ML20082E034

Responds to NRC Re Violations & Deviations Noted in Electrical Distribution Sys Functional Insp Repts 50-254/91-11 & 50-265/91-07.Corrective Actions:Temporary diesel-driven Compressors Will Be Used in ILRT

ML20082E034
Person / Time
Site:	Quad Cities
Issue date:	07/24/1991
From:	Kovach T COMMONWEALTH EDISON CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9107310036
Download: ML20082E034 (41)

Text

ON Comm:nw:alth Edison g [ ) 1400 Opus Place (ONs Downers Grove, Illinois 60515 July 24,1991 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555

Subject:

Quad Cities Nuclear Power Station Units 1 and 2 Response to Electrical Distribution System Functional ort 50 254/91011,265/91007 Inspection (EDSFI)

NRCl)ocket NosJ10-2 Reb >iand1t0-265

Reference:

H.J. Miller letter dated June 24,1991 transmitting NRC Inspection Report 50 254/91011,50-265/91007 This letter responds to the EDSFI conducted at Quad Cities Station.

Attachment A to this letter provides Commony calth Ec ison Company's (CECO) response to the Notice of Violation. Attachment B contains CECO's response to the Notice of Deviation. The referenced letter discusses a number of unresolved and open items, and requests each of these issues be addressed along with the reply to the violations and deviation. Attachment C contains the requested responses.

Attachment D discusses CECO's plan to address the overduty concern with the 41W and 250V systems. Finally, Attachment E provides details of CECO's short term corrective actions as well as the long term plans of action to address identified weaknesses such as design calculations and breaker overduty concerns associated with the 4 KV and 250 V systems.

In addition, some items in this response relate to documentation of design issues. CECO plans to meet with Region IIIin the near future to discuss our plans for future control of design documentation, particularly for our older stations.

Currently an EDSFIis being performed at Dresden Station. The concerns identified at Quad Cities will be reviewed in conjunction with those identified at Dresden Station. If evaluation of common issues leads to a new resolution or different plan of action than that stated in the enclosed response, CECO, as necessary, will prmide the NRC with a supplementary response.

If your staff has any questions or comments regarding this letter, please refer them to Rita Radtke, Compliance Engineer at (708) 515-7284.

e

/

(( -se

.J ovach Nuclear Li ensing Manager cc: A. Bert Davis, Regional Administrator - Region III T. Taylor, Senior Resident Inspector '6i L. Olshan, Project Manager - NRR ,

o , 910731003/_ 910724

'A

/

6;; O PDR O

ADOW Or,0003/i PDR i

[ l

s l

l NI'TACHMENT A RESPONSE TO NOTICE OF VIDIATION 254/91011 08 254/91011-13 l

reg /1086:1

s ATTACHMENT A RESPONSE TO NOTICE OF VIOLATION NRC INSPECTION REPORT 50-254/91011;265/91007 ViolatiertAl2&4/91011-08)

During an NRC inspection conducted on April 1 through May d 10,1991, violations of NRC requirements were identified. In accordance with the General Statement of Policy and Procedure for NRC Enforcement Actions",10 CFR Part 2, Appenc'.ix C (1991), violation A is listed below:

10 CFR 50, Appendix B, Criterion XVI, requires that measures be established to assure that conditions adverse to quality are promptly identified and corrected.

Contrary to the above, since May 1980 the licensee failed to implement adequate corrective action to replace General Electric (GE) type CP120A relays which were failing as a result of higher than rated voltage applied to the relay coils.

This is a Severity Level IV violation (Supplement I).

Discusion In 1980, the station initiated an actior. item record (AIR) No. 4-8014 to investigate the failure of several CR120A relays. In October 1981, the Nuclear Engmeering Department completed its review of the subject AIR and recommended replacement of the CR120A relays with GE CR120B relays. Modification M-4-1(2)-85-017 was initiated to implement the replacement. The scope of this modification included a complete rewiring and possible panel redesign due to the difference betwe<m the two t, pes of relays. In 1987, this modification was put on hold due to lack of funds. In September 1988 a review of the CR120A relay failure history was l conducted, this review concluded that during the previous 3 years the failure rate

! for the safety related relays was approximately 1.6% Based upon this evaluation, a revised corrective action plan was developed to replace the coils rated for 115VAC with coils rated for 120VAC. The priority of these corrective actions, however, was not considered to be particularly high based upon the observed failure rate. This action plan was approved by Corporate Engineering in February 1989.

Modification M-4-1(2)-85-017 was cancelled at this time and was replaced by work requests Q74173 - Q74176. These work requests were for coil replacement of CR120A relays in panels 901(2)-40 and 901(2)-41. The initial schedule called for the replacements to be accomplished during the cycle 10 refueling outages (Q1R10 and Q2R10).

"'he Unit One work scheduled for Q1R10 was deferred due to manpower and schedule constraints. The Unit Two work was accomplished as scheduled during l Q2R10. Five of forty-five coils were replaced during Q1R11 and the remaining l forty coils were deferred again due to schedule constreints.

1

' reg /1086:2

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

4 ATTACHMENT A (continued) l In April 1991, Unit One entered an outage to repair the reserve auxiliary transformer. During this outage, fourteen additional coil re 31acements were accomplished. The remaining twenty si) coils are schedulec to be replaced during the next Unit One refueling outage Q1R i2.

Actions _Taken_To_ Correct The Deficiency As stated in the discussion nineteen of the forty five relay coils have been replaced leaving twenty-six remaining relay coils which will be replaced during the next Unit I refuel out-age or sooner iflinit I would enter an outage of sufficient duration. All Unit 2 relay coils have been replaced.

CorrectivelctionalakerLTo_1?rerenLEurtherlioneompliance All corrective actions to identified problems of this type are now entered into the Nuclear Tracking System in accordance with QCAP 1780-G. Completion dates are established accordmg to available resources and a risk based matrix which was implemented in January,1991, This matrix provides for an evaluation of the probability of occurrence and consequences of an occurrence should a corrective action not be completed. This methodology allows for determination of relative priorities, and completion dates are now assigned accordingly. Progress toward item completion is monitored by the Regulatory Assurance Department and unsatisfactory progress is now identified to station management.

An action plan will be developed by September 30,1991 to outline the steps required to evaluate all items in the Nuclear Trackiag System that were not prioritized according to the new methodology.

For outage related activities, Work Planning and Regulatory Assurance coordinate I

preparation of the outage schedule to ensure that all action items and commitments for the outage are incorporated into the schedule. Regulatory Assurance monitors the outage work which is being considered for deferral to identify any action items or commitments, and notifies station management so that l a decision may be made with regard to the acceptability of deferring the work.

DatdhenhlLComplianceMillbeAchieve Full compliance should be achieved prior to the end of the next Unit I refuel outage

. which is currently scheduled to begin September,1992, l

reg /1086:3

\

ATTACHMENT A (continued)

Violation B1254/91311-13)

During an NRC inspection conducted on April 1 through May 10,1991, violations of NRC requirements wine identified. In accordance with the General Statement of Policy and Procedure for NRC Enforcement Actions," 10 CFR Part 2, Appendix C (1991), violation B is listed below:

B.10 CFR 50.59 requires licensees that make changes to their facility without prior Commission approval to maintain a written safety evaluation which provides the bases for the determination that the change does not involve an unreviewed safety question.

Contrary to the above, prior to May 10,1991, the licensee made a change to the facility by connecting the integrated leak rate testing (ILRT) motor driven air compressors to Unit 2 safety bus No. 23, without providing a written safety evaluation for ILRT testing of Unit I containment with Unit 2 at 1007c power.

This is a Severity Level IV violation (Supplement 1).

Discussion In 1975, modification M-4-75-71 was completed in order to provide a power supply from Unit 2 4kV Bus 23 for two 600 horsepower motor driven air compressors utilized for integrated leak rate testing (ILRT). The safety evaluation conducted at this time covered ILRT testing of Unit 2 containment while in refueling. During the next Unit I containment ILRT test the same 4kV electrical supply was utilized to power the motor driven air compressors. No safety evaluation ciocumentation exists for this configuration with Unit 2 at power. This configuration was incorporated into plant procedures and became accepted plant practice.

Actionslaken Toforrect_The Deficiency During the inslection Quad Cities Station Management committed to discontinue the use ofILTC motor driven compressors which were connected to the 4kV Bus 23 until an adequate technical and safety evaluation was performed. The station plans to utilize temporary diesel driven compressors for future ILRT tests. These diesel driven compressors operate independently from the stations 4kV electrical system.

The feed breaker on Bus 23, which was previously utilized, has been taken out of service (OOS) and a note added to the OOS sheet indicating that this breaker is not to be utilized without performing an adequate technical and safety evaluation.

l reg /1086:4

q ATTACHMENT A (continued)

Existing station procedures relating to the set up and performance of the ILRT test are being revised to delete all reference to connecting and using air compressors connected to Bus 23. Procedure revisions will be completed prior to the next ILRT which is scheduled for the upcoming Unit 2 outage.

CorrectimActiumLTaken Toltryenthtther Noncompliance In 1991, Commonwealth Edison initiated significant enhancements to the performance of 10 CFR 50.59 evaluations. The enhancements were based on the

,rinciples contained in NSAC 125, " Guidelines for 10 CFR 50.59 Safety 3 valuations." The new process expands the evaluations by assuring that the reviewer documents the FSAR research and reviews specific criteria in determining whether an unreviewed safety question exists. The review process, therefore, does not merely state the response for the criteria defined in 10 CFR 50.59 but rather provides a logical evolution for conducting the evaluation. The safety evaluation documents the safety analysis logic, judgments and results. For example, under the current program, the review is required to answer the following:

1. Describe how the change will affect plant operation when the changed systems, structures or components (SSCs) function as intended (i.e., focus on system operation / interactions in the absence of ec uipment failures).

Consider all applicable operating modes. Include a c iscussion of any changed interactions wit h other SSCs.

2. Describe how the change will affect equipment failures. In particular, describe any new failure modes and their impact during all applicable operating modes.

This process provides further assurances that a thorough evaluation is performed and that all operating modes are considered.

In addition to the expanded review criteria, the enhanced evaluation methodology provides the qualification rec (uirements for personnel who prepare and review safety evaluations. The quahfication requirements assure that personnel understand the objectives of the safety evaluation and have the appropriate background to conduct or review the evaluations.

The enhanced program for safety evaluations provides further assurance that a detailed evaluation is performed and that all operating modes will be considered

, during the evaluation. - This methodology will ensure that the evaluation considers l

the appropriate safety provisions, design criteria and specifications pertinent to the existmg design and consideration of the potential effects of the change on the safety of the plant.- Quad Cities Station has implemented this process in procedure QAP 110012 " Conduct of 10 CFR 50.59 Safety Evaluations and Screening".

Rate When.EnlLCompliance Was Achievsd Full compliance was achieved on July 24,1991. The feed breaker was taken out of service on February 18,1991 and the compressor was subsequently disconnected, i The note indicatin g that the breaker is not to be used without performing an L adequate technical and safety evaluation was added on July 2(,1991.

l l

l reg /1086:5

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

ATI%CHMENT B RFEPONSE 'IU NOTICE OF DEVIATION

'254/91011-02a; 265/91007-02a 254/91011-02b; 265/91007-02b t

i-l-

l I

l reg /1086:6

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

, ATTACHMENT 11 RESPONSE TO NOTICE OF DEVIATION NRC Inspection Report 50-254/91011;265/91007 Deviation During a NRC inspection conducted on April 1, throu ;h May 10,1991, a deviation of your Final Safety Analysis Report was identified. En accordance with the

" General Statement of Policy and Procedure for NRC Enforcement Action",10 CFR Part 2, Appendix C (1991), the deviation is listed below:

FSAR Section 8.2.2 states,in part, that all protective circuit breakers are sized according to standard electrical industry practice where maximum current interrupting capabilities of circuit breakers exceed the available line-to-line or three (E) phase short circuit current taking into account the impedance of the generator, transformer and other electrical system components.

Contrary to the above, as of May 10,1991, breakers in the 250Vdc system were sized such that a fault current condition could exceed their maximum eurtent interrupting capability by up to 80%; and breakers in the 4kV system could experience a fault current condition that would exceed their maximum current interrupting capability by up to 9%.

Discussion 41W.EyatemDxerduty As discussed below and in Attachment D, CECO acknowledges the technical concerns underlying this deviation and has taken comprehensive actions to address those concerns. CECO's technical review has concluded that it is in compliance with its FSAR commitment to size breakers according to the then existing standard electricalindustry practice Given the c uality of documentation and uncertainty of the precise practice during that period, however, considerable c! Tort and resources would need to be expended to confirm this moint with certainty. In light of our plans to correct the concern, CECO current:y does not alan to pursue this confirmation. Moreover, based on the wordmg of the c eviation CECO is concerned that the NRC may be interpreting the FSAR commitment to be a commitment to current, rather than 1967, standard industry practice. In light of the above, the following additional information is supplied for NRC consideration.

Prior to the EDSFI Inspection, CECO self-identified concerns rogardin ; potential overduty and the quality of the calculations su aporting the de ign of the electrical distribution system. The concerns were that a .though the original calculations appeared to set forth reasonable and accurate results, the quality and detail of the calculations did not meet current industry practice. CECO discussed this issue with the EDSFI team, together with the identified overduty issue, and its plans for addressing these issues, meluding short and long term corrective actions.

reg /1086:7

)

ATTACHMENT B (continued) 1 CECO has reviewed the deviation and it is unclear whether the deviation is premised on a l'mding that the circuit breakers at issue were not sized according to

- 1967 electrical industry practice or to cuirent practice. CECO's interpretation of the FSAR commitment is that the existing configuration must conform to industry practice as it existed in 1967. As a result, subject to the above qualification, Quail Cities meets its licensing design basis, Specifically, CECO has reviewed Quad Cities' FSAR and the original GE design calculations. The plant licensing basis and design basis for the 4KV buses require that the equipment be capable of interrupting the worse case 3 phase short circuit current under both the normal operating and the design basis event (LOCA + LOOP) conditions. We have also concluded that non-normal or transient (short time duration) events were not anal; sed nor do we believe were required to be analyzed under industry practice as it existed in 1967.

Although CECO has determined it meets its 1967 FSAR commitment, CECO nonetheless recognizes and shares the NRC's concern with respect to the overduty issue as it relates to current electrical standards and practices, and has undertaken comprehensive actions to address the matter.

250Y_Systentosenluty A potential 250V short circuit concern was also brought to CECO's attention in 1990 as a result of auxiliary power system studies performed by S&L to address Generic Letter 88-15. Based on these studies CECO performed an operability assessment and a failure mode effects analysis on the 250V system.

The impact of the overduty concern is currently being evaluated for the 250V system with respect to meeting 1967 FSAR design basis criteria. This evaluation will be complete August 8,1991.

In conclusion, Attachment D discusses CECO's plans for addressing the identified 4KV and 250V overduty issues.

l l

l l

reg /1086:8 i

ATTACHMENT C RESPONSE 'IV OPEN AND UNRESOLVED ITEMS 254.91011-01a; 265/91007-01a 254/91001-01b; 265/91007-01b 254/91011-03;265/91007-03 254/91011-04;265/91007-04 254/91011 05;265/91007-05 254/91011-06;265/91007-06 254/91011-07a; 265/91007-07a 254/91011-07b; 265/91007-07b 254/91011-10;265/91007-10 254/91011-11;265/91007-11 reg /1086:9

UNRESOINED ITEM 254/91011-01a; 265/91007-01a "The team was concerned that the degraded voltage relays were set too low. The plant documentation reviewed was unclear as to the switchyard voltage value that was used to establish the degraded voltage relay setpoints. An April 24,1981, submittal to the NitC established the switchyard critical voltage as being 328hV The team calculated that the actual degraded voltage relay settings were based on a critical voltage of 325kV in lieu of the required 328kV. However, the team could not make a clear determination on this matter since the actual calculations to support the voltage studies in the submittal were not available for review during the inspection. Pending further anelysis by the licensee and subsequent NRC review, this item is unresolved."

RESPONSE

CECO established the existing degraded volta ze setpoints at all nuclear plants to meet tha following criteria,1) proviie adequate operating voltage at the terminals of the most limiting Class 1E circuit device and

2) to provide sufficient time for an operator to take action to address the degraded voltage condition without prematurely disconnecting off site power. These setpoints were formahzed and accepted into each stations' Technical Specifications. As discussed in Quad Cities USFAR section 8, the most conservative critical voltage was determined to be the minimum voltage to prevent malfunction of the 400 volt motors or 414 volts. All contactore or relays will operate at this voltage. Their minimum gick-up voltage is 408 volts with a 480 volt nameplate voltage. The critical control voltages (102 volts and 414 volts) related back to the 345kV bus voltage level assuming the worst case auxiliary power distribution system loading is 315kV Hence the degraded volta ge setpoint of 325kV (as calculated by the NRC above)is conservative with regard to actual equipment performance capabilities.

However, in light of the concerns raised during this ins pection and those related in NRC Information Notice 91-29, Deficiencies I dentified During Electrical Distribution System Functional Inspections, CECO plans to revisit the adequacy ofits degraded voltage setpoints and supporting documentation at all nuclear sites. It will be assumed that the voltage on the buses being sensed by these relays can remain at a leveljust above the relay setpoint, Wherever necessary, formal calculations will be put in place to establish system voltage values. These calculations will consider the most limiting Class 1E loads. An engineering estimate is presently being prepared for performing this analysis for Quad Cities down to the 480 volt level. It is expected that a preliminary report and development of an action plan for resolving any setpoint or documentation deficiencies will be completed by September 30,1991.

reg /1086:10

UNRESOLVED ITEM 254/91011-01b; 265/91007-01b "The team noted, during the system voltage review, that voltage

" studies" were being used as supporting documentation instead of formal calculations. The " studies" were not com plete calculations that could be evaluated for technical adequacy. The following elements were not addressed:

(1) The basis for expected / assumed voltage values.  !

(2) Acceptance criteria and the basis for this criteria.

(3) References for field data such as transformer tap settings and system loads.

(4) Worst case loading conditions.

(5) Calculation details including input data, methodology, and results.

(6) Minimum allowable voltage at equipment terminals. l Pending further analysis by the licensee and subsequent NRC review, this item is unresolved."

RESPONSE

Two separate calculations addressing system voltages were provided te

~ the inspection team for review. The first calculata a,.DBD 08-274

-(performed on November 1,1979) for worst case motor terminal voltage was created in response to an incident described in IE Information Notice 79 04. The second calculation, Electrical Auxiliarylystem (DBD

, 08.054),' evaluating the performance f .he electrical distribution system, i was performed on January 31,1990. I sis calculation urovided in put to study report SL-4515. SL-4515 was commissi aed to c etermine the performance of the electrical distribution system including the fault duty

. of the 4160 volt switchgear, motor starting volta ge and motor running voltage down to the 480 volt switchgear level. The study report transmitted the results of calculation DDD 08.054, which was performed, reviewed and approved under Sargent and Lundy's quality -

assurance program.

Commonwealth Edison agrees that neither of the aforementioned calculations " stand alone . In both cases, additional data was provided

~ to the inspection team to supplement ' he information found in the calculations. The general weaknessu identified with many of the calculations reviewed by the team are being addressed separately in the E response to this inspection (Attachment E). The following discussion is intended to clarify each of the elements which were identified as "not addressed",

reg /1086:12

-. .= - - - -- . . - . - . - . . -

a 4

Responce for item 1 "The basis for expected /assumul voltage values."

The voltage levels provided in DBD 08.274 and DBD 08.054 were determined by calculating the voltage drop through auxiliary power system impedance under expected maximum plant loading conditions given the lowest estimated switchyard voltage. The inspection team questioned why diiTerent values of switchyard voltage were used in the two analyses. The value used for the lowest estimated sw tch  :

voltage is determined by the Commonwealth Edison System lanning I;vard Department using a mainframe based computer pr9 gram to determine system power flow. This system power flow analysts is performed annually to determine the afTect ofload increases and changes to the transmission system, such as new lines or generating units. For nuclear i stations, the calculation of the minimum switchyard voltage assumes the outage of both units and the next worst single contingency (the outage of a transmission line, transformer or generating unit) with maximum transmission system loading. Therefore, as Commoawealth Edison's transmission system changed between 1979 and 1990, different minimum expected switchyard voltages were used as input to DBD 08.274 and DBD 08.054. i Response for item 2 Acceptance criteria and the basis for this criteria.

The acceptance criteria for the voltage range at the motor terminals is based on equipment ratings as defined by the National Electric Manufacturers Association (NEMA). NIBMA standards require that the maximum rated motor voltage should he limited te 110% of equipment rated voltage and that the minimum voltage for a running motor be at least 901 The motors connected to the 4160 Volt buses are rated at 4000 Volts; therefore, the voltage range should be 4400 Volts maximum, 3600 Volta minimum. The motors connected to the 480 Volt switchgear and motor control centers are rated at 460 Volts for un acceptable range of 506 Volta maximum,414 Volta minimum. Note that the voltage calculated in DBD 08.274 and DBD 08.054 is given in terms of either a per unit (P.U.) value or as a percentage of the equipment rating as well as the absolute value for a quick comparison of the available voltage to the lower limit of 90%

Response to item 3 Refemnces for field data such as transformer tap settings and system loads.

References for field data used in calculation DBD 08.054 were transmitted by means of Design in aut Transmittals (DITs). DITs utilize filed information (verified by walkc owns) for such items as transformer tap settings and motor nameplate data. Tim BHP requirements for a pu up motor are normally obtained from another calculation. If BHP is not known, rated (nameplate) horsepower for the mump motor is assumed. Maximum system loads are determinec by performing a system overview to determine which motors or other loads are running for a given plant condition. Copies of the DITs referenced in DBD 08.054 were provic ed to the inspection team.

Hesponse to item 4 Worst case loading conditions.

Worst case system loading is determined utilizing a systems approach to determine which loads are running under specific plant conditions reg /108613

and to determine when a given motor may be required to start. This methodology applies to both DBD 08.274 and DBD 08.054. For DBD 08.054, eight conditions were analyzed to determine which conditions result in interrupting duty above the breaker rating. DBD 08.054 provides tabulations of the running loads for each analyzed condition and concludes that running and starting voltage values were acceptable for all conditions.

Response to item 5 Calculation details including input data, methodology, and results.

Calculation details on DBD 08.054 were discussed with the team in response to information request N 161. As indicated in the response to item 3, input data was referenced in the form of DITs. References were provided under request N-192 which addressed calculation methodology as well as a specific question ecacerning the fault contribution from the l recire MG set. Additional questions on methodology were responded to i under request T-218. These questions concerned the efTect ofload cable impedance, the references for total X/R, cable temperature, Bus 14-1 fault duty and the ratings of the GE 250 MVA gear. Conclusions are '

reached in the calculation for voltage and fault duty on pages 19,26,33, 40,47,55,63 and 71 of the calculation. Specific recommendations were provided to resolve the fault duty issues on pages 80 through 87, Response to item 6 Minimum allowable voltage at equipment terminals.

As indicated in the Response to item 2, the criteria for the acceptable voltage iange at the motor terminals is based on equipment ratin gs as defined by the National Electric Manufacturers Association (NEEA).

In addition, to provide adequate torque for motor starting and to prevent contactors from dropping out at the 480V motor control centers, minimum voltage under starting conditions must be limited to 75% of motor rated voltage. This issue is discussed in DBD 08.274.

l l

l 1

l

[

reg /1086:14

UNRESOINED ITEM 254/91011-03;265/91007-03 "The team identified a number of weaknesses with calculation No.

7318-33-19-3, dated October 5,1990,in regard to the capability of the DG to meet load sequencing and total load capacit Weaknesses identified in the calculation included:y requirements.

Lack of dynamic performance analysis to demonstrate adequate DG 3erformance during the entire range ofload sequencing conc itions. The calculation provided a voltage dip analysis based on a dead Ic,ad pickup curve supplied by the vendor. The curve only defines the voltage performance of the DG for a period of one second followin g the instant a sequenced load is applied. However, the cumu ative and dynamic load effects on the DG output voltage will continue beyond one second as other loads are applied to the DG,

. The calculation did not include a frequency analysis.

The calculation attempted to demonstrate the ability of the DG to start large loads usmg the dead load pickup curve, This demonstration however, was not listed as an objective of the calculation and acceptance criteria for this objective were not explicitly defined.

- The calculation did not include an analysis for atarting the service water pump (SWP). In response to this concern the licensee provided a draft calculation demonstrating the 6ility of the DG to start the SWP based on the dead load rinup curve. The draft calculation did not include an aralysis of the effect of the SWP start on existing loads.

- The calculation indicated that certain loads could drop out due to voltage dips during the loading sequence. However, there was no attempt to quantify the power re-quired to restart lost loads or to assess the effect that this would have on DG performance.

- The calculation did not compare the static load profile to the DG's output capacity. ,

It was unclear how a designated design reviewer could determine calculation acceptability since key acceptance criteria of the calculation and DG loading capability criteria were not stated in the body of the calculation and could not be found in the references stated in the calculation. Pending further analysis by the licensee and subsequent NRC review, this item is unresolv'ed."

l l

reg /1086:15

. RESPONSE The following revisions and additions will be completed in order to clarify the weaknesses in the Quad Cities DG loading calculations 7318 33-19 3(-1,-2).

Response to item 1: Lack of dynamic performance analysis to demonstrate adequate IX1 performance durin g the entim range ofload sequencing conditions. The calculation provided a voltage dip analysis based on a dead load pickup curve supplied by the 4 vendor. 'lhe curve only de6nes t.he voltage performance of the DG for a period of one second following the instant a sequenced load is applied.

However, the cumulative and dynamic load effects ~

on the DG output voltage will continue beyond one accond as other loads are applied to the DG, An improved description of the capability of the DG to start motors, including limits, acceptance criteria and documented references will be contained in the revised calcstation. A more complete discussion of voltage recovery above 80% ano the effect of reduced voltage on protective devices will also be included.

Response to item 2. The calculation did not include a frequency analysis.

Frecuency response of the diesel generator under motor starting cond itions is a function of the governor response. The original specification for the diesel generator did not address frequency response. Instead, the abihty to start motors based on voltage recovery was specified, Industry codes and standards in affect during the original design specified voltage recovery as a basis for successful starting.

Frequency response later became an issue under Regulatory Guide 1.9.

Data for such an analysis is not available for the equipment installed at Quad Cities Station. Motor starting capability is demonstrated via a Technical Specifications surveillance (4.9.E.2) where a loss of offsite power is simulated and the ECCS pumps are sequenced on to the diesel generator. Although the pum p discharge is in nummum flow for this

- test, starting during an actual design basis event would he under ec hr conditions. ' The discharge valves receive an open signal when reactor depressurization to the mjection pressure setpoint occurs. Full flow to the vessel occurs after all ECCS pumps have started on minimum flow, and the discharge valves stroke open. Therefore, the periodic test adequately demonstrates the abihty of the diesel generator to start ECCS motors.

- Since Quad Cities Station is not committed to Regulatory Guide 1.9, and since the Technical S 3ecification surveillance testing adequately demonstrates the abi;ity of the Emergency Diesel Generator to start the ECCS motors, a frequency response analysis during motor starting will not ne included in the revised calculation.

L l

i reg /1086:16 l

p Hespc nne to item 3. The calculation attempted to denwmdrate the ability of the DG to start large loads using the dead -

load pickup curve. 'Ihis demondration however, was not listed as an objective of the calculation and acwptance criteria for this objective were not emphcitly defined.

A more detailed comparison of calculation results to acceptance criteria for both loading and voltage recovery plus justification for the acceptance criteria will be included in the revised calculation, Justification of the assumption of constr; 'WA load during motor starting will be included. I Hesponse to item 4. 'Ihe calculation did not include an analysis for l starting the service water pump (SWP). In response to this concern the licensee provided a draft calculation denenstrating the ability of the  !

DG to start the SWP based on the dead load pickup l r.urve. The draft calculation did not include an analysis of the effect of the SWP dart on exiding loads.

Starting of a service water pump will be included in the analysis.

Response to item 5. The calculation indicated that certain loads could dmp out due to voltage dips during the loading sequence. However, there was no attempt to quantify the power required to restart lost loads or to assess the effect that this would have on DG periormance.

The revised calculation will include an expanded discussion of dropped out loads due to voltage dips during the loading sequence.

Response to item 6. The calculation did not compare the stat.ic load profile to the IXi's output capacity.

A clarification ofload table reference for which the DG is certified to sturt plus a correlation ofload profile values to Electrical Load Monitoring System (ELMS) 3rintouts will be included in the revised calculation. There will also ao increased attention in the revised calculation to the subject of powering cross tie loads.

The anticipated completion date for the revised diesel generat. r loading calculation is the end of the first quarter of1992.

Regarding the team's concern on how a designated design reviev er could determine calculation acceptability since key acceptance criteria of the calculation and DG loading capability criteria were not stated in the body of the calculation and could not be found in the references, the issues of calculation quality are discussed elsewhere in the response to this inspection report (Attachment E). -

reg /1086:17

1 4

OPENITEM  ;

254/91011-04;265/91007-04 "The team noted that in order to meet Technical Specification requirements for Diesel Generator testing, the Diesel Generators are paralleled with the offsite power system. The team reviewed the generator protection scheme and ouserved that during Diesel Generator

-(DG) parallel operation, the DGs could be overloaded. In the event of a LOCA followed by a loss of offsito power (LOOP), the degraded voltage relays on the associated vital bus would not sense loss of voltage from the offsite power source and load shed (trip) the necessary circuit breakers. As a result, the DG would be subjected to a sudden overload as the DG attempted to pickup the loads arevioasty supplied by offsite

>ower. This condition could damage the DG or trip the DG output areaker. The team recognized the low probability of a grid collapse during monthly DG testmg. However, the team considered this condition to be a generic design weakness. Pending further NRC review.

this item is considered open.'

RESPONSE

On October 15,1990, Quad Cities Unit 2 tripped on a moisture separator high level signal. This event is described in LER (90-010). Power was transferred from the Unit Auxiliary Transformer to the Reserve Auxiliary Transformer as expected following the turbine tri ). Several unexpected actuations occurred during the event; of princip' e interest to the inspection team was the Unit Two Diesel Generator Autostart.

Quad Cities Station Technical Staffinitiated an operability review of the Unit 2 diesel generator prior to restart. This review concluded that the diesel generator was operable.

A study was initiated to review the logic, determine the cause of the DG Autostart and recommend alternatives to prevent a recurrence. This study is scheduled for completion by September 30,1991.

During an initial phase of the study (which reviewed the existing logic),

the exposure of the DG to damage during testing with a concurrent LOOP /LOCA was identified. A Probabihatic Risk Analysis (PRA) was initiated utilizing Dresden Station to model the significance of this exposure. The PRgconcluded that the probability of this event is on the order of 1.67 X 10- per year. The expected probability is approximately the same for Quad Cities Station. The probability of this event, therefore,is a negligible contributor to core damage frequency.

Commonwealth Edison is including this improbable event as a preliminary design requirement for any proposed modification to the autostart logic. The actual inclusion of this design requirement will be based on economic considerations and the results of a technical review of the impact of this logic change to damage frequency of the DG.

reg /1086:18

The autostart logic alternatives reviewed to date have all introduced new concerns to operation under design basis events and have been ,

rejected.

In accordance with ENC Procedure QE 00," Design Modifications" the final study results will be presented to the Station Modification Iteview Cornmittee (SMllC). The SMllC will then deterruine which logic alternative, if any, will be initiated as a modification to correct the unexpected autostart actuation.

l l

l l

l' l

l l

l I

reg /1086d0

1 UNRESOINED ITEMS 25W91011-05: 265/91007-05 "The team wat, unable to determine whether cables were properly sized to provide suflicient ampacity. The licenseo stated that cable sizmg was  ;

established using various architect engineer (AE) standards. The AE '

standards were based on industry standards; however, the particular industry standards used to develop the AE stundards were not identified. As a result, the team could not effectively evaluate cable sizing and cable fill requirement. Pending further analysis by"the ,

licensee and subsequent NRC review, this item is 'nresolved.

Hesponse From the original design period through the present, Sargent and Lundy (S&L) has used various revisions ofite * .d ards to conservatively size cables. These S&L Standards were s.: , tinue to be based on accepted industry practices and documents in efTect at the time of tlieir isst.c.

Due to the level of documentation typically required during the design of early vintage plants, specific reference to the blandard and Revision was not mode. De ermining the Standard and Revision applicable in the specific time frame requested oy the team was not possible. Therefore, to address this concern, the Sargent and Lundy Interactive Cable Engineering (SLICE) program feature which assesses cable ampacity wil be utilized to confirm that currently installed power cah' s are ade.quntely sized to carry their load current. The initJa! run of SLICE database used a library with default values for cable ampacity. The default values used are the madmum allowable values for a given cable size as a conservative approach. As such, the SLICE program using these very conservative values, indicated that some routing points were thermally overloaded. A more thorough analysis of the overloaded routing pointa using actual runniniload currents, as opposed to a maximum reted load current per the S&L library, is bemg done and it is anticipated to be completed by August 31,199L The SLICE progam cable ampacities are based on the STOLPE method of calculating the cable amkincity. In this method an allowable heat intensit(y per unit cross sec tonal area of the tray is determined solving Jm non linear equations for the conductmn, convection and by radiation heat transfer associated with the cable. Each cable is allotted an amount of heat based on its cross sectional area and the allowable heat intensity for the cable tray depth of fil' at the aarticular routing

,oint under consideration. The allowed amount of beat is used to find the ampacity of the cable from the conductor resistance at rated eperating temperature.

In the SLICE pro gram, the depth of fill for a routing point is calculated i

from the input ca ile diameters. A piece wise linear approximation is used to translate the results of the solution of the non linear heat transfer equations into a curve of allowable heat intensity versus cable tray depth of fill, i reg /1086:20

The SLICE program uses a refinement of the STOLPE method in order to calculate the power cable ampacity for trays containing mixed lower and control cables. The control cables are modeled as a separate ayer of cables lying on top of the layer of power cables. The non linear equations for this model are translated into another piece wise linear approximation which is a function both of the power cable depth of fill and the control cable depth of fill.

l The SLICE program was developed under the control of S&l/s Quality Assurance (QA) program and continues to be maintained in accordance -

with S&l/s current software verification and validation (V&V) procedures.

S&I/s soRware QA program meets the requirements of10 CFR 60, Appendix B and 10 Cim, Part 21. It addresses the sonware development process,and configuration control and errorby teS&I

. All aorting?s General soRware development implementation is governed Office Procedure 41 Rev. 4," Software Development & Maintenance Requirements" and supporting Computer Services Di+1on procedures.

The software development process incitides doctimentation requircinents for each phase of development including a Software requirements Specification (SRS), a Sonware Design Description (SDD), a Software Test Plan and a Sonware Verification and Validation Report (SVVH).

Each document is independentl defines what is to be developed,y reviewed and approved. The SRSadd as in put and output, hardware, and performance. The SDD describes how the program will be orgam.:ed to meet the requirements addressing the planned sonware structure, program module functions, applicable logic and applicable data base structure. The test plan defines test se's and applicable procedures necessary to test the specified requirements.

The SVVR presents the results of all tests and discusses any deviations from expected results.

SoRware used by S&L for client projects ato licensed for external use is entered into a configuration control system. Required software development documentation, user documentation, program source code in both electronic and hard copy format, and execu ;able code are submitted to our computer sonware library (CSL) for inclusion in the system. When production software requirements are met, CSL installs or distributes the executable code for production use.

The S&L soRware QA procedures also include requirements for error reporting. Software errors are classified based on the potentialimpact with respect to successful program usage and conseguences of error.

Ikrors that could potentia ly produce non conservative results for design related activities are formally reported with an indication of potential corrective action for applicationn of the program version that contained the error. Errors that do not inhibit successful use are corrected in planned subsequent versions.

Current S&L V&V procedures apply to all new software and modifications to existing software, l

reg /1086:21

OPENITEM 254/91011 06;265/91007-06 "The tenin observed that fuse Nos. 6 and 10, installed in primary untainment isolation (PCI) panel No. 90140, were rated at GA and 5A respectively,instead of f>A and 1 A specified in design drawin 4 No.

4E 1764A. Also, fuse Nos 6,9, and 13 installed in PCI pane: 00141 were rated at 4A,5A, and GA res iectively,instead of f>A,1 A, and 5A s pecified on design drawing No. LE 1764h. In addition, the team noted that neveral fuse types specified on the desi gn documents did not conform with the l'une types installed in PCL panel 90141. The incorrectly installed fuses did not affect equipment operability.

The licensee had previously identified fuso installation discrepancies and has initiated a fuse walkdown to identify similar discrepancies. Also, a fuse list and a procedure to control and maintain fuses were being develepud. Pending NHC review of the licensee's actions to replace the inemrect fuses and establish an effective fuso control program, this item is considered open."

lllCSPONSE in 1990 Commonwealth Edison initiated an efTort to upgrade the fuse control program at all ofits Nuclear Stations. Ily the end of 1990 CECO had developed a controlled fune list and distributed the list to Quad Cities Station.

A walkdown program is currently on going to ensure the necurney of the i fuse list. The Site Engineering group, which is performing tho l walkdown, has completed approximately 75% of the Unit I list and 30%

of the Unit 2 list. The review of the fuse list is expected to be complete in the third quarter of1992.

Fuso discrepancies that were identified by the inspection team have been evaluated and found to be adequate to protect the circuit. A l

recommendation has been made, however, to replace the installed fuses with fuses of the correct (design) size and type.

l The Quad Cities procedure QAP 300-40 " Control and Maintenance of i

Fuses and the Fuse List Procedure," has been issued. This procedure l

defines methods utilized for evaluating and replacing blown or discrepant fuses.

To incorporate lessons learned from this inspection, the Nuclear Engineering Department is working closely with t he Station to impr ove the effectiveness of the ongoing fuse control program. The following procedural enhancements will be made to QAP 300 40 to ensure timely resolution ofidentified fuse discrepancies.

A. Station walkdown teams will process fuse discrepancies in a timeiy manner. A technical evaluation of an identified discrepancy should be r mpleted in ap?roximately 30 days.

Any deviation from this timeframe wil . require upper management approval, reg /1086:22

P

11. When a discrepant. fuso evahuition finds the installed fuse acceptable but recoinmends the fuse be replaced, replacement shouM be com sleted as soon as practical based on plant i operating coniitions. Fuse teplacements shall be completed orior to restart from the next refuel outage. Any deviation u Erom this tinneframe will require upper nuutagentient approval.

C. l'rocedural changes will also be made to better define the interfaces necessary to efketively coordinate fuso ebnnge out.s during operation.

These procedural changes will be completed by September 30,1991.

reg /1086:23

. AM(. .

hj OPENITEM 254/9101107a; 265/91007 07a i

i

~)

"The teant identified the following concerns during the as built drawing review:

. Cable Nos.15153 and 18125 were shown on currer.t elect rical drawings even though these cables had been reinoved froin the field during ituplementation of past modifications.

. Installed circuit breaker ratings diWered from design drawings in 480 Vac MCC No.191 cubicles H1, H5 and E3; 480 Vac MCC No.18 2 cubicles F1, F2, F3, F4, G2, G3, and G5; 480 Vac MCC No.19 2 cubicles E1, E2, E3, E4, F2, F3, F4, and F5; and 120/240 Vac ESS distribution panel and instrument bus panel No. 90149 circuit Nos. 32 and 33.

. 120/240 Vac Ess distribution panel and instrument bus panel No. 901-40 circuit breaker Nos. 27,29, and 31 were mislabeled as Nos. 25,27,29, respectively.

. HPCI valve, MO 23019, limit switch contact (7-70) was shown connected oh schematic diagram No. 4E 1529, Sheet 3, and wiring diar;lled in the field This contact was also shown on the'am No. 4 never insta Control lloom Critical Drawing which reflects the as built plant condition.

. Motor power leads Al and A2 were shown reversed on the drawings for 250Vdc MCC No. \A cubicles F02, G01, H02, and 101.

. Critical control room drawing No. 4E 2318B did not show the Unit 2125Vdc auxiliary battery and the Unit 1 125Vdc auxiliary battery was not labeled.

Pending further analysis by the licensee and subsequent NRC review, this item is considered open.

RESPONSE

The above concerns identified by the inspection team have been assigned to the ap,plicable system engineer for review and evaluation.

Appropriate design document revisions or Work Requests will be initiated to address the concern if required. All items will be entered on the Quad Cities Nuclear Tracking System (NTS). Items 1-4 are currently under review and all disere aancies will be resolved. A Drawin

• Change Request (DCR) has >een submitted to correct concern

5. The Init 1125Vdc auxiliary has been labeled, and additional information has been added to the Critical Control Room Drawing identified in item G to address the inspection team's concern. Corrective actions to address the overallissue of configuration management are discussed in Open item 254/01011-07b; 205/91007-07b.

reg /1086:24

OPENITEM 254/91011-07b; 26fd91007-07b "The team identified the following concerns during the field inspection:

- The licensee had revised the schematic and wiring diagrams to inclade proposed plant modifications that had not been installed in the field (referred to by the licensee as " chaining" methodology). As a result, the existing electrical drawings did not reflect the field installation (as built). The licensee informed the team that prior te performing any activity on the EDS, a field wiring verification was required by procedures.

. The interim document information system (IDIS) data for drawing NO. 4E1764A contained numerous errors. It was difficult to determine the as. built condition / status of the documents posted against the drawing using IDIS. The team was concerned that data errora may exist throughout IDIS.

Pending further analysis by the licenw>e and subsequent NIIC review, this item is considered open."

RESPONSE

The following responses have been developed for this open item.

Response to item 1 The licennec had rwined the schematic and wiring diagrams to include propomxl plant nwdifications that had not been installed in the field (referrxxl to by the licenace as " chaining" methodology). As a result, the existing electrical drawings did mit reikx L the field installation (es built). The licenmm infornux! the team that prior to performing any activity on the EDS, a field wiring verification was required by procedures.

Commonwealth Edison acknowledges the weaknesses associated with the " chaining" methodology associated with modi 6 cations. In order to minimize this weakness,61 " chaining" methodology modifications have been cancelled E Quad Cities Station. In addition, Commonwealth Edison recogniz s the need to remove drawings in which interim (non completed) chaining modifications exist.

Two methods are beinputilized to remove the " chaining" drawings. The first method consists of reviewing how the " chain" affected each revision of a particular drawing. As a par t of the review of each drawing revision, the impact.ot other completed and non completed modtlications on the particular drawing must. be resolved.

reg /1086:25

The second method of removing the affects of the " chaining" modification to a particular drawing is to perform a complete as. built walk-down of the systems reflected on the drawing. Once this as built walk-down is l completed a totally new drawing is created.  !

Each of these methods, as well as a combination of both, are being utilized b,y Quad Cities to remove the affects of the " chaining" modification process on drawings.

Effective January 1,1990 the Engineering Change Notice Methodology was implemented at Quad Cities Station. To assure the as built j condition of the drawings, a Document Change Request will now be generated at the time installation is complete to assure incorporation of the design change onto the original design documents.

The current schedule for completion of the mechanical and electrical portions of the walk down program is the end of1992. The scope of this work activity includes the completion of the Total Job Management (TJM) data collection. As built drawings will he completed in a parallel process as data collection progresses in both the mechanical and electrical areas.

Current resources for the Quad Cities walk down propam include approximately 30 personnel from both Commonwealta Edison and qualified contractors.

Response to item 2 The interim document information system (IDIS) data for drawing NO. 4E1764A contained numemus ermrs. It was difficult to determine the as built, condition / status of the documenta posted against the drawing using IDIS. The team was conwii.ed that data errors may exist thmughout IDIS.

The construction status of the IDIS report has been revised so that a cancelled modification can be more easily determined. To irovide information relating to ECNs, FCRs ami DCRs a new Desi n Document to Design Change Document status is now being utilized. 'o assure more reliable data a duplicate entry error report has been produced.

In addition, approximately 1755 drawing revisions have been reviewed and are curren"ly being back fitted into the IDIS data base to provide a more com archensive status of all drawings. Revised computer reports which include additional and reformatted data have been created to allow easier use of the data base. Finally, trainin g in the use ofIDIS is scheduled for completion by the end ofID91 for all Central File and selected engineering and 'I echnical Staff personnel, reg /1080:26

OPENITEM 254All011-10; 20M)100710 "The team deterinined that no infoimation existed to vetify that the quad Cities Station electrical penetrations met the intent of llegulatoiy Guide 1.03. NitC letter dated November 30,1981 to Conunonwealth Edison Company, indicated that Dresden 2 (a sister plant) was meeting the intent of these requirements. This is an open item pending NitC follow up on the licensee's determination that the Quad Cities Station penetrations are similar to the Dresden 2 penetrations."

ILESPONSE The Quad Cities and Dresden low voltage penetrations are the same (GE type NSO4) for the valve motor operator selected by the NItC for review at Dresden. The protection devices (circuit breakers and overload relays) used, however, differ slightly in type and rating. Also, the enblo length difTers which has a modest elTect on the analysis.

In order to determine if the Quad Cities penetration is sufhclently similar to the Dresden penetration, an analysis of the protection provided to the low voltage penetration conductors by the circuit protective devices has been performed. This analysis utilizes the same methodology used for the Dresden Eystematic Evaluation Program (SEP) topic.

As a part of the SEP in 1981, the NHC seh eted the 480V circuit of Motor Operated Valvo (MOV) MO2-0202-5A for evaluation against the latest design criteria for containment penetration assemblies. Sargent &

Lundy (S&L) analyzed the seh'cted penetration circuit and concluded in its report. of. June 22,1981, that the combination of molded caso circuit brenher and the thermal overload device provide adequate primary protection of the penetration conductors for all values of fault current except for faults within the 200-600A rango. Since the circuit was to ho environmentally qualified, there was no need for providing backup protection for (Im penetration conductors. Tests were perl'ormed by CECO) nuuntenanco qualified personnelConunonwealth to determine theEdison Company nature and capa (bilities of the overload devices in providing primary protection to low voltage penetration conductors for fault currents m the range of 200-600A. Based on the results of these tests, the issue oflow voltage penetration protection was considered to be adequately addressed for Dresden Station.

Utilizing the same methodology developed for the Dresden SEP response, the penetration and protection devices associated with the valve motor o aerator for Quad Cities MOV 202 5A has been analyzed. A summary of t m results of this analysis follows.

reg /1086:27

The penetration conductor size (#10 AWG) and the penetration design (GE type NSO4) for Quad Cities Units 1 and 2 are the same as the Dresden Unit 2 equipment. The maximum steady state drywell environment temperature before LOCA and the temperature time characteristic of the drywell environment following LOCA are essentially the same at both Dresden and Quad Cities. Therefore, the capability curve established for #10 AWG penetration conductors at Dresden Unit 2 is applicable for #10 AWG Penetration Conductors at Quad Cities Units 1 and 2.

The Unit 1 penetration of MOV 202 5A is protected by a GE type TEF 136M040 magnetic only breaker and a GE type CR1240 thermal overload relay utilizing CR1230180B heater element. The TEF breaker is presently set at Lo tcp. The Unit 2 penetration is protected by a GE ty?e TEC 36050 magnetic oni,y breaker; however, the thermal overload re.ay and the heater element is identical to the Unn 1 circuit. The TEC breaker is set at position 5.

The settings and characteristics for these devices were reviewed against the hveakers, overload relays and heaters used at Dresden. The Dresden relay and heater bound the Quad Cities relay and heater for lower currents. At higher currents, the breakers provide adequate protection for the penetration.

The attached figure 1 ws the thermal capability of the #10 AWG penetration conducto: Sin figure also shows the operating characteristic of TEC s her when set at position 6 as used on Quad Cities Unit 2. The bi .t. 3r used on Quad Cities Unit i is bounded by this curve as it,is a lovror rated breaker (40 Amps versus 50) and is set lower (position 1 versu s position 5). The time current characteristic of the Quad Cities overload relay and heater are compared to the Dresden relays and heaters.

It is therefore concluded that the combination of CR1240 relay with CR.'.23C180B heater and the TEF or TEC magnetic only breaker provides adequate thermal protection for the Quad Citics #10 AWG penetration conductors.

l reg /1086:28

Ol'EN ITEM 254/91011-11; 26fd9100711

The team found several deficiencies in design documentation associated with the capacity of the fuel oil system:

No documented fuel consumption tests existed for the Quad Cities DGs.

The fuel consumation rato calculations did not consider reduced fuel availability Lo the DGs as a result of fuel consumption by the diesel driven fire pumps.

. Possible impact of transfer pum a suction limitations on effective available volume of the storngo Lanks was not considered, such as vortex considerations and transfer pump suction head requirements.

The calcultin n did not address the imps ' of storage tank orientation (i.e., taaks are installed with a slope along longitud'nal axis) on the accuracy oflevel instrumentation. .

The calculation did not address differences between fuel consumptior test conditions and actual operating which could affect fuel storage volume requirements.

The team concluded that the design documentation deficiencies did not constituto an immediate operabihty concern since the licensee administratively maintains a much higher level of fuel oil in the fuel oil storage tanks. l'ending further analysis b NitC review, this item is considered open.,y the licensee and subsequen 1 RESPONSM Itesponao to item 1 No documented fuel consumption tenta existed for the Quad Ciuen DGs.

Quad Cities Station will perform EDG fuel consumption tests. These tests are scheduled to be completed by September 30,1991.

Hesponne to item 2 The fuel consum ption rate calculations did not consider reduced fuel availability to One IX1s an a renuit of fuel consumption by t.he dienol driven fire pumpn.

Conunonwealth Edison has contracted to complete a review of the l consumption calculation considering the reduced fuel oil availability due n to the diesel driven fire pumps, This calculation will be revised if necessary by September 30,1991.

Hesponse to item 3 Possible impact of transfer pump suction j limitations on effective availablo volume of the storage tanks was mit considermi, such as vortex considerations and transfer pump suction head requirmnenta.

reg /1080:29

._ m _ . _ _ _ _ _ _ _ - . . _ . . . _ ._ _. _ _ _ . ~ _ . __ __ _ _ _ .___ _._ __.

l l

. l The resolution of the impact of transfer pump suction limitations on the e%ctive available volume of the storage tanks will be included in the review identified above. The calculation will be revised if necessary by September 30,1991.

Response to item 4 The calculation did not addreas the impact of storage tank orientation (i.e., tanks are inntalkwl with accuracy a sk+flevel o anstrumentation,c along longitudinal axis) oj l

ABB 1mpell Calculation No. 0591523 001 Rev. O provided data for i calibration of the level measurement instrumentation, in this '

calculation, which was provided to the inspection team,it was assumed that the tank tilt haa negligible effect on the usable volume of diesel fuel in the subject storage tanks. Therefore, no calculation was performed to show that the amount of unusable volume would be any greater if tank tilt was taken into account.

ABB Impell Corporation, in a letter addressed to T.P. O'Brien (CECO) dated May 2,1991 addresses the impact of storage tank orientation on the accuracy oflevelinstrumentation for the UnLt 2 and 1/2 diesel fuel oil storage tanks. Based on this evaluation, the following conclusions are made:

1. The calculation conservatively assumed that a 1"levelloss would exist around the ruction piping due to a possible vortex.

Based on a review of the velocity of fuel oil in this piping it ap pears that this assumption is conservativo due to .ow fluid velocity. This conservatism would more than account for any effects of tank tilt.

2. Based on a review of vendor drawings, the suction pi aing and level measurement instrumentation may be mountec slightly off-center of the tank. Since the instrumentation used for level indication uses the pressure due to the weight of the fluid above the bottom of the suction pipe as the basis for determining remaining level, and since this instrumentation piping is positioned very near to the suction oiping, it is concluded that the off-center mountings have no efYect on the calibration range of the level instrumentation.

Response to item 5 The calculation did not addrens differencen between fuel consumption test conditions and actual operating which could affect fuel storage volume requirements.

Based on the response to item 1, the fuel consumption tests at Quad Cities are scheduled to be completed by September 30,1991.

Engineering will evaluate the results of the tests and ensure that the actual DG fuel oil consumption is consistent with assumptions made in desi m calculations. This evaluation will be completed by October 31, 1991, reg /1086:30 t

i A'ITAClIMENT U PIAN 'N) ADDRESS TIIE 4 KV AND 250V SYSTEMS OVEllDUTY l reg /1086:31

4 PI AN TO ADDitESS Tile 4KV AND 250V SYSTEMS OVEllDLTI'Y Prior to the Quad Cities EDSFI, Commonwealth Edison identified several issues concerningde and 250 Y- systems.the Dresden/ Quad Cities Auxiliary Power System for both For the 4100 V ne system the issues can be categorized into the following areas:

1. Short-circuit Overduty of Switchgear Various studies identified plant operating conditions where the short circuit rating of certain 41,V,250-MVA switchgeur (Dresden lluses 23,24,33, and 34; Quad Cities Iluses 13,14,23, l and 24) as exceeding current industry codes and standards.

2. Availability of Spare Parts and Qualified Components The 250 MVA switchgear was originnily manufactured by GE.

GE has stated tlint it will no losiger supply view clictiit breakers or spare mrts for this switchgear. Of particular concern is the unavaila'iility of arc chuten.

3 Improved System Performance Activities associated with innroved system performance include provisions for future load adiitions, improved separation of safety related loads and non safety related loads, and the development of an integrated approach to the implementation of the S130 design.

In response to these issues, Conunonwealth Edison has prepared an Auxiliary Power System Enhancement Study. Several alternative solutions have been provided in this study and are currently under review for the best course of action to take, based on resources available, to improve the system.

l reg /1086:32

' ' ' W e'y m ygy ,. ._ __

O Resolution of these issues will require a long term program to assure succesaftil conspletion, hiilestones schedtileu for this progrant liicitide

)

I

. Completion of a detailed cost benefit comparison of alternatives by September 1,1991, ,

. Obtain budget approval by Febniary of 1992, and 13egin detailed design activities and prepare a finalized  :

project scheduto by April of1992.

This schedule has been develo 4ed so that required switchgear replacement / load transfer moiifications can be coordinated with the Stallon Blackout (SBO) modifications.

For the 250 V de system similar issues have been identified. Due to the similarities between the issues, an analogous program for the enhancement of the 250 V de system has been initiated.

The initial remort for the enhancement of the 250 V de system is duc October 31,1991.

b reg /1086:33

._ _ _ _ . _ . _ . . . _ . . _ . . _ _ _ _ _ . - _ .._ __ . . _ . _ _ . _ _ .-. _.m.. . _ _ _ _ . _ _ . .

4 ATI'ACilMKNT E PR(X1 HAMS'IV ADDRESS ELECTRICAL DISTRIHtfrlON SYFTEM CALCUIATIONAL WRAKNESSES t

I-reg /1086:34

e PROGitAMS TO ADDitESS l ELECTitICAI, DISTIlillUTION SYSTEM CALCUIATIONAI, WEAKNESSES The team determined that shott circuit calculation No. 8445 00 EAD 1, dated January 31,1990, contained the following weaknesses:

. Nonconservative methodologies were employed without justification such as not including the effects oflowest load motor impedances on bus fault currents.

Assumptions were not stated or justified, such as the effects of cable temperature on the fault current calculation.

. References for key input data were not provided, such as the name plate data for 250MVA breakers.

Formulas were used without defining the equation variables.

. The available short circuit current was evaluated at the load equipment terminals instead of at the load breaker's output terminals.

The licensee performed supplementary calculations that demonstrated that the above factors would ' at have a major efTect on the overall results. Ho=ver, the assumptions underlying these nonconservatisms should have been clearly stated and justified in the original calculation.

The team considered tlus concern a weakness relative to design documentation."

RESPONSE

While this specific design weakness was identified by the EDSFI team, it actually represents a more general concern regarding the total quality of calculations supporting the design of the electrical distribution system. Prior to the EDbFI Inspection, this concern was identified through the performance of a qualitative review of the design calculations associated with the 4160 V ac,480 V.ac,250 v de, and 125 V-de systems. This review determined if revision or updating was warranted to reflect current plant design and configuration. The inspection team was presented with these results, as well as a chronology of observations regarding weaknesses in the total quality of calculations and several preliminary actions taken by CECO to mitigate this concern.

reg /1086:35

Based on this review, Commonwealth Edison has determined that it is most appropriate to take a prospective approach to improving the total of design calculations. In general,it is not Conunonwealth quality's Edison intent to retroactively upgrade calculations aport to suexistin

• design. If however, a design calculation in identified whic i is technicaly inadequate, then a revision to that calculation will be made, and the impact of that revision will be evaluated. Commonwealth Edison will, however, improve the process of performing design calculations so that future design calculations will be created to a higher standard of excellence than those created in the past.

Since the conclusion of the Quad Cities EDSF1 inspection, significant additional work has been completed by CECO to improve the process of performing design calculations. This work includes:

1. More clearly defining the problem associated with calculation quality.

2. Identifying root causes of the problem t.rmeciated with calculation quality.

3. Developing short term corrective actions to prevent continued problems with calculations.

4. Developing long term corrective actions to insure all root causes will be addressed and that the quality of calculations will meet current expectations for the industry.

In order to address these issues a CECO /AE Workgroup on Calculation quality was established. The workgrou, consisted of representatives of CECO and each of the three dedicated A3 firms (ABB Impell, Bechtel, and Sargent and Ltmdy). The workgroup's review included determining industry standard pr actice by obtaining smnplea front otlier utilities of their calculations and their procedures or guidance for performing calculations.

DefiniUon of the Problem It is CECO's belief that, although the calculations that have been reviewed contain reasonable and accurate results, the quality of the calculations, as measured by level of detail, do not meet current industry expectations.

The review of calculations shows that they appear to fell short in one of two categories:

A. A. calculation.should be legible and belogically.composedmThey should.he_readily_ understandable and suitable for_ reference and useby_a technically _ qualified _individuaLwithout recouretto..the preparer.

The majority ofidentified deficiencies fallinto this aien. These are calculations with weaknesses, but do not contain results that are wrong. Rather, the reasoning that led to the results is not presented in sufficient detail for a third party to follow or to reconstruct the thought processes. Examples of weaknesses in this area:

reg /1086:36

. Assuinptions not stated, not verified, or not justified

. Engineering judgments not stated or not justified

. Iteferences not listed, sources of equations not presented

. Purpose, scope, or conclusions not clearly stated

. No logical flow from step.to step and from assumptions to conclusions

. Writing that is illegible or too light to be copied or microfilmed

. No documentation of software types, versions, or operating systems No reference to sollware verification and validation

11. A. calculation should.be technically _ accurate.und based on sound engineering practices.

Only a small percentage of the deficiencies occurred in this area. These are calculations that contain results that niay not be reliable,in that the reasoning behind the results contains some error. Examples of potential weaknesses in this area:

. Math error

. Incorrect equation or formula

. Invalid technical approach Out of date t eference material

. In crect codes or standards ,

. Use of unsuitable software application

. Use of software application for which no verification and validation was performed

- Unsupported conclusion l

l-reg /1080:37 1

6 Identification of the Hoot Caumw The workgroup assessed the nature of the problem to identify potential root causes, 'I he following provides a summary of the identihed root causes:

A. For Total Quality _of Calculations.

The conclusion was that:

The existing procedures for performing calculations do not provido adequate guidance on the " total quality" of calculations.

B. FotLack of_ Calculation Itequireinents.

The conclusions were that:

. There is no industry standard or other source for objective requh ements,

. There are no standardized methods for performing calculations.

. The problem (of poor calculations) is not fully understood,

. Engineers did not perceive the need for improvement of calculations.

. No resources have been allocated to develop requirements.

. The " audience," or reviewer, has changed.

Short Term Corrective Action The following short term corrective actions have been completed:

A. Prepare and issue a report of the findings and activities of the workgroup. This re port and its lessons learned is being made available to, and will be discussed with, those CECO and AE personnel that perform and review calculations.

B. Prepare and issue a draft 'High Ixsel" technical requirements document that provides the requirements on calculation content and level of detail, lemg Term Corrective Action The recommendations for long term corrective actions have been reviewed against the root causes. CECO believes that those actions will raise the quality of calculations to fully meet the current ex pectations of the industr categories: y, The long term corrective actions fall into the iollowing ,

reg /1086:38

t A. Formally issue the technical requirements document. Reference to this document will be incorporated into the AE Guidebook as

" Interim Guidance."

This action item will be completed by October 1,1991.

B. Generate "Model Calculations" to be used as examples of calculations that fully meet CECO engineering's expectations.

This action item to be completed by January 1,1992.

C. Provide detailed procedures that describe the quality requirements for calculations. These will include the following:

1. Revise ENC Procedure QE 81. Hoview of Asstunptions and Judgments for Architect Engineer Supplied Design Evaluations. This procedure presently guides the CECO cognizant engineer through the process of reviewing assmuptions andjudgments wit iin project calculations. It addresses principally the process now but gives no benchmarks against which to assess (Ite adequacy of assumptions and judgments.

The revised procedure will be expanded in scopo to cover the purpose and conclusions. It will also include a checklist or other tool for judging and trending the quality of calculations. Finally,it will include an evaluation standard. This review is not intended to be an additional level of mathematical verification. Rather,it is intended that the review determine the acceptability of the calculation.

This revised procedure will be issued by December 1,1991.

2. Revise ENC Procedure QE.51.D. This procedure is limited to certain NED calculations, and presently guides the CECO cognizant engineer through the process of performing project calculations. This procedure currently addresses the process flow, but gives na benchmarks against which to judge a calculation's adequacy.

The purpose of the revised procedure will be to provide the process flow and administrative requirements (numbering convention, paragraph headings and their order, format, typing versus handwriting, authority level for review and approval, etc.) for calculations performed by CECO. This procedure will be applicable only to CECO, not to the AEs.

The revised procedure will reference the technical requirements document.

This revised procedure will be issued by January 1,1992.

reg /1086:39

e c.

3. C cate Departinental " Design Guides." These Design Guides describe specific types of calculations and list the computational steps that are applicable to each type. They include references to the applicable codes and standards and a list of those ste under consideration. ps conninon to the type of calculation The creation of" Design Guides" will, by necessity, he an ongoing process.

D. Train CECO and AE personnel on the docuinenta described above.

Initial tt aining is to be coinpleted by the end of the second quarter,1992.

%'t reg /1080:40