

4. Action Plan IA & IB Investigation & Troubleshooting Report,

, Rev.1 (6/25/85).

4 1A & IE Report

i. _ . - - .- __. -. . . . . - . . . _ - . , . . . , . , _ . . . - . _ . . . . _

I

/^T U

f ) ( 3 Steam Steam Generator u; Generator N o.1 7, No.2 MSrO7A MSIO6A IMSIC6 I MSl07 AF599

/

, A N J s g AF3870 ,, AF3872

~

s' L

~

t; .e AF3871 r,

AF3869 Aux. Steen } I4

" 273 >4

>4 MS733 MS728 MS730 MS729 l

x%

l S /- $

C- - - - - - -----

-3 AFP AFP No.1 No.1 No.2 No.2 N

l SCHEMATIC REPRESENTATION OF l AUXILIARY ' FEED PUMP TURBINE STEAM PlPING SYSTEM Figure 1 l

- - _ .< - ,__ ,-.- - -y-- , , - - - , . - - -

' , , :ji i

,~i 5

_ 4 A b __ _ __

- - 0

- _ 4 H

5 3

P T G

- - - - A R N _

. ~ _ _ __ j_ -

R A I .

_- _ _ _ P __

.. C T -

_ ) - .

._ S

. . _ _- . _ _ _- _ e D P I

E E N

E N D _

P L I

-__ . _ B O S

/ L R

U C

- _ I 0 3

- _ _ .. _- . _ _ _ 3 _

_ A T H C T

-nT '

I F I _

N P 0 W .

- . _. _ . _ / Y .

T

_. _ .. _- _ _ .t

_r 5

.M - _ -

2

. _. M ) .

. _-m___

R. _-_ C. 2 E

. __ _ - S e V. '

- ( r r' .

u

. . 1~ - 0 E 2

i r~

.1 MF.

I l

-_ J T J.

W

-Y ..

_ _ _ =

_ - ,m" s _

_ O- 1I' i

. _ . - ]I lt1

\IL

_ __ _ - p - 3 -

_ ( 1h III i

- Igl)

-_ . g \(

. - - 0\

\ 1t'

___ -_ o P '5

. "\ t M8 . = l I

U/ _ y\

P P. - \

\.,'

i(

_. D/ (= .

1 \ gr '

E _. )L \

E  : _

(I IIIgIf1 F 5

.ET _ _

(  ;,I I I3 lf X

U M [

< f ' g A _ _ I '

o 5 0 5 0 5 0 5 0 5 5 4 4 3 3 2 2 l 1 0 O O8

• 2QO .OWWO(.O iij;ji' i il'lll, >it. .; 4i,!,Ii' 1

,, .1 ' > 1 l

,C' 5

_ 4

_ ~ ~

~~ _

. pd_. _

\-

=

0

_ -- -- 4

- -- - \ , -

, i F -

O i

- . - ~ H -

H T

- - P I - _ .-

\ ,

A W -

=_

. _ Y ,, , - = - R G T G

N - - - _ -

. -. R D A I

P - -

/

m p -

E E

T S

I P - - ._

P

_ =

- - S E T - - _ - 0 N O 6'_

-- . 3

-- -_ /p <

- L A

I B

H - _

s

- C R -

. m I U _ ,

i a

y -

P Y

T

- sy T 5

- s - -

2 s

y, - -

)

- C.

E 3

- =

- N S

( e 0 E r

u

-- _ J

_ i 2

M i g

r F

-_ ---- _ p I T

_ /,

. /

.- . . _ ,g - -_

W.-

. (it " -

5 T .--.

/ - 1

_ Y~

(

0 _----

r c

- . .- . - - f

. iT .

Jx g

- T P 0 MM U / _.

0l f IJ t/

I /

1 P 9

.- D /

'\n '

E.

L 6 -

.(

> \

1

. I E -

. (JlII,II 5

.T X A -. .

f 9 I I UD o A -- ,

- 0

> 5 0 5 0 5 0 5 5 4 4 3 3

. o.

2 2 1 i OOO ,2k OWW1o t j,'1 j; ' 4il' 4I ' l1 l:]>l;!1; > i- ,)l1 i j:l ffi*

,l

).

0 0

0 5

0 0

5 5 0 4 0 )

t r

1

' t 0

/ ' e 0 C

- fE 0 S l 0 B R I . D 01 5 e U

n 0

4  : D 5 D5LJ "

S m E E 1 EO S a L P 8

/

9 0RL 0PF

' sR e t 0

I F

P I

. /

6 P (S 0 R

C C [ C T

._ N 0EEE 6T 5

O 3A T T HA o 3 5 P

/L L s 1 S

R DU %U s U t

E e 0 P T E E C_ E_ C_ S 6 E PR HA .

M 0 N A SC LC ) E R

3 V S

A P

l l aa TT PP H

l F AHA

] l m a T T PP r l' 7W 0 O 0 C.

D E

N I

R D q, E r N u y.

_ 5 S(

B Q P 2 I F

A y R f E

F 1 r U 0

O L

m 0 i

T T

H E 0 l

. D 2 0 O

H R

E

' N 0 0

~

T ' 0 T U

5 N P

f E M

~

1 I

O S C N 0 A R

0

' . 5 T n 0 8 o 1 9 9 D t i / 1 I - E c 0 N '

_ O a 0 2 U V rP )

n 0 - J a . 8 e e 5 0 m t u S  :

l o f E T

V

(

o 0 A 0 -

D 0

0 0 0 0 0 0 0 0 0 0 8 6 4 2 8 6 2 0 2 1 1 1 1 0 0 0 0

,  ;  ; l 4l  ! f;i{jj4; 1 ; j11 $

'3i 0 _

0 0

)

5 t

) e t l e n 0 l I 0 r

I m 5 0 a 4 0 m e 1 0 a t t 1 e S 0

/ t (

0 C S 0 B

!E > ( E D l 0 P U1 U 0 4  : D 5 /

D5u W S S

E E 5 1 EO O L P 8 0RL L E I P

/

9 0PF F R 0 F I

/ f - P 0 R 6 C C [ [

. T g 5 N

0E 6T EE T T x 3 5 O

S R

E 3R DU

/L RR LL UU

^

' 0 0

P 1

f U

P T E C.

E C. C. .

PR HA N f\\

E - 0 1

SC CC E D 3 )

S V W

J 22 22 E O I mm m = E N R 5 TT TT P 0 O D PP PP S " 0 C E e E

l l i l ' . r AAHA ' 5 S( N u T 2 I g P ' ) B i

" m

{/V F R F A ' a E e 1 t U F t 0 I T O S 0 T H

L E

0 T D D 2 O O I B M O R

E T

U V

[ ,

0 0

5 T

N P 1 E I

- M S O

C N 0 A 0 R

. 5 T 0 8 1 9 S 1

- E 0 N 0 2 U 0 - J

. 8 5 0 E

T 0 A 0 D 0

0 0 0 0 0 0 0 0 0 0 0 0 8 6 4 2 0 8 6 4 2 0 2 1 1 1 1 1 0 O 0 0 0

FINDINGS, CORRECTIVE ACTIONS AND GENERIC IMPLICATIONS REPORT

\

TITII: Toledo Edison - Source Range Nuclear Instrumentation NI-1, Channel 2 REPORT BY: M. Borysiak (TED) PLAN No.: 15A P. Alleman (B&W)

PAGE 1 0F 31 i

l CHAIRMAN REV DATE REASON FOR REVISION BY TASK FORCE I

M. Borysiak

( 0 8/23/85 Initial Issue P. Alleman L. Grime /)

M. Borysiak 1 8/30/85 Title Change P. Alleman ( _

M. Borysiak s Added Data b 2 9/16/85 P. Alleman C V

~

1 T ~ . . .

TABLE OF CONTENTS P_a.gg I. Issue / Concern 3 II. Basic Principles of Source Range Nuclear i Instrumentation NI-1 Channel 2 4 l i

III. Summary of Troubleshooting and Investigation 5 A. Field Actions Performed 5 B. Significance of Findings 19 IV. Results/ Conclusion of Findings 19 A. Direct Causes 19 i

B. Root Causes 20 4

V. Technical Justification of Finding 21 VI. Specific Corrective Action 22 A. 22

( Required Corrective Action

. B. Additional Planned Actions 23 VII. Generic Implication 26 VIII. Generic Corrective Actions 27 l

f Figure 1 Source Range Instrument String 28

Figure 2 Triax Connector 29 1

Attachment 1 l2 i

l l

1 w- ~.--p.9c.w ,-.-qt **--S'r* - ' " - - * ~ ^ * - -*T %--*4**w'+-'eW'TM+='-'9~-mPur'-'-,-w-e-, -

p I. ISSUE / CONCERN Both NI Source Range Channels were inoperable during ;he Reactor Trip on June 9, 1985. This had no significant impact on plant operability but was an added distraction to the operators.

On June 4, 1985, source range out of core detector NI-1 was declared inoperable due to it reading an elevated count rate ~ 1.5 decades above NI-2 with the detector high voltage cut off. A reading of 10-1 is normal for this condition. Spiking was also associated with the elevated count rate. NI-1 was still inoperable during the trip on June 9, 1985. After the trip when high voltage was applied to NI-2, i it failed to respond and remained inoperable for approximately I hour. Technical specifications require that at least one source range channel be operable or shutdown margin be verified by the operator.

'l Even though there was no significant impact during or following the transient on June 9, the inoperability of both NI-1 and NI-2.did create an undesirable added distraction to the operators in having to verify the shutdown margin requirements per Action Statement 6 of T.S. 3.3.1.1. From this standpoint, it is highly desirable to locate and correct the problem with NI-1.

II. BASIC PRINCIPI.E OF OPERATION .

\

The Source Range instrumen'.ation consists of two redundant count rate instrument strings designated NI-1 and NI-2 which measure neutron flux.

The subject of this report is NI-1, Source Range Instrument String l2 as shown in Figure 1. NI-1 is located in Nuclear Instrumentation /

Reactor Protection System Channel 2.

The Source Range Signals originate in high sensitivity BF3 proportional counters located in thimble wells in the reactor vessel cavity.

The proportional counters are connected to preamplifiers located in the reactor building. The preamplifiers are used to shape and amplify the low-level pulses from the proportional counter. The signal output of the preamplifier is sent to the count rate amplifi-er module (CRAM) located in the Nuclear Instrumentation / Reactor Protection System Cabinets.

The CHAli which contains a discriminator and log count rate amplifier further shapes and amplifies the signal pulses. The discriminator i

discards background and gamma pulses. A log count rate amplifier l-converts the neutron pulses to a signal suitable for displaying the log count rate level both locally and remotely to the operator. The display range is from 10-1 to 10s counts per second.

s The Rate of Change module has differentiator amplifiers which

-/

)

calculate the rate of change of the log count rate for display locally and remotely to the operator. An interlock is provided for control rod withdraw hold and alarm on high startup rate.

III.

SUMMARY

OF TROUBLESHOOTING AND INVESTIGATION Nuclear Education and Training Services, Inc. (NETS) from Ohio State University was requested to review action plans and Troubleshooting 2 as an independent input. The NETS input is included in this report as Attachment 1.

j A. Field Actions Performed Maintenance Work Orders 85-2092-03 and 01, 1-85-2192-00 and 01 Action Plan Step 1A and IC - The object of this step is to measure the CRAM pulse input, CRAM pulse amplifier output and l CRAM output. These were observed at normal and elevated count

, ra tes . The source used for these tests was reactor residuri i

neutron flux. It was determined by performing cable and connection inspections that'the high count rate preblem was upstream from the CRAM. After this, only the CRAM pulse input was monitored. Oscilloscope picturea were taken indicating both normal and elevated count rates. See action plan Step ID for the results evaluation.

l

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

Action Plan Step 1B - The object of this step was to repeat readings of 1A above with the reactor at power (Mode 1) and the high voltage off. This step cannot be completed when the reactor is shutdown.

Action Plan Step ID - By a slight movement of the Amphenol connector between the MI cable and Triax cable at the detector well, the elevated count rate could be made to come and go away. This was repeated three times and establishes the i

assembly of the detector, MI Cable or Triax cable interface as a possible cause of the elevated count rate.

.~

The existing detector with attached Triax cable and Amphenol connector has been removed and sent to the vendor for determi-natica of root cause. The vendors analysis indicates that the detector has no problems and that a cause of elevated count rate appears to be in the Amphenol connector. Westinghouse

, Test and Evaluation Report number NC TR 85-19 has been pro- 2 l

l vided to Toledo Edison Co.

A new detector with integral MI cable and Amphenol connector has been installed in NI-1. An elevated count rate with spiking conditions has been observed since replacement of the detector and connector.

9

,. ,- ., -, - - - , , . . , - - . - - ,.,_-,..,--n, . ~ , -,-,-,Sw, ,_n,.,_ n,,

l l

1

~s Action Plan Step 2A - Time Domain Reflectometer (TDR) traces

3. were taken using a Tektronix Model 1502 Time Domain Reflectometer. The time domain reflectometer measures imped- ,

ance vs. cable length by sending.a high frequency signal down the cable and monitoring reflections coming back.

The cables checked are as follows:

1. RPS Cabinets to the Preamp

, a. Signal Cable 1

b. High Voltage Cable

c. Low Voltage Cable i

2. Preamp to Detector

a. Signal /High Voltage Cable The results of the TDR tests are:

1. The low voltage cable from the RPS cabinets to the preamp shows very small impedance changes from the penetration to the preamp. The small impedance changes are judged not to be a cause of the elevated count rate.

9

I 8-

2. The high voltage cable from the RPS cabinets to the preamp shows'very small impedance changes from the penetration to the preamp. The small impedance changes are judged not to I

be a cause of the elevated count rate. [

t b

i Action Plan 2B - The concern is that NI-1 cable length is terminated in a %, \-or full wave length to be susceptible to

- resonating at a frequency of 45 to 60HZ. The elevated count rate normally fell in the 45 to 60 counts per second range and is the basis for selecting the resonating frequency range of 45 j to 60HZ. The calculation of resonant frequency of the cables  ;

1 I

) need not be performed due to determination of the root cause 2 f as explained by Significant Problem #8.  ;

1 Action Plan Step 2C - An inspection of conduit- and cables i

looking for improper assembly of connectors, NI cables located close 'to high power cables and other things that could contrib- l l ute to the elevated count rate problem has been completed except for two connectors per cable internal to the penetration as outlined by Section VIB Additional Planned Actions Item 8. 2 l

As part of the troubleshooting procedures some repairs were made with approval of the Region 3 NRC representatives. The problems listed below were detected during this inspection l

and other troubleshooting procedures:

i i

l l

..._.._.,____,_-..___._...____,._,__,_,_____,..,.....-__._,..__._m_.-

\

l 1 l l

Significant Problems

1. RPS Cabinet (Cabinet 1) (5755F) has a loose connection to 1

. the station ground bus. Tightening of this connection 2

! has been completed.

2. The center pin on the preamp detector cable connector was

pushed in approximately k" and was off center. This connector has been replaced using installation instruc-I tions written in the Maintenance Work Order. Figure 2 i

, shows a typical connector assembly.

2

3. Operation of Reactor Protection System (RPS), Safety

Features Actuation System (SFAS), and Steam Feedwater i

Rupture Control System (SFRCS) cabinet door switches 2

(which provide annunciator indication of open doors) frequently cause high level spikes visible by use of an oscilloscope at the input to the CRAM. No alarms have I'

resulted. Spiking is aggravated by the elevated count rate problem and is not a problem when the elevated count f rate is not present. The monitoring recorder temporarily j connected to the output of the CRAM does not respond to these spikes.

4. No grounding wire is connected to the outside box which contains the preamp.

l l -

4

• 5. RPS Channel 2, Cabinet 2 has a loose connection where the Cabinet I safety ground bus tiire is connected to the j safety ground bus by a single bolt. Tightening of this 2 i

connection has been completed.

i

6. Every Amphenol connector inspected was tarnished and many contained metal flakes from the connector assembly.

2 l

7. Between August 24 and August 26 NI-1 failed due to elevat-

) ed count rate. This was after the new detector and -

1 interfacing connector had been installed. Subsequent to this it was determiced that connectors installed on the j

preamp over top of paint was not providing an adequate

~

f ground connection. It was determined by test that re- 2 7 sistance values, sensitive to forces on the connector, from infinity to less than one ohn existed between 1) the i

Triax cable inner shield and the preamp inner box and

! 2) the Triax cable outer shield and the preamp outer box.

l l Both items 1 and 2 abcve should be low resistance (<1 ohm) connections. This condition was corrected by removing the

! paint from the preamp boxes at the locations where the connectors are att' ached.

i 9'

,_.._...___...,_---__.,-.._.._.__;_._,:.....:r.;;..rn.u,__,_.-._ _ . _ _ . . . , . . , _ . . . _ - . . . . ~ . . - -

d i

t fs 8. Operation of some motor operated SFAS isolation valves w_ / have been identified as causing some spiking at the output

'i of the Rite of Change Amplifier Module (ROCAM). Observation 4 of spiking vs. valve operation is continuing and is being i logged by the reactor operators. This observation will I continue until a month of power operation is accomplished.

Observations j 1. The power supply high voltage output connection has a i  ;

, crushed "O" ring. ,

i i

2. The preamp had no "0" rings on any of the cable connectors.

! l

3. The detector cable bushing on the preamp outer box has a [

j-5~.ry small clearance of ~ 1/32 inch from the detector cable connector. The concern is that these points may short causing a potential ground loop. These points are electrically insulated from each other. Additional 2 spacing was provided by installing a gasket on the mount-i ing for the detector cable bushing.

4. The TDR trace for the low voltage cable shows very small impedance changes from the penetration to the pream, The small impedance changes are judged not to be a cause of the elevated count rate problem.

_..w.we-.mn-e--,-w.,.-,,w.m.ar.u-,.,_.

- ,.,,.n.,,,n.,,.nn

-~s 5. The detector center line location versus midplane of reactor core for both NI-1 and NI-2 was investigated. By comparison of the exposed MI cable in the thimble connec-tion boxes it looked like one detector was not located properly on the core midplane. This investigation has been completed and both detectors 'were located properly.

6. The detector and high voltage cable connectors at the i

preamp appears to be nickel (shiny) plated versus silver ,

plated. Silver plated connectors are specified by the i vendor.

7. The TDR trace for the high voltage cable show very small

\

impedance changes from the penetration to the preamp. The l small impedance changes are judged not to be a cause of 1

the elevated count rate.

8. In some cases fiber (shipping) washers have been left on i

the bulkhead connectors installed at the Preamp and RPS I

Cabinets. Thds fiber washer prevents the connector from being tightened in place properly. The result would be that the conmactor body would turn during tightening of the connector. The fiber washers were inadvertently left in place during initial construction. All bulkhead 2 connectors have been checked and fiber washers removed where they existed.

, 13-s 9. The Blue Ribbon Connector on the source range high l

voltage power supply in the RPS cabinet is chipped 2 and cracked.

t 10. The Masonite spacers used to block the area around the detector signal cable in the detector thimble plug were approximately 5 inches long. They should have extended 2 1

i the full length of the detector thimble plug. When the new detector was installed the correct length spacers were i used.

I Action Plan Step 3 - This test is to remove modules in RPS Channel 2 to determine if noise is being introduced from >

l sources external to NI-1.

! x.

This test was to determine the source of a high frequency noise signal with an amplitude of approximately 100 millivolts and is I

a separate action from the elevated count rate problem. The

! input to the CRAN was the monitoring point in all cases. The source range test module and rate of change module were removed and reinserted individually with no change in the monitored noise level. The high voltage, low voltage cables and signal j cable were removed at the RPS cabinets individually with no

! change in the monitored noise level.

i-l The CRAM was removed and bench tested. No input noise was

[ present during this test and the module tested OKI.  ;

mm,-w-, y- -,.wv.,,,. , , ,- . , - . n --

,,p- re,e-p+es---pt,ye.-,v-e,-+w.r-r., ,..w----,.ww, ..m,..--,.g.

NOTE: The noise present on NI-1 source range signal even

( though within specifications was higher than that observed for NI-2. It was felt that'if the source of this noise could be determined in conjunction with other troubleshooting being performed that overall system operation could be improved. This noise is i

present with and without the elevated count rate present and does not contribute to the count rate

problem.

Action Plan Step 4 - The results of Step 3 testing did not indicate that removal of individual boards within modules would be necessary. l2

) Action Plan Step 5 - The object of this test was to perform I

extensive tests on control relays within the NI-l instrument string modules. None of the previous results indicated relays were a problem therefore this test need not be performed. l2 Action Plan Step 6 - The object of this test was to investigate all RPS Channel 2 power supplies for sources of noise. The DC buses within the RPS cabinets were tested for sources of the noise. The plus 15 volt bus in both RPS Channel one and two contained a ripple frequency of 7300 Hz with a 10 millivolt amplitude with the detector high voltage power supply turned on. This noise signal disappeared with the detector high voltage power supply turned off. The 10 millivolt amplitude of

+---------,..~,----w.w.--- --,,-m- ,,,-n..,_n- n.n,-,-. ,.~,..-..m.,--,, ,g,--, .-

l I this signal is within specification and is not significant to i

the elevated Count Rate Problem.

i

> i s

To isolate system power supplies as a possible cause of the 2 elevated count rate problem the source range high voltage and

the 15 VDC cables to- the preamp were disconnected one at a time at the back of the RPS cabinets. An external power supply was '

connected to provide power to the preamp. The elevated count

! rate was still present in each case when the external power i

supplies were providing power to the preamp. This test elimin-ates the system power supplies as the cause of the elevated count rate problem.

Action Plan Step 7 - This test objective was to determine if the CRAM was going into oscillation. With the signal cable l

disconnected at the preamp the count rate was < 10-2; however, i e when a resistive / capacitive network (to simulate the detector) was connected to the signal cable the count rate saturated at

> 108 counts. It was concluded that the resistive / capacitive I

f network was inadequate to complete.this test. A new detector

] was then used to complete this test. -See Action Plan Step 11.

~

1 Action Plan Step 8 - A spectrum analyzer was used to compare i

the CRAM input of both NI-1 and NI-2. A DC component and also a 60 Hz component was present. The frequency spectra obtained l

for NI-1 and NI-2 were identical.

16-

/"~' Action Plan Step 9 - The Object was to increase discriminator f

settings with elevated count rate to determine if the noise could be discriminated from the signal. The discriminator setting was increased in steps to a dial setting of 5. It was

. concluded that the signal to noise ratio was such that the noise could not be discriminated from the signal when the elevated count rate was present. The existing discriminator setting of 0.7 for NI-1 currently exceeds the existing setting of 0.56 for the NI-2 instrument string.

Action Plan Step 10 - The object was to transfer modules f

between NI-1 and NI-2 to determine if the associated problem in each channel follows. The CRAMS and the detector high voltage l2 power supply modules were interchanged between NI-1 and NI-2.

Noise level and count rate remained unchanged in each respec-tive channel when the modules were interchanged.

i i

i l The preamp in NI-1 was replaced by a new preamp with no ob-served changes in the associated problems. The preamp removed l from NI-1 was bench tested successfully in accordance with the l

vendor manual preinstallation acceptance test procedure. No l2 other interchanging of modules is planned.

I Action Plan Step 11 - The object is to connect a spare detector at the preamp location and evaluate NI-1 noise level. A new detector was connected at the NI-I preamp location with no elevated count rates being observed after monitoring for four i

-%4 - , ,---wer- , ,.-,- , ,y -

.--..,-,_,_.es.yw.,.r-.---e, erg , mva,,,.-,.mw,.. ,--w-,m,,,.m..w,-=,m-

days. The output of the CRAM was recorded. Completion of

- Action Plan Step 11 satisfies the requirements of Action Plan j Step 7 in that no oscillations or elevated count rates were 3

observed during the four day monitoring period.

l Action Plan Step 12 - The object is to remove NI-1 modules while observing CRAM input for elevated count rate and also perform bench test of NI-I modules. The source range test i module and rate of change module were individually removed with no change in count rate level observed at the input to the CRAM module. The CRAM module was removed and a bench test was performed successfully in accordance with vendors manual j preinstallation acceptance test procedure. No anomalies were noted during these tests.

1 l Action Plan Step 13 - The object of this test was to inter- ,

change the high voltage cable and signal cable between the RPS and the preamp to determine cable effects on the elevated count rate. This test was performed with elevated count rate present 2 and the elevated count rate remained after interchanging the signal and high voltage cable.

(

l l

Action Plan Step 14 - The object of this step is to:

1. Remove the NI-1 Source Range detector, interfacing connec- 2 tor and a short section of the triax cable to be returned

- ' 1 to the vendor for evaluation. ,

8  :

l l_._. _. _ _ _ . __ __ . _ - _ .... _ __ _ _ . . . _ . _ _ _ _ _ _ _ _ _ _

j i

2. Install a new detector in the thimble well.

i

3. Install a new connector on the triax cable at the thimble l well and at the preamp.

4. Place channel back in operation and continue monitoring for elevated count rate and spiking.

i i

This action item was completed and an elevated count rate problem occurred as described by Significant Problem Number 7 under Action Plan Step 2c.

4 l .

i Action Plan Step 15 - The object of this step is to inspect, i clean and/or replace connectors as required. Each connector 1

i cleaned or replaced is to be documented and monitored for l

4 elevated count rate and spiking.

I r

This step has been completed except as described by additional ,

+

planned Action Item 6.

2 Action Plan Step 16 - The object of this step is to correct loose Station ground connections in cabinets C5755E&F and monitor for elevated count rate and spiking.

i This action plan step was completed with no. change in the

. elevated count rate problem.

i l

- , . . , . . . , , - . . - - - - - , , , - - _ - ,c. m,--. ,,n.., , , - , - , - , , - , , - - . - ,,.e,,- .-mwg--,,y., ,w.--,v.r.w--rye, , - ,m y y m ,, p-, w e- w-

I B. Significance of Findings O.

b The significant findings are:

1. Movement of the Amphenol connector between the MI cable and Triax cable at the detector well would cause the elevated count rate to occur.

2. Improper installation of Triax cable connectors and the poor state of maintenance of connectors in the NI-1 instrument string contributed to the elevated count rate 2 problem.

i

3. High resistance connections existing between Triax cable

)

shields and preamp inner and outer boxes contri? yted to the elevated count rate problem. This condition was caused by failure to remove paint or use star washers properly by the manufacturer when installing connectors in the inner and outer preamp boxes.

IV. RESULTS/ CONCLUSIONS OF FINDINGS A. Direct Causes t

1. A direct cause of elevated count rate on NI-1 was deter- 2 mined by Westinghouse to be improper installation of the Triax cable jack which mates with the detector cable

s. -

'~' This Triax cable jack was X-rayed by the Exam connector.

Company which verified the Westinghouse conclusion.

2

2. Another direct cause of the elevated count rate on NI-I was determined to be high resistance connections existing between Triax cable shields and preamp inner and outer boxes. This condition was caused by the failure to remove paint or use star washers properly when connectors were installed by the manufacturer on the inner and outer preamp boxes.

3 .' Some spiking alarms are clearly related to the annunciator associated with RPS cabinet door switches when an elevated

) count rate is present.

Troubleshooting efforts are continuing to further identify the sources of NI-1 startup rate spiking.

B. Root Cause

1. Inadequate grounding of the connectors where attached to the preamp housings caused by failure to remove paint or 2 i

improper use of star washers by the manufacturer was a cause of elevated count rate.

t

%s d

--es - - -

, - - - -+,- , - - - - - - -..- L ,--ye.. ..w--,er

2. Improper installation of the Amphenol connector at the

(

5

% interface between the detector MI cable and the triax cable at the detector thimble well. This was concluded by 2 the vendor evaluation as a cause of elevated count rate. ,

i

3. Other contributing factors to the elevated count rate was degraded condition of other triax connectors.

1 V. TECHNICAL JUSTIFICATION OF FINDING Moving the MI cable at the detector well of the source range instru-i ment string identified that an elevated count rate problem could be s

induced in the detector MI cable or the MI cable to Triax cable interface connector.

The preamp and modules were interchanged between NI-1 and NI-2 and bench tested in accordance with vendor procedures and no abnormali-ties were found; however, by further testing it was determined that resistance valuer, sensitive to forces on the connector, from infinity to less than one ohn existed between 1) The Triax cable inner shield and the Preamp inner box and 2) The Triax cable 2 outer shield and the preamp outer box. It was determined that when the high resistance conditions were present an elevated count rate existed. Monitoring has been in progress for 15 days after removing the paint from the preamp at the connector locations with no elevat-ed count rate problems occurring.

1 J -

l

.. _ . _ . - - ,, - , , . ..-_____-.--__,_,_._,.,,__--...m_. -

The power supplies were evaluated for sources of the problem. No

' abnormalities were found.

l The assembly of the detector, MI cable, and MI cable to Triax connector has been evaluated by Westinghouse.

The Westinghouse conclusion was that:

1. The detector was in good operating condition.

2 >

2. The center pin of the triaxial jack that was mated to the detector cable connector was positioned incorrectly. This conclusion was verified by X-ray of the connector after return to the Toledo Edison Company.

VI. SPECIFIC CORRECTIVE ACTION A. Required Corrective Action l 1. The NI-l detector has been replaced. Inspection, clean-1 itg, and replacing (as necessary) shielded cable con-nectors from the detector to the RPS cabinets is in l

progress. All connectors for the NI-1 string from the I "As necessary" means replacing those connectors that cannot be cleaned

) properly, the tarnish removed or shows other signs of not being a l

quality connection.

1

l l

l I

detector to the RPS cabinets have been inspected / cleaned /

% replaced except two per cable which are inaccessible at 2 the penetration. The connectors at the penetration are operating satisfactory are not a cause of the elevated count rate problem but will be replaced as an additional  :

i planned action prior to startup after the next refueling outage to assure that all connectors in the NI-1 instru-ment string are installed to the latest procedures. This work is being done with approval of the Region 3 NRC representatives. The connectors are being replaced using installation instructions written in the Maintenance Work Orders. Selected personnel have been instructed in the proper installation procedures by the system vendor representative.

2. Recorder monitoring of the CRAM output is in progress for assurance that the elevated count rate problem has been corrected. Monitoring will continue until a month of 2 power operation is accomplished, i

B. Additional P,lanned Actions:

1. Revise or prepare new procedure for installing triaxial cable connectors. Current procedure is MP 1410.24. Due (2 to the generic corrective actions as outlined in Section VIII, this item will be completed prior to the start of the next refueling outage.

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

2. Provide cable connector installation training for station 2 personnel. This item to be completed prior to further cable connections. Work is planned to commence during next refueling outage and training must be completed prior to this.

3. Add grounding wire to the outside box which contains the 2 preamp. An engineering avaluation will be completed this fuel cycle to determine if this ground wire is required.

The ground wire will be installed during the next refuel-ing outage if required.

4. Review and revise (as required) procedure number MP 1504.04 2 for adequacy in determining if the seal plate rubber gasket, for the connector box at the top of the detector thimble, requires replacement. The existing procedure provides for applying RTV sealant to the seating surfaces ensuring a watertight seal. Gasket replacement concerns I

will be a'ddressed prior to next refueling outage.

i

5. Replace Blue Ribbon Connector for the source range high 2 voltage power supply. The chipped / cracked Blue Ribbon Connector would not effect system operation and could be replaced any time, however it will be replaced prior to this startup.

l l

l

l

. 1 l

~' 6. Replace Source Range NI-1 Amphenol penetrations with new 2 s_- ! Conax penetrations. This action is necessary in that the existing penetrations contain Triax connectors that 4

require penetration removal to inspect or replace the connectors. The new Conax penetrations will eliminate one connector per cable and the connectors will be accessible after penetration installation. The replacement of the penetration modules will be completed at the next refuel-

! ing outage.

7. Revise procedure number IC 2002.02 so that it adequately 2 addresses all areas of NI detector removal and installa-tion. No further detector removal / installation is planned.

This item to be completed prior to next refueling outage.

8. Revise procedure number IC 2002.04 to address current 2 requirements for source range discriminator and high voltage settings. This item will be completed as part of the annual procedure review or prior to next refueling outage, whichever comes first.

9. Revise procedure number MP 1410.24 to address current 2 requirements for installation and removal of Raychem heat shrinkable tubing. This item will be completed

, as part of the annual procedure review or prior to the next refueling outage, whichever comes first.

l

\s l

l l

~

l 4

f .

10. Continue observation of spiking vs. SFAS isolation valve operation and look for/ correlate with any other equipment i

actuations that cause spikes at the output of the ROCAM 2 l

l and perform tests to determine source of spiking. This

] observation will continue until a month of power operation is accomplished.

t i

i 11. Prepare an interim administrative procedure for operating with SFAS isolation valve spiking present and any other ,

j

equipment actuations that cause spiking pending (if i

j possible) location and elimination of the source of the 2 spiking. The SFAS isolation valves would not normally be d'

operated during plant startup or shutdown therefore this would not present a problem to operation during startup or shutdown.

I VII. GENERIC IMPLICATION f 1. Triax shielded cable connectors on the other NI Systems may l2 have conditions similar to those found on the source range instrumentation.

f 1

i i 2. Inadequate grounding of preamp connectors may effect other 2 l plants using the same model number preamp.

t

, . . - . - - ,-,..._--.-,,_m.,,,.--_.~.r,.m,..,wwm,.y,,,m , , , - , . , , , _ w.,w we r w w www- ww es=w-

• 27-VIII. GENERIC CORRECTIVE ACTIONS 2 O

1. Inspect, clean and replace (as necessary)1 shielded cable connectors on NI Intermediate Range Instrument strings.

2. Inspect, clean and replace (as necessary)1 shielded cable  ;

connectors on NI Power Range Instrument strings.

3. Inform B&W resident engineer of the grounding problem associated 2 with connector installation on the source range preamp. This information should be distributed to other plants which use the same model preamp.

~.

i i

i 1"As necessary" means replacing those connectors that cannot be cleaned properly, the tarnish removed or shows other signs of not being a quality connection.

2 More time is required to get an adequate supply of replacement connectors.

No known problems exist in the Intermediate or Power Range Instrument

strings which are less susceptible to elevated count rate and noise 2

, problems and have performed satisfactorily to date; therefore, this work will be performed at the next outage.

I bj e/17

,+-----,-r-- --

w *-.* -

gagm---.-w,.wy--,r. y- -.--- ,.w,-. ,w -y.---+wgamp.-yWw

_ . _ _ _ _ _ _ _ _ _ _ _ _ - - - . . - - - _ - - - - - - - - - - - - _ . - - --w. - - - - - - - - -

_3 _ . a .--%,, _ we _.a .+w -u._ _ e na -u--

E-+

• $h 9 e

t N# 43* f,48/it/ &[e(cdhb

2. Twintaxisi,dsbk

3. Monin/fnsu/*ted

!{ l (MI) odele ._.

~~

5 l l f__ _

I i

I g

!i l

' i t I l l

! l 1 g

l  ;

9.t'tr Ig i '

~~

I

! l -

1 II i 4

a 1

I I M'!A l, )ra,,,4TiL t

l I i f4.. a d i N i i

l 1 s 1 6,. s Ai) l n /g n.;4A i - ' l

• l A

A.r 4.s .~ is.i i

_.;.,'..n :18" = *E* '

_. i ~

j -**** ~

i /

I a ,

I  ; i M l Tu l

aakatarl=tsk!

~

fEMF-Ifdll I 3

-C:+

j .., if r,aow i

l i.e N.4s '

l l l ~~ I

, 8 -

4 - *l*** -

l 1 -s l l in i l -

l l< -g,.act.r Bu Id>~.9-

>g

, g g a n es s ts x

i l l '

i gaaet.r BuoIdiio)

I re,,atra tion

!. SovRcE RANGE i

TNSTRUMENT STRING 1 Fiji 2

I i

e 29

\SSEMBLY 8' #

$7RUC710NS n , , .

j l $ l surre j

is "

siiG

.e .

ga !' dpgiin" != 15!!!l7-ear waanes sasarr same saouse tusu w es name innee sus i :ww n vaansa c w ee er com mas u.e nue seev , sw er nue maats

r :sstact susuurse eser acx canacr seer h,7 prf }l!

,+ e -> Siice nut, wasner ana gasxet over ca:te. Cut off cutssce ;acxe (usmg ra:nr :taca or were sm:cers: s cimenston a Maxe a M $ ggg' cean c:st. :emg very careful nct ts mcx traic. C.:t first :raic to I 0 M. ,

E SUde first braid c:amo over traid us to sacxet of ta

.%=* l- )lla araid eacx cver c:ama, mammg sure trato is evemy cismeuted

..- aver me surfaca cf tne c:amo. inm see:no lacxet to cimen

!U c. agam catrg very careful not to mcx craid.

hI $

he Tnm second traid to emensi:n d. SIide en cu:er grcuna wasn h7

@@M E

e t?

msulator and second crarc c:amo. hta sec:nc 3. aid tacx stand r ame, again maxmg sure tnat trato is eventy cismtutec ever rctaca of c:amo.

l 3 4R$

E N

inm cante dieteemc to dimensica e.

g ig Ems g m

g gg,- >t4 Tiri ma inside hole of tDe cmtzer. fin wer g tze: anc solder. Remove any excess sorcer.e at,c insert mto

{j g =:E r. ;

.m - Se sure cante cierec.

g- g sii , lj ,

tnc ts not heated excessivery and swollen so as to prevent diesemc front emenng Omry of nimig.

$O

!="'F'sq .,E_;. p==I 5 gbP? Plug onty Placa front irtsulatsr anc outer c=ntact M ^_ -_, g cacx et cannec:nr tocy ann casa rrmr crecerInsert crace.assemcly ca:te- - int IQlMh,ig

"~

i contact assemely imo cocy. screw nur atts cocy witn wrenen

@d i E *An untif moderatefy egnt.

1 l sinoam, amm =twa4i na eeminunen y c. _-

~. "% i l I '

sea.!r tree l s, it free

[ h 5 _m 1

'*Ll I_ S )

a rucA

Y*

m -

b ksn.s)

~i _ _ mM '

e nu.zi t!.11 N f 4.31 tMis.1) 1*

I q d IMe,71 tt.3) h+s T g)

_ ===4 e it.

, .,,; , . lg agg.p -

f M:gula a.8)

s73) e W2.2)

- i s u..e rt,.

s'# -

i e r ros

( ' C Qt

• a D

iwis.ti  : % n.oi I

N'"." ~^ ~

_ r-

• :2w8.3) is.m t 5.1) mo6 4 t- )

• J- __

__r - ---- q , f w ..

h:i2.4 c". ,>31 l  :.21 f)Cf UtCOand Uf!::tyq3 si: *1a*4 ,** * ;*N

~

g.. g uns cimension is .147(4.51 a s =, s : :n ,g  :, .;)

AOb est&1f (9

\ / f W 9( NUCLEAR EDUCATION &~

TRAINING SERVICES, INC.

4500 CROMPTON DRIVE COLUMBUS, OHIO 43220 August 21, 1985 g8f-So3cf To: Mr. Mike Borysiak i CE!VF.D The Toledo Edison Company 300 Madison Avenue AUG 271985 Toledo, e Ohio 43$ -

NUC. FAC. ENG.

I It '!

From: on M le J g

Subject:

Summary of Meeting at The Davis-Besse Nuclear Power Plant, August 15, 1985, 10:00 AM - 4:00 PM.

NI-1 So'urce Range Nuclear Instrument Channel.

This memo is intended to summarize the activities, and our conclusions and rec.ommendations resulting from the recent meeting to discuss the operational difficulties with the NI-1 Channel.

Meeting Attendees: Steven A. Arndt, Joseph W. Talnagi, Don W. Miller, and Mike Borysiak.

Activities:

(1) Reviewed and discussed color photographs of many of the key

\ elements in the Channel between the BF3 Proportional Counter Neutron Detector (including the detector connections) and the Count Rate Amplifier Module (CRAM). The photographic study was directed primarily to the triaxial cabling and connectors since it is expected that the observed difficulties with NI-1 have their root cause(s-) in the signal lines between the detector and the control room. The study included significant attention to the penetrations and their installation.

(2) Reviewed and discussed the measurements, and data recor'ded by the Davis-Besse technical staff during' the past two months. These data included strip chart recorder time records, oscilloscope trace pictures, and time domain reflectometry (TDR) information.

(3) Observed strip chart recordings and oscilloscope traces of both NI-1 and NI-2 in operation.

Conclusions and Recommendations:

Two general and qualitative conclusions were made.during and following the completion of che above activities. First there are several potential contributors to the operational difficulties of NI-1 Channel.

These include a high stress on the cable at the input to the preamplifier, a recessed pin (.25") on the preamp input connector, high

! noise on the CRAM input signal, uoise pulses occuring with cabinet door f'

t ,

closing or opening, possible ground loop connection at the preamp input, cable connectors incorrectly installed or maintained, and low signal cable recistances. Second, there may be more than one root cause of the intermittent operation of the NI-1 Channel.

I

. Paga 7,1 l /"'s Since there are several possible items, many of which may be interactive, that may be contributing to the difficulty with NI-1, the j i combinatorial difficulty introduced into trouble shooting represents a l significant challenge to the Davis-Besse technical staff. As a i consequence, we endorse the Davis-Besse staff decision to proceed on the

' straight forward, but tedious and time consuming task of replacing the detector, and cleaning and maintaining each cable, connector and penetration in the NI-1 Channel beginning with the detector and ending at the input to the CRAM. Following each maintenance activity, the NI-1 ,

Channel response will be observed and changes recorded. By following

! the planned procedure there is a reasonable probability that the root

! cause(s) of the difficulty with the NI-1 Channel will be determined and .

1 a high probability th'e Channel will be repaired.

Although this approach is " brute force", there appears to be no other

)'

reasonable solution to the problem. We therefore agree with the approach being taken by the Davis-Besse technical staf f. Additionally, we recommend that laboratory evaluation of replaced components be made

, wherever possible, as is already planned for the detector-cable 4

assembly.

l t

cc: S . A. Arnd t

, B.K. Hajek

! J.W. Talnagi -

I t

4 i

l l

l l

4

!____._._______________-.____.~___,.___._,,_._________. . _ _ . -

)

l

[ FINDINGS, CORRECTIVE ACTIONS and GENERIC IMPLICATIONS REPORT l

\

TITLE: Toledo Edison - Source Range Nuclear Instrumentation NI-2 Channel 1 REPORT BY: Jack DeSando (TED) PLAN NO.: ISB Paul Alleman (B&W)

PAGE 1 0F 26 l

WRITTEN APPROVED

REV DATE REASON FOR REVISION BY BY J. DeSando and 0 8/26/85 Initial Issue P. Alleman L. Grier) 1 8/30/85 Title Change J. DeSando and P. Alleman -A M

J. Desando and ,

2 9/19/85 Added Data P. Alleman  %

I e

i i

! \

- . = _ - . .. -. - , - . - . -- ... _.. .-_. - .. .

l l

1 TABLE OF CONTENTS

?:.a_ge, I. Issue / Concern 3 II. _ Basic Principles of Source Range Nuclear 4 Instrumentation NI-2 Channel 1 III. Summary of Troubleshooting and Investigation 5 A. Field Actions Performed B. Significance cf Findings

} IV. Results/ Conclusion of Findings 18 A. Direct Causes 18 f B. Root Causes 19 V. Technical Justification of Finding 19 1

VI Specific Corrective Action 19 t i A. Required Corrective Actions 20 j B. Additional Planned Actions 20 VII. Generic Impl'ication 24 i

VIII. Generic Corrective Action 24 r Figure 1 - Source Range Instrument String 25 Figure 2 - Triax Connector 26 i

I i

0

, - - - - - - ~ , , - - , - - , - - . , , , - , - , ---+--~am--- ---e-- -- , .- w -----~~m4-,, ~r- e--- e-,,,,--,---w -_,,e,-.w---w---e,w,'w v w- - w e w w e, -- w

• w m w w -o n , ~--e

I. ISSUE / CONCERN Both NI Source Range channels were inoperable during the Reactor j Trip on June 9, 1985. This had no significant impact on Plant I operability but was an added distraction to the operators.

i Prior to the reactor trip on June 9, 1985, NI-2 indicated less than j 1 x 10~1 counts per second with the detector high voltage cutoff in effect. Source range high voltage is turned off when neutron flux level exceeds 1 x 10~9 amps as detected by both intermediate range f channels. A source range indication of less than 1 x 10 ~1 counts i

Per second is normal for this condition.

During the reactor trip on June 9, 1985, as neutron flux level de-creased to less than 1 x 10 s amps, as detected by both intermediate range channels, and high voltage was applied to the source range i detectors, the NI-2 indication remained at less than 1 x 10~1 counts i per second. Under these conditions, NI-2 should have registered S

l approximately 10 counts per second. NI-2 remained inoperable for approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after which a normal NI-2 indication was ob-served on the Control Room Indicator and on the local meter on the i

Count Rate Amplifier Module (CRAM) located in the Nuclear Instrumen-I tation/ Reactor Protection system cabinets. On two previous oc-casions, once when the RPS cabinet door was opened and once when tapping on the front of the CRAM, the count rate level indication returned to normal.

1.__.,--,.._

l Even though there was no significant impact during or following the transient on June 9, the inoperability of both NI-l and NI-2 did create an undesirable added distraction to the operator in having to verify the shutdown margin requirements per action statement six of T.S.3.3.1.1. From this standpoint it is highly desirable to locate I

and correct the problem with NI-2.

) II. BASIC PRINCIPLE OF OPERATION The source range instrumentation consists of two redundant count rate instrument strings, designated NI-l and NI-2, which measure neutron flux.

i The subject of this report is the NI-2 source range instrument string as shown by Fig. 1. NI-2 is located in,the Nuclear Instrumentation / Reactor Protection system channel 1.

l The source range signals originate in high sensitivity BF3 propor-tional counters located in thimble wells in the reactor vessel cavity.

The proportional counters are connected to preamplifiers, located in the reactor building. The preamplifiers are used to shape and J

amplify the low-level pulses from the proportional counters. The signal output of the preamplifier is sent to the count rate ampli-fier located in the Nuclear Instrumentation / Reactor Protection system cabinets.

l

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

The CRAM which contains a discriminator and Log Count Rate amplifier further shapes and amplifies the signal pulses. The signal pulses are sent to discriminators where background and gamma pulses are discarded. A log count rate amplifier converts the neutron pulses to a signal suitable for displaying the los count rate level both locally and remotely to the operator. The display range is from 10-1 to los counts per second.

The Rate Of Change Module (ROCM) has differentiator amplifiers which calculate the rate of change of the log count rate for display locally and remotely to the operator. An interlock is provided for control rod withdraw holu and alarm on high startup rate.

III.

SUMMARY

OF TROUBLESHOOTING AND INVESTIGATION i

l Maintenance Work Order #1-85-2030-00, 01, 02, 03 Note: As part of the troubleshooting procedures some repairs were made with approval of the Region 3 NRC representatives.

+ .

Action Plan Step 1 i

j Pictures were taken of cabinets, wiring, etc. to document existing conditions. Readings were taken at test jacks on all associated modules. The source used for these tests was reactor residual neutron flux. There were no anomalies noted in any of the readings.

i 6-f During NI-2 inspection and other troubleshooting the following problems were detected.

Significant Problems:

1. It was determined by test that penetration W 2 PIL1L for NI-2 signal cable ILRPSB03C contained a resistance substantially

, higher than the penetration for NI-1 or spare penetration 2 1

W 3 PIL1L when connected into the NI-2 instrument string.

I Operation with W 2 and W 3 (spare) penetration interchanged indicates that the W 2 penetration contains an anomaly.

2. The top BNC Coaxial Cable Connector (JC) on the Count Rate Amplifier was not locked in properly. The back of the CRAM has l

two BNC coaxial cable connectors as interfaces to the CRAM signal input and 15V output to the Preamp. The top BNC coaxial i cable connector could not be turned to its fully locked in i

f position. This probles has been corrected. l2 1

4 i

l, 3. The preamp low voltage cable was loose in the connector body.

This connector has been replaced. l2 i

i

4. The signal cable was loose in the connector body at penetration i

PIL11. This connector has been replaced. 12 i

O 9

-,,,,--.--,,,-,-.,,m . --mv--- , - + - , - - - ,i,,r-,,,yw .3

. ,w -...,-,,,.,,-r, ,,.7 -.-,-cmp-_ww.p.- , w--er w-o w e-e r v n=w r*ww - eee- ev w w.wm v w~~r

i i

l i S. Operation of Reactor Protection System (RPS), Safety Features Actuation System (SFAS) and Steam Feedwater Rupture Control System (SFRCS) cabinet door switches (which provide annunciator

, indication of open doors) frequently cause high level spikes 2 visible by use of an oscilloscope at the input to the CRAM. No alarms have resulted. The monitoring recorder temporarily connected to the output of the CRAM does not respond to these I

spikes.

l 6. The high voltage connector at the preamp is extremely loose (with no "0" ring in place). Wita the knurled connector nut tight, movement was possible between the two halves of the connector. This connector has been replaced. 12 f 7. No grounding wire is connected to the outside box which con-tains the preamp.

8. Every Asphenol connector inspected was tarnished and many contained metal flakes from the connector assembly.

i

9. Operation of some motor operated SFAS isolation valves have been identified as causing some spiking at the output of the l Rate of Change Amplifier Module (ROCAM). Observation of 2 i

j spiking vs. valve operation is continuing and is being logged

by the reactor operators. This observation will continue until a month of power operation is accomplished.

i l

Observations:

t

1. The high voltage cable connector on the back of the power supply appears to be nickel (sbiny) vs. silver plating. Silver plated connectors are specified by the vendor. This connector has been replaced. 2

2. The high voltage cable connector on the back of the power supply in RPS cabinet #1 is missing the "0" ring. Figure 2 i

j shows a typical connector assembly. This connector has been a

l replaced. 2 1

3. Time Domain Reflectometer (TDR) traces were taken using a i

Tektronix Model 1502 Time Domain Reflectometer. The Time Domain Reflectometer measures impedance vs. cable length by

, sending a high frequency signal down the cable and monitoring reflections coming back.

i 3

When the TDR trace for the signal cable was taken, very small impedance changes through the penetration to the preamp were 2 l

observed. The small impedance changes are judged not to be the cause of the loss of Count Rate problem.

t

4. The preamp high voltage cable connector does not have an "0" ring. This connector has been replaced. 12

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

_. _ .. - . . = .

1

5. The preamp detector and signal cable connector appears to be N

nickel (shiny) vs. silver plating. Silver plated connectors are specified by the vendor. These connectors have been 2 I replaced.

i i

, 6. Plastic bushings were not installed in the ends of the 2"

! conduits that enter into the back of the preamp outer box. The bushings protect cables from abrasion by internal threads on the conduit.

7. The detector cable bushing did not have a gasket where it is attached to the outer preamp box. This problem was corrected by installing a new preamp. 2

\

8. The four bolts which mount the detector cable bushing to the outer preamp box did not have lockwashers. This problem was corrected by installing a new preamp. 2 i .

l 9. The detector, signal and high voltage connectot , mounted on the outer preamp box were loose. The nuts could be turned with j ,

only fingers. This problem was corrected by installing a new '2 1

preamp. The paint was removed where the connectors attach to the preamp housing due to a grounding problem on NI-1.

10. Rust was observed in the connection box at top of the detector thimble. The seal plate apparently leaked at some time.

l t

11. The TDR trace for the low voltage cable shows very small imped-ance changes through the penetration to the preamp. The small l2 1

impedance changes am judged not to be the cause of loss of f Count Rate.

)

12. The printed circuit board mounting for the preamp was loosely mounted off of four slip over metal standoffs. This problem was corrected by installing a new preamp. 2 ,

i

13. In some cases fiber (shipping) washers have been left on the

=

i j bulkhead connectors installed at the preamp and RPS cabinets.

l This fiber washer prevents the connector from being tightened in place properly. The result would be that the connector body 1

i would turn during tightening of the connector. The fiber washers were inadvertently left in place during icitial construc-tion. All bulkhead connectors have been checked and fiber washers removed where they existed. 2 i

Action Plan Step 2 I

i To reproduce the count rate failure the NI-2 detector power supply was turned off and on while monitoring readings on the associated 4

j modules. This procedure was rapeated approximately three times per 1

I day over a 14 day period. No failures were observed.

l 1

l i

i l

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

l Action Plan Step 3 i

To reproduce the count rate failure, the associated cabinet doors l were opened and closed and the module face plates of the associated i

modules were tapped lightly while observing the applicable readings on the associated modules. No failur s occurred during this procedure.

. t Action Plan Step 4 i

j The Bench test of the CRAM was not performed so that the problem with the BNC coaxial cable connector would not be destroyed, see 2 significant problem 2. The CRAM was tested while installed and was I

eliminated as a possible cause of the failure (see Step 6). The l l CRAM BNC coaxial connector was corrected so'that it locks in place l properly. 2 4

l l

j Action Plan Step 5 l

l The detector high voltage power supply was removed, bench tested l

under load and monitored for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. No anomalies were found.

i l

l Action Plan Step 6 The object of this test was to verify proper operation of relay K1.0 in the CRAM. The CRAM relay K1.0 was monitered for proper operation while switching the test module from " operate" to " test" and back to 4

e

.ww-w - ,w-,- ,-:m , ymm -m -,,pww-wwr,,.mm--,w ,. en c- g-, , m e-- -vc=.-.w.-wwa,-rw-w,mm w-w w _w emwn m wwsm me-y e--e

i

" operate". Relay K1.0 contacts connect the test signal or operating i signal to ground dependent on whether the " operate mode" or test

" mode" has been selected. During this test no anomalies were found.

Action Plan Step 7 A recorder was connected to NI-2 monitoring the CRAM output, the Rate of Change Module output and the detector high voltage divided l by one thousand test jack. The intermediate range high voltage cut off bistable was " tripped" and " reset" in an effort to reproduce the loss of count rate problem. This procedure was repeated approxi-mately four times per day for 10 days. One failure occurred which lasted for approximately 4\ minutes.

O An additional channel of the recorder was connected directly to the detector power supply high voltage output and monitoring was started with the detector high voltage on continuously.

During continuous monitoring over a period of approximately 20 days, the following three failures occurred:

1. The count rate failed to 10~1 for three minutes then returned to normal and then failed for an additional four minutes before returning to normal All other recorded readings were normal.

2. Approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the failure in 1 above, another .!

count rate failure occurred for less than one minute then j j returned to normal. All other recorded readings were normal.

3. Approximately 7 days later a new preamp was installed and the recording was continued. Within 24 he trs another count rate ,

failure occurred which lasted for 1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. During the failure

• the low voltage (+15) was measured at the preamp and was found I  !

to be within limits. When the high voltage was measured at the preamp the count rate returned to normal. The connectors at the preamp were checked for tightness. When the signal output (to the RPS) connector was rotated, spiking and elevated count I i rate occurred. The elevated count rate remained for some period of time. No problems were observed with the other i

i preamp connectors.

l 4. After failure #3 occurred triax cable connectors at the preamp, the detector MI cable /triax cable interface and accessible '

) connectors at the penetration were being inspected / cleaned or j ,

replaced as necessary while maintaining operation of the 2 i channel. During this time the channel operated for 12 days i without loss of count rate then the following failures occurred:

Lost count rate for 8 minutes; 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> later lost count rate for 10 minutes; 42 hours4.861111e-4 days <br />0.0117 hours <br />6.944444e-5 weeks <br />1.5981e-5 months <br /> later lost count rate for 2 minutes; and 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> later lost count rate for 1 minute.

[' On the basis of resistance tests taken under Action Plan Step 9, Item b (see below), the penetration was indicated as possibly containing an anomaly. Penetration FW 2 P1LIL was interchanged with spare penetration FW 3 P1L1L and the channel placed back in operation. The channel operated in this config-uration for 9 days without loss of count rate.

During the time spare penetration FW 3 P1LIL was connected into the NI-2 channel an independent test was run on penetration FW 2 P1LIL per the following: A 0.5 volt DC source was l 2 applied to the FW 2 penetration center conductor, inner shield and outer shield. Each of the three conductors was monitored by a separate channel on a recorder for the loss of the 0.5 VDC signal applied. During the six day duration of this test three separate failures (0.5 VDC was lost) occurred on the center conductor.

The penetrations were then returned to the original configura-tion (FW 2 connected back into the circuit) and after 21 hours2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br /> operation the following failures occurred: Lost count rate for 55 minutes; 25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> later lost count rate for 7 minutes; and 47 hours5.439815e-4 days <br />0.0131 hours <br />7.771164e-5 weeks <br />1.78835e-5 months <br /> later lost count rate for 17 minutes.

f Action Plan Step 8 (a) The NI-2 preamp was removed and bench tested. The pulse ampli-tude for the times 10 range was slightly high on bench test.

In accordance with the vendor manual pre-installation acceptance

4 l

I test procedure, the output pulse amplitude limits are -1.9 to

-2.1 volts for the times 10 range. The reading for this preamp was -2.2 volts. No other anomalies were noted.

(b) A new preamp (obtained from VEPCO) was installed and monitored for three days. No failures occurred. The NI-1 preamp was removed and installed in the NI-2 channel and monitored for 1\

days. No failures occurred.

(c) The original NI-2 preamp was reinstalled and monitored for two days. No failures occurred. At this time the preamp was removed and taken to the vendor for complete evaluation. 2 Vendor test data has been issued by meno from Bailey Controls Company (Harold Sternberg) to Toledo Edison Company (Jack

)

Kasper) dated 8/22/85. The Bailey Controls Company conclusion is that the preamp operation is satisfactory.

A new preamp (obtained from Consumer's Power Company) was installed and the failure as described by Action Plan Step 7,

+

1 failure #3 occurred.

Action Plan Step 9 (a) Time Domain Reflectometer (TDR) traces were taken using a i

Tektronix Model 1502 Time Domain Reflectometer. The Time Domain Reflectometer measures impedance vs. cable length by sending a high frequency signal down the cable and monitoring

reflections coming back

16-The cables checked are as follows:

1. RPS cabinets to the preamp

a. signal cable

b. high voltage cable

c. low voltage cable

2. Preamp to detector

a. signal /high voltage cable The TDR test results are:

1

1. The signal cable from the penetration to the preamp shows very small impedance changes. The small impedance changes are judged not to be a cause of loss of Count Rate.

2. The low voltage cable from penetration to preamp shows very small impedance changes. The small impedance changes are judged not to be a cause of the loss of Count Rate.

(b) Resistance readings on the high voltage and signal cables were taken at the RPS cabinets. The high voltase and signal cables were disconnected at the RPS and the preamp. 2 Appropriate jumpers for the conductors being measured were

, - , - , - - + , - - - ---- - .,_m .ywv.,,,w,-y- -wye,.,

connected on the cable at the preamp. Resistance readings were taken from the center conductor through the inner shield, center conductor through the outer shield and inner shield through outer shield. Comparative readings were taken of NI-1, NI-2 with penetration W2 PILIL connected and NI-2 with spare penetration W3 PILIL 2 connected. It was determined that the Center Conductor resistance for NI-2 with penetration W2 PILIL connected was more than double that measured for NI-1 or NI-2 with spare penetration W3 PILIL connected in place of pene-tration W2 PILIL. Operational tests were also performed, see Action Plan Step 7, Failure #4 for details.

, Action Plan Step 10 i

Detector checks were taken for both NI-1 and NI-2 at the preamp locations. Readings for NI-1 and NI-2 were similar and did not show any anomalies.

l l

i I

k J

4 Action Plan Step 11 Inspection, cleaning and replacement (as necessary)I of connectors has started. See Section VI A. 1. required corrective action for details.

4 B. Significance of Findings The significant findings are:

1. Failure of penetration W 2 PIL1L l2

2. The poor state of maintenance of the Asphenol Triax

, Connectors. -

i IV. RESULTS/ CONCLUSIONS OF FINDINGS A. Direct Cause

1. The direct cause of the loss of cou~nt rate was failure of 2 penetration FW 2 PIL1L.

2. A secondary cause of loss'of count rate was the degraded 2 condition of triax connectors.

1 "as necessary" means replacing those connectors that cannot be cleaned properly, the tarnish removed or shows other signs of not l being a quality connection.

l i

B. Root Causes

1. The direct cause of the loss of count rate is the pene-tration, however within the penetration the root cause cannot be established until the prenetration can be  ; 2 removed and evaluated.

2. A secondary root cause of loss of count rate has been established to be the degraded condition of the triax con-nectors. The degraded condition is center pins out of 2 place due to improper assembly, missing parts and lack of proper cleaning before mating connectors.

b V. TECHNICAL JUSTIFICATION OF FINDINGS The cable resistance test data included as part of action plan -

step 9b and operational test performed on Action Plan Step 7 is justification that the direct causes are the intermittent penetra-tion and degraded condition of triax connectors. 2 On removal of the penetration it will be evaluated for the specific fault.

VI. SPECIFIC CORRECTIVE ACTIONS Corrective actions are being performed with the approval of the Region 3 NRC representatives. ,p

h A. Required Corrective Action

1. The penetration has been determined to be intermittent and will be replaced prior to startup after the next refueling outage. In the interim spare penetration i

FW 3 P1LIL will be used for NI-2.

2. Inspection, cleaning and replacing (as necessary)2 shielded cable connectors from the detector to the RPS cabinets is in progrest. All connectors for the 2 NI-2 string from the detector to the RPS cabinets have been inspected / cleaned or replaced except at the penetration. The connectors are being replaced using installation instructions written in the Maintenance Work Orders. Selected personnel have been instructed

.1 I in the proper installation procedures by the system vendor representative.

I

! B. Additional Planned Actions i

l

1. Revise or prepare new procedures for installing triaxial cable connectors. The current procedure is MP 1410.24.

Due to the Generic Corrective Actions as outlined in 2 I

. 2 "as necessary" means replacing those connectors that cannot be cleaned properly, the tarnish removed or shows other signs of not 2 being a quality connection.

J e

l

Section VIII, this item will be completed prior to the next refueling outage. l

2. Provide cable coriaector installation training for station personnel. This item to be completed prior to further cable connections. Work is planned to commence during the 2 next refueling outage and training must be completed prior to this.

3. Reinstall the 2" plastic bushings in the conduit that enters the back of the Preamp Outer Box. This wo;k will be accomplished during the next refueling outage. 2

4. Add grounding wire to the outside box which contains the Preamp. An engineering evaluation will be completed this fuel cycle to determine if this ground wire is required. 2 The ground wire will be installed during the next refueling outage if required.

I i 5. Replace the seal plate gasket for the connector box at the top of the detector thimble. This gasket will be replaced prior to this startup. 2

6. Replace source range NI-2 Amphenol penetrations with new Conax penetrations. This action is necessary in that the existing penetrations contain triax connectors that 2 i require penetration removal to inspect or replace the

- - - , - .- --- , - n.-- ---. , --

w---.- .-. --- --.-,n, . . - n-, . - , ,n,-,, ,_e.---,w. . , .

l l

1

, ) connectors. The existing penetration is intennittent and a spare is being used in the interim time period. The new Conax penetrations will eliminate one connector per cable 2 and the connectors will be accessible after penetration installation. The replacement of the penetration modules will be completed at the next refueling outage.

7. Review and revise (as required) procedure number MP 1504.04 for adequacy in determining if the seal plate rubber gasket, for the connector box at the top of the detector 2 thimble requires replacement. The NI-2 rubber gasket will be replaced prior to this startup. The gasket replacement s concern will be addressed prior to the next refueling outage.

8. Revise procedure number IC 2002.02 so that it adequately addresses all areas of NI detector renoval and installation.

No further detector removal / installation is planned. 'This 2 item will be completed prior to the next refueling outage.

9. Revise procedure number IC 2002.04 to address current requirements for source range discriminator and high voltage settings. This ites will be completed as part of 2 the annual procedure review or prior to the next refueling outage, whichever comes first.

l l

(_,, 10. Revise procedure number MP 1410.24 to address current requirements for installation and cremoval of Raychem heat shrinkable tubing. This ites will be completed as part of 2 the annual procedure review or prior to the next refueling outage, whichaver comes first.

i

11. Continue observation of spiking vs. SFAS isolstion valve operation and look for/corrolate with any other equipment actuations that cause spikes at the output of the rate of 2 change amplifier module and perform tests to determine source of spiking. This observation will continue until a month of power operation is accomplished.

12. Prepare an interim administrative procedure for operating with SEAS isolation valve spiking present and any other equipment actuations that cause spiking pending (if 2 possible) location and elimination of the source of the spiking. The SEAS isolation valves would not normally be operated during plant startup or shutdown therefore, this would not present a problem to operation during startup or shutdown.

l l

13. Recorder monitoring of the NI-2 instrument string will continue until the start of power operation. 2 i

VII. GENERIC IMPLICATION Triax shielded cable connectors on the other NI systems may have conditions similar to those found on the source range instrumen-tation.

3 VIII. GENERIC CORRECTIVE ACTION l2

1. Inspect, clean and replace (as necessary) shielded cable 12 connectors on NI Intermediate Range Instrument Strings.

2. Inspect, clean and replace (as necessary)' shielded cable l2 connectors on NI Power Range Instrument Strings.

I bj e/18 3 More time is required to get an adequate supply of replacement 2 connectors. No known problems exist in the Intermediate or Power Range instrument strings which are less susceptible to elevated count rate and noise problems and have performed satis-factory to date; therefore, this work will be performed at the next refueling outage.

3 4 "As necessary" means replacing those connectors that cannot be V cleaned properly, the tarnish removed or shows other signs of not being a quality connection.

Noles : 1. uAuv rr:a cat.

2. ne;><asiai o,pa

s. u,sawr,,sawa l (MI) caak . _ . - -

F__

3 '

I l .

I l 1 '

I l .

I

~

I l i- I TW M *lf I p -

l

^

I l l l l m,. 4 l- l l

I

""" I I ll gg'd,., g jc6 ,

4 I

I j.. -* 46~ " ' 8 v i I e 1 l 0 2 l ,,,,.., I  :

.~ hdiff,rHZ ! =

E.MEx l

l k n,;,,44 4 l 9/

• !E. I J= faker l 5 n l l l sde i

% $*./>! N @ I s. <

p y'r7

rc#:+r 4

l i' u,e.-

\ behet,y locatcJ I l sn U.kWe l l

.g<a i -

l i

I l

l- l l ni ll .

I 1

< geact.r bullh>;9 l l l Aff Cdh58/5 g~

>l l r

haet.r 51usId*kJ

&,,,S tan

! sov8CE RANGE

.TNSTRUMENT STRING g .

-e S

1 F.36 3 L s-- A 7

, . Page 26

, &y n

y , 1, SSEMBLY 8' #

l LNSTRUCTIONS n ,

~w w

$ 58=.1ElW l 39 l h, h h 5, *~

i  : "

jugestaE cumes wasmas cuw =e couract Ea rce

~

sYc EITET Noi Slide nut, washer and gasket over cab!a. Cut cff outside Jacket g ft-)-

g (using razor blade or wi.'t strippers) to dimensica a. Make a g,4[ gg clean cut, being very careM not to nick brard. Cut first braid to g gg, sin _ ,g dimension b.

hN b 8--> Slide first braid efamp over braid up to jacket of cable. Fold first s

hI g braid back over c: amp, makmg sure braid is evenly distributed over me surfacs of me c! amp. Trim second jacket tc dimension

c. again being very careful not to nick braid.

h  ! O (e>

• Trim second braid to dimension d. Slide cri outer ground washer, insulator and second braid ciamp. Fold second brald back over l 5 q'yE E f braid c: amp,,agam making sure that braid is evenly distributed over surface of c! amp.

gl Q $? Trim cable dielectric to dimension e.

Jg g g g ,- >fg Tin me inside hofe of me ecmact. Tin wire and insert into con-(4) =g tact and solder. Remove any excess solder. Se sure cable dielec-

] g g F :

Q a

tric is not heated excessnrery and swoften so as to prevent dielectric from emenng body of fitting.

Plug on!y: Place front insulator and outer centact assembly into

$: 11 1 5 15 back of cennecter body and push ir:ta proper place. Insert catte-comact assemtly into body. Screw nut into bocy with wrer:ch 5y i aje.g until moderately tight.

i OI El Sincomo dias. =WQ.4) incnes (tmnimeterst Pfuas I saA.19 Type l L 11 Type I"*__*C "'il h-__.i i F.5i,3 i %ni'i

~

f

% a .

l %3.5) %f3.2)

J- __~- Jacks 54A. 39 Tree l 8 Troe h h ans. M- .? ~ J B  % 1 I e 1

4) .9

-an=== e 4(6.4m W7.9)  ;

%Jf! - - . -__- __ _ _ n f ):dZ.4, 4 (3.21

. ES' l!!' cs!!

u- -

Os ' _ ator 531co and 53150 civos e tar 344co and 34375 iacas tms emension is . s7(4.53 tras annension u . :o<3.33 1

. I

FINDINGS, CORRECTIVE ACTIONS AND GENERIC IMPLICATIONS REPORT TITLE: TOLEDO EDISON - SERVICE WATER TRANSFER REPORT BY: TIMOTHY CZUBA (TED) PLAN NO. 26 BRUCE HICKMAN (TED)

JAMES TABBERT (TED)

PAGE 1 of 13 WRITTEN APPROVED REV DATE REASON FOR REVISION BY BY Tabbert Czuba * #'"

0 8/17/85 Initial Issue P. Hickman _

Tabbert Czuba O 1 2

8/23/85 8/29/85 Corrective Actions Added Testing Clarification B. Hickman Tabbert Czuba B. Hickman L. Grime L. Grime Updated Root Cause and

, 3 9/9/85 Technical Justification L. Grime D. Mominee

~  !

O 4 9/17/85 Final Issue B. Hickman 3 ,.fg u e r-= ~

1 l

l l

l l

s

, - - - - ..-,-.,,m.-----,-c- ~, - - , , - -., . ~ . . - +---r-,.._-y- < - - - - - . , - , - . - - - , - , ,,--*w e- ..'gt'-w-v

- en v w-Mr '-rr --*

2-TABLE OF CONTENTS I. Issue / Concern 3

II. Basic Principles of Motor Actuator 3 III. Susanary of Troubleshooting and Investigation 4 A. Field Actions Performed 4

, B. Analysis Performed 5

)

C. Testing Performed 8 D. Significance of Findings 8 IV. Results/ Conclusion of Findings 11 A. Direct and Root Causes 11 V. Planned Additional Actions 11 VI. Specific Corrective Action 11 A. Required Corrective Action 11 B. Additional Planned Action 12 VII. Generic Implications and Corrective Actions 12 VIII. Generic Implications and Corrective Actions 12 Figure 1 Auxiliary Feedwater Pump Suction Schematic 13 Figure 2 Auxiliary Feed Pump Suction Pressure 14 Figure 3 Auxiliary Feed Pump Suction Pressure - August 21, 1982 15 Testing l

l l

l i

e e

- _ - - _ c.,._ . ..,y._. . -y_,_,,,,_,,-.p..,e,,,,. , , , ,,_m,. , . - . , , , , , . , . , , ,.-,,,_.-n..,m,_. .,m. , , , ...,.m,w, ,, ,m.,-.,,

l i

-3 l l

I. ISSUES / CONCERN:

During the June 9 transient, an inadvertent automatic transfer of the suction of Auxiliary Feed Pump 1-1 (AFP 1-1) from the condensate storage tanks to the service water system was actuated.

i The automatic transfer of auxiliary feed pump suction from the condensate storage tanks to the service water system occurred prior to the actual loss or low water level in the condensate storage tanks. Service water will allow the auliliary feed water pumps to perform their safety function. Service wrter is chlorinated lake water which is chemically undesirable for use in the steam generator.

m J

% II. BASIC PRINCIPLES OF OPERATION i In the event of loss of the water supply from the condensate storage tanks, an automatic backup is provided from the service water system.

Two independent low pressure switches provided on the auxiliary feedwater pump suction line, upon both sensing low pressure, will I

automatically close the valve from the condensate storage tanks and l

! open the valve from the service water system. Each auxiliary feed-water pump has an independent set of pressure switches, supply valve from the condensate storage tank, and supply valve from the service j water system. These components function independently with their respective auxiliary feed pumps. See Figure 1.

A U

l

.- , - - , , _-. - - - , . . _ _ - , . _ _ , - _ . _ _ - . , , . _ _ . _ . _ - . . , - _ . . _ - - . _ _ _ _ . _ - - _ , + . _ _ . . _ _

( III.

SUMMARY

OF TROUBLESHOOTING AND INVESTIGATION 1

A. Field Actions Performed NWO 1-85-2130-00 and 1-35-2133-00 checked the time response, I

calibration, condition of the switch, and wiring of PSL 4928A, PSL 4928B, PSL 503, PSL 4929A, PSL 4929B, and PSL 507.

l j PSL 507 was found to actuate at 9.36 PSIG. The setpoint for PSL 507 is 11 PSIG. PSL 507 monitors the auxiliary feed pump 1-2 suction pressure downstream from the strainer $206. PSL 507 '

provides an alarm indication in the control room, but no control function.

PSL 4928A was found to have a loose terminal connection. It was determined the connection was making contact and had no effect on the operation of PSL 4928A. Review ci the circuit diagram shows the loose wire would not affect the actuation of the automatic transfer.

MWO 1-85-2146-00 verified there were no obstructions in strain-i ers S201, S206, or S257. The strainers were disassembled and no obstructions were found.

NWO 1-85-2144-00 verified that upon sensing low pressure at PSL 4928A and PSL 4928B, the suction to auxiliary feedwater pump 1-1 would transfer from the condensate storage tanks to the service

f\

J water system. Low pressure was applied to PSL 4928A an'd PSL 4923B simultaneously and the suction transferred correctly.

B. Analysis Performed The effects of vibration on PSL 9428A and PSL 4928B were investigated. The evaluation concluded that vibration was not a credible cause of inadvertent actuation of PSL 4928A and PSL 4928B. The corclusion was based on vibration analysis of the supports and switches.

Another analytical study investigared the pressure drop in the pump suction line as a potentia $ cause of the automatic trans-

's_,- fer of auxiliary feed pump suction from the condensate storage tanks to the service water system. The study concluded that 4 the conical strainer, S257, (which was added to the system in 1981) may significantly reduca the steady state suction pres-sure and thus increase the probability of suction pressure dropping below the setpoints of the pressure switches PSL 4928A

! and PSL 4928B, activating the automatic transfer. It should be noted that the calculations were initially based on a calculat-ed pressure drop for the conical strainer S257. Subsequent testing (see Section C.) showed that at rated flow of 2100 gpm the suction strainer would be estimated to reduce the steady pressure at the pressure switch by 4.6 psi.

s

\

's A study was conducted to verify the circuitry for the automatic transfer of 'the auxiliary feedwater pump suction was of correct design and logic configuration. The study concluded:

• The circuitry will provide the automatic transfer of the auxiliary feedwater pump suction upon sensing low pressure at both low pressure switches. (PSL 4928A and PSL 4928B for AFP 1-1. PSL 4929A and PSL 4929B for AFP l-2).

t

• No single failure or inadvertent actuation of any single device will cause the automatic transfer of the auxiliary feedwater pump suction. The automatic transfer requires two (2) out of two (2) actuation of the low pressure switches.

• Barring multiple or common mode failures of the pressure 4 switches, actuation of the automatic transfer of the auxiliary feedwater pump 1-1 suction was caused by a low i

suction pressure which was sensed by PSL 4928A and PSL 4928B.

l Investigation of the June 9th and previous documented auxiliary feed pump suction automatic transfers from condensate storage tanks to the service water system has shown:

.. - - ... -- . , . - . . , . , . - . , , - . , . - - , . , . . . . . . . , - - . . - - . _ - ..y-,m- - . . . -

w, . - -

. 7 r%

• The automatic transfer has occurred four (4) times: July 30, 1981; November 9,1983; January 15, 1985; and June 9, 1985. The earliest transfer was on train 2, the three later transfers all occurred on train 1.

• All transfers occurred after the addition in 1981 of strainer S257.

The three early transfers occurred when the auxiliary feed flow.

was increasing, i.e., as a pump was accelerating. The transfer 4 on June 9, 1985, however, occurred coincidently with the cut-off of steam to AFPT 1-1. The alarm printout for June 9, 1985, shows the automatic transfer took place at 1:58:40. The Csn nd same alarm printout shows that the AFPT 1-1 trip throttle valve left the open position at 1:58:39. As explained in the find-ings report for Action Plan ID, this corresponded to the spurious trip on the improperly reset trip throttle valve and the rapid cut-off of steam to AFPT 1-1. Data from the June 9, 1985, incident show that the flow from AFP 1-1 to the steam generator dropped from about 500 gpm to zero in less than two seconds at that time.

I

r C. Testing Performed Test procedure ST 5071.01 (T-mod 9544) (performed on September 6, 1985) determined the steady pressure losses in the line from the auxiliary feed pumps to the condensate storage tank. This was done with AFPT 1-2 operating on auxiliary steam and recir-culating through a large recirculation line as well as the normal recirculation path. Based on these data the pressure at the pressure switch location can be calculated for other flow rates 2nd other levels in the condensate storage tank. These results are shown in Figure 2.

Observations of transient pressures during startup of AFPT 1-2 4 were made in August 21, 1982 using test procedure TP 520.48.

In this test a pressure transducer was placed in the pump suction and the pressure was recorded as the SFRCS was tripped and auxiliary feed pump 1-2 started. Figure 3 shows the dip in suction pressure which resulted as the pump accelerated. This trace also shows other fluctuations in pressure which are' evidently related to flow through the system, but are not the i

result of acceleration of the turbine.

l D. Significance of Findings Investigations have shown that the automatic transfer of the A auxiliary feedwater pump suction was not due to a mechanical or

~ electrical malfunction of PSL 4928A or PSL 4928B. PSL 4928A l

l

_ _ _ __ _ _ _ _ ._ _ __g...___ __

9-i and PSL 4928B are mechanically and electrically independent.

An automatic transfer of the auxiliary feedwater pump suction I caused by a malfunction of PSL 4928A and PSL 4928B would require both to fail at the same time. Therefore, a transfer of the auxiliary feedwater pump suction by a malfunction of PSL 4928A and PSL 4928B is not credible.

Review of historical data shows that three suction transfers had taken place prior to the one on June 9, 1985. However, these transfers were fundamentally different from the transfer which took place on June 9, 1985. The other transfers took place when an auxiliary feed pump was increasing in speed. 4 Consequently, it is concluded that they were a direct result of

{ the suction pressure dip which testing has confirmed occurs when a pump is accelerated. They probably also involved conditions which made this dip deeper than usual, involved i

j other hydraulic pressure fluctuations, or involved lower

, condensate storage tank levels. The addition of the suction

, strainer S257 to the system reduced the margin between the dips i

and the pressure switch setpoint so that instances when the dip l

! could actuate the switches became more probable. Although the

{

dip in suction pressu're is necessarily brief (2 to 4 seconds) 1

! this is long enough to actuate both switches. (The switches have no intentional delay.) The duration of the suction

! pressure depression is limited because a high rate of flow acceleration must be terminated or the pump will trip on overspeed.

.,.-g- -,- - , , - . + v--ew=-y-% -

, gr e m -e.9- --m+, w+ g-e-ehe,ye a g up -pg ev--e-e=--tww-w-M*t-'Tv-M W- -*-*'*w-*w-'e*g - t % =wr-e-=er w mew,vwa m _ww=--u+*,-ewa--wa.--e

- -10' The automatic suction transfer on June 9,1985, was different from the earlier transfers; therefore, its direct cause was also different. There will be a significant hydraulic tran-sient in the auxiliary feed system associated with the closure of the trip throttle valve for AFPT 1-1 while AFPT 1-2 is running at essentially full speed. In particular, check valves in both the outlet and inlet will be actuated as the flow is decelerated and potentially reverses. The rapid closing of check valves is capable of generating significant pressure pulses (both positive and negative).

The magnitude and duration of the pressure transients which j ' occur from check valve closures depends on the particular 4 parameters of the valve and system. The detailed inf'ormation on the flow chang.s during the June 9 incident which would be necessary to calculate with confidence the pressure transients is not available. However, it has been repeatedly demonstrated in fluid systems that the waterhammer from check valve closures can easily result in pressure pulses of the magnitude (about 20 psi) which would be needed to reduce the static pressure to the switch setting.

(

l l

e l

C-I

• t

l IV. RESULTS/ CONCLUSIONS OF FINDINGS 1

1 A. Direct and Root Cause The automatic transfer of auxiliary feedwater pump (AFP) 1-1 on June 9, 1985, was a direct result of the hydraulic transient in the auxiliary feed system which resulted when the turbine trip throttle valve for the turbine for AFP 1-1 spuriously closed shutting down the pump.

The root cause is considered to be a switch and control system

]

configuration which was capable of responding to rapid hydrau-lic transients and pressure fluctuations even though the 4 C suction transfer was not required, or even desirable, as a result of those conditions.

V. PLANNED ADDITIONAL ACTIONS l

l, Hypothesis 7 is to be investigated.

, VI. SPECIFIC CORRECTIVE ACTION A. Required Corrective Action ,

l i i  !

i

1. Add a time delay to the actuation of the suction transfer  !

l to assure that brief drops in suction pressure from  !

I hydraulic system transients do not cause the system to )

I actuate. i r

~

[ 2. Reduce the suction system pressure drop as much as practi-  !

\s ~

cal by increasing the suction strainer (S257) mesh size.

B. Additional Planned Action Proper operation of the switches and the operating margins will be confirmed by monitoring suction pressures (both steady state and dynamic) during auxiliary feed pump testing. This testing will be conducted after the corrective actions have been perfo rmed.

VII. GENERIC IMPLICATIONS AND CORRECTIVE ACTIONS 4

The fluctuations in pressure in the auxiliary feed pump suction system during brief hydraulic transients may affect the operation of other pressure switches in that system. These other pressure switches and associated control systems will be reviewed to add appropriate delays and, if necessary, make changes in setpoints to assure that automatic actions and alarms represent conditions which require such automatic action or oterator concern and are not merely the normal consequence of system operation.

se e/17 '

i

~

2 I

AUXILIARY FEEDWATER PUMP SUCTION SCHEMATIC 31.5'l' CCKDEtGATE , ,

l STf.RME TANth

(

5W1372.

pg SERVIG WATER Y PDSH Sol A92sG Pool - - .

PCO B --O P00to rw7sto l AFPl-1 7

S20\

/Ml l

!O P5L 49285 O .5EX7/E LJATER Sid 1395 h73g

PDSH 50a pgg.

l PCOS - - -

'807 ru n o ,'

are n 1. , /-K ,*

szou ,'

LO P3L 9121A V

STRRT UP FEED PMP.

FIGURE 1 l

F J

'V 35 . ,

g a-  ;

- .. .. . .l.,

. . .,l .. -

f.. j; i !

[ ..

. Ap

,g 30- .f '

CD lI m- ---

+=  ;.. w= LE #W Q+,. .j , +.

~*

r j , / -

n.

e

/,

Ih h

<,< . ~

/ p/

x : ;It

{)/ ..

I I

,, fI h#y hf f . u ,{e.'.

~ ~

15 j,f,4'/-

, _, .; g,, 4 i , ,,;

k_-_.,+,L l ,'

$ , / ./ t ,/: ,

l::  ! ,!:, ,'

.h f /.' / -,- yt", ,

>. +

yd :yd

=  %, mi 4 .

Mi H! I '# . '

C) o  !* 44

M-'fD 'I / / 'a/ .+ ii'+F # l ll;;

if N@$';6/ l ' )f 4 i!/y' it I

!) ..t!p ill!

lii:

Il i '

e',<

/ g H-

/

v A .

>N r-

,j l. n _: p*.

in n-

_1 i, g

--y-

.L .

-/

l.--.l ,

.4 7

4

-}, 'i i,..

i *l T qo

p. .

.1L!} l

_)( :1 11 r

h. . .U :: 1 II;l  !'i: ..p;l..

}.

l- .} / l'/5 l l[' ' l  !  !! ll',

'I'l l!'

2#--. .lI1

... . 1 . . I .l -l ! I l / -

'--l ( -

't 4

l}![ 'jj lj

~' j .!. !,!! !.i.'.

jj/

}pf

/ . ___. .. . .

l77

.M___,._.. ., 4 l

i

...l i  !;-

__.... 7

..-:. 1li'

. . .~:.. .. - . .-

.i

-10 0 10 20 30 4'O 50 60 Invel in Condensate Storage Tank (Ft)

O o

AUXIUARY FEED PUMP SUCTION PRESSURE FIGURE 2

i Time After SFRCS Trip Seconds

~ 17 18 19 20 21 22 23 L L L

.a

. . i, e ... . ,L r

- - - r

- - - - r L

r-

, ...L r - - - - i 30 k I IE II ! I!il 1Ii i!II EE 2

" -~

N= i  !!!! I II!5 ~=  ! ~

• i 'III !!! E ifl!

md'

.a 27- *:= Ei- ir fiji !M ?;l t:4 T;i Cii n; , iii -it' -:

nu m

t .r; hi 'J! lI 4

l- i. , ii*

i -

24_

g so .

c ..

-

-a e-n-

=.

= -, - - - .. = p a...

2_  ;.  :

n a

, ip_fi l .

gg j =- 23;;g

g'.g j g; 21 yg gg g .=,.

g .;;. .ry y g

.. ;3 4; 3;  ;;;3 j jag.:;*= gd 7 3

= _._.  ;;33 ..

g

_3 ,

?:t r r H= ]s rr inm g =

tm . = i v ie a .:.__  :  ;.i. .

- g .; p.; :m  :

- rg p,. . v

ur 'j s .

g,, g .q g. _

jp  ; _  :.gj-p g g gj 1 .

g_ .

3,; . .

p g  ;  :.. .

a .; _

g g _

g e

6 M iii --

*4 i-ii nn '

i-s t: _

4 !:;  ;-d  :+ -e .- . ._

is ~; b =-! 16 ,

4 4 .

--- = -

fI u..:  :- .

-u .  :-- . .. .__ .. ,. r_:

I =:- ~i p

'I

. . . . 2

"- '" ~~ * *' ' -~~ ~ ~

A '

~

' ' v' 3 '

I' fi" IEI jE- I

' ~~

i ' ! P'

~ ' ' '""

{

~ '

2" '

F I 3 i l" EIi '

E5  ?

l H"

0 t . .

e s a e

u 1

e u a s y a e e a w e w e e e a s W 8 u e v 5 e * * * '

• u *

• 4 1

1 4

Note:

- Pump Starting Dip - Fluctuation f At final speed, pressure Approx.10 psi Approx. + 1.5 psi _

was approx. 18.5 psi. ,

4 i

i i i

4 i

AUXILIARY FEED PUMP SUCTION PRESSURE AUGUST 21,1982, TESTING FIGURE 3

\ _

TOLEDO

%mm EDISON Docket No. 50-346 JOEWILLIAMS JR.

Serw Ves Pres 4 set Ahrlear License No. NPF-3 {$ll$$.l$$$

Serial No. 1188 October 1, 1985 Mr. Harold R. Denton, Director Office of Nuclear Reactor Regulation United Stated Nuclear Regulatory Commission Washington, D. C. 20555

Dear Mr. Denton:

G The enclosed materials comprise Revision 1 to the Davis-Besse Course of Action submitted to you on September 10, 1985 (Serial No. 1182). A list y of effective pages and document control receipt containing instructions for revising this document are included in the revision package. The document control receipt must be returned to Toledo Edison Company to ensure compliance with our document control procedures.

F ~

Very truly yours, 9 4,

. Williams, Jr.

Senior Vice President, Nuclear JW/VJM/bjs Enclosures cc: DB-1 NRC Resident Inspector oil

., s THE TOLEDO EDISON COMPANY EDISON PLAZA 300 MADISON AVENUE TOLEDO OHlO 43652 L_ _

J