ML17199T439

Responds to Open Items Noted in NRC 870916 Interim SER Re Pump & Valve Inservice Testing Program.Revised Inservice Testing Program Will Be Submitted in First Quarter 1988. Relief Requests PR-1,PR-1A,VR-22 & VR-16 Encl

ML17199T439
Person / Time
Site:	Dresden
Issue date:	11/17/1987
From:	Silady J COMMONWEALTH EDISON CO.
To:	Murley T Office of Nuclear Reactor Regulation
References
NUDOCS 8711240109
Download: ML17199T439 (9)

Text

e Common..

th Edison One First NatiJl!r Plaza, Chicago, Illinois Address Reply to: Post Office Box 767 Chicago, Illinois 60690 - 0767 November 17, 1987 Mr. Thomas E. Murley, Director Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory commission Washington, DC 20555

Subject:

Dresden.Station Unit 2 and 3 Response to the +nterim Safety Evaluation for the Dresden Pump and Valve IST Program NRC Docket Nos 50-237/249

Reference:

Daniel R. Muller to L. b. Butterfield Letter dated September 16, 1987 transmitting the subject Interim Safety Evaluation Report

Dear Mr. Murley:

In the referenced letter, the NRC identified several open items concerning the Dresden !ST Program.

The following discussion and attachments respond to e*ach of those items:

1)

Dresden will adopt the ANSI/ASME OM-6 (Draft 8) document as the basis for establishing vibration acceptance criteria for all pumps in the program except the High Pressure Coolant Injection (HPCI) pumps [2(3)-2302].

See the revised Relief Request PR-1 (Attachment A).

The acceptance criteria for the HPCI pumps will continue to be based on the ASHE Section XI criteria due to the inherent vibration levels found in each HPCI pump.

See new Relief Request PR-lA (Attachment B).

Impeller replacement on the associated booster pumps is currently scheduled at the next refueling outage for each unit.

This is expected to reduce pump vibration and may result in levels within OM-6 guidelines.

2)

3)

4)

Based on information provided by the NRC and their consultant (EG&G), Dresden has revised Relief Request VR-22 to include only the 2(3)-305-115 valve and has amended the alternate testing method for that valve to conform to the recommendations made by the NRC.

Relief from testing the 2(3)-305-138 is no longer required because the valve must close eac~ time a control rod is moved.

A control rod could not be moved if the 305-138 valve were to fail.

Therefore, as part of the control rod exercise surveillance performed during normal operation, the 2(3)-305-138 valve will be exercised.

See revised Relief Request VR-22 (Attachment c>. ~a; 7112tarci9' -eii1~fT*t...,.., __,*~,

1

• An*ocu 05000~... 7 * * * * '

same as 2).

\\

!PR, :

. ~

P.DR,,-L __ ;. -.. i

. -. l As required ~Y the~ NRC,.Dresden -wil-1 wisual-ly **inspect the_

2( 3)-'2301-39 'va'lve on a schedule of once per* refueling outage until inspection results allow lengthening the inspection interval.

Relief Request VR-21 may*be deleted and a new relief request will be submitted when inspection results warrant.

5)

As required by the NRC, Dresden will change the definition of fast-acting valve from "those that stroke in five seconds or less" to "those that stroke in two seconds or less".

See the revised Relief Request VR-16 (Attachment D).

6)

As indicated in the reference, a number of administrative/clerical changes to the program have been previously agreed to which do not affect the program's content.

on August 17, 1987, an NRC inspection team from AEOD began a two week diagnostic evaluation of Dresden Station.

As a result of the diagnostic audit, a recommendation was made to re-evaluate the Dresden Station In-Service Testing Program.

In response to this recommendation, Dresden Station has decided to submit a new and revised program for NRC review.

This revised program will be submitted to the NRC when completed in the 1st quarter of 1988.

CECo believes that an onsite meeting for detailed discussions with the NRR and EG&G reviewers would be very beneficial after the revised program is submitted.

Until NRC approval of the new program is received, Dresden Station will follow the existing IST Program as approved by the NRC.

Yours very truly,

~1.~

Nuclear Licensing Administrator cc:

M. Grotenhuis - NRR NRC Resident Inspector - Dresden Regional Administrator - Region III 3841K/bs

Attachment A RELIEF REQUEST NO. *PR-1.

PUMP NUMBER:

All centrifugal pumps in program except High Pressure Coolant Injection Main and Booster Pumps 2(3)-2302 (see Relief Request PR-lA).

SECTION XI REQUIREMENTS:

IWP-3100 Inservice Test Procedure; IWP-4500 Vibration BASIS FOR RELIEF:

Pump vibration and bearing temperature are required to be measured to dete~t any changes in the mechanical characteristics of a pump..

This is to detect developing problems so repairs can be initiated prior to a pump becoming inoperable (.i.e., unable to perform its function).

The ASME Code minimum standards require measurement of the vibration amplitude displacement in mils. (thousandths of an inch) every three months and bearing temperatures once per year.

Dresden Station proposes an alternate program which is believed to be more comprehensive than that required by Section XI.

This program consists of performing the required vibration readings in velocity rather than mUs displacement.

This technique is an industry-accepted method which is much more meaningful and sensitive to small changes that are indicative of developing mechanical problems.

These velocity measurements detect not only high amplitude vibrations that indicate a major mechanical problem but also the equally harmful low amplitude - high frequency due to misalignment, imbalance, or bearing wear that usually go undetected by simple displacement measurements.

In addition, these readings go far beyond the capabilities of a bearing temperature monitoring program, which requires a bearing to be seriously degraded prior to the detection of increased heat at the bearing housing.

The vibration velocity readings on a schedule of once every three months achieve a much higher probability of detecting developing problems than the once per year reading of bearing temperatures.

Data gathering on bearing temperatures also is *not without its own problems.

The enforced thirty minute run time, [i.e., IWP-3500(b) - three successive readings taken at ten minute intervals that do not vary more than 3\\], causes problems with pumps having no recirculation/test loop.

Variations in the temperature of the pumped fluid also complicate analysis of bearing temperatures when attempting to trend any developing problems from year to year.

Improper interpretation of results could result in unnecessary pump maintenance.

In addition, it is impractical to measure bearing temperatures on many of the pumps in the program.

Some specific examples are as follows:

e

• • e A)

CORE SPRAY 2(3) - A, B 1401 These pumps are lubricated by the pumped medium.

The temperature of the pumped liquid could have an"adverse affect on the accuracy of any yearly trending.

B)

LPCI 2(3) - A, B, C, D 1502 Same as A) above.

C)

CCSW 2(3) - A, B, C, D - 1501-44 These bearings are contained in an oil-filled reservoir.

The ambient temperature of the pump space is changeable, thereby varying the start temperature of the data.

Results would be extremely difficult to trend from test to test.

D)

DIESEL GENERATOR COOLING WATER PUMP - 2, 2/3, 3-39038 The bearings on these pumps are internal to the unit because the pump and motor are directly attached to each other and at no point is a bearing accessible for temperature readings.

The foregoing reasons demonstrate that the proposed program of vibration measurements is a more practical method of testing which exceeds the requirements of the ASME Code.

ALTERNATIVE TESTING:

The alternative testing described herein for pump vibration monitoring was developed using ANSI/ASME OM-6 (Draft 8) as a guideline, with the intent of incorporating into the program any additional requirements of OM-6 to those found in ASME Section XI.

Pump vibration measurements will be obtained and recorded in velocity (inches per second), and are broadband (unfiltered) peak readings.

All monitored locations are clearly marked to identify the specific point at which the transducer. is to be placed while taking vibration measurements using portable equipment.

The readout system and transducers used to take vibration measurements are capable of frequency response in the range of one-third minimum pump speed to at least one-thousand hertz, and they have a minimum accuracy over that range of +/~5\\.

All centrifugal pumps in the program will have vibration measurements taken in a plane approximately perpendicular to the rotating shaft in two orthogonal directions on each accessible pump bearing housing.

Measurement will also be taken in the axial direction on all bearing housings when accessible.

Reciprocating pumps will have vibration measurements tak.en approximately perpendicular to the cran~shaft *and the line of plunger travel, including the axial direction when accessibie oh each pump bearing housing.

The Acceptable, Alert, and Required Action Ranges.of OM-6, Table 6100-1 will be used in lieu of Table IWP 3100-2 of Section XI as shown below.

Corrective actions will be taken in accordance with Article IWP-3230.

Vibration Ranges Pump Type Acceptable Range Alert Range Required Action Range Centrifuga1(3)

Os_V5_2.5VREF (1) 2.5VREF<V5_6VREF V>6VREF But Not > 0.325 in/sec<2)

But not >.7 in/sec(2)

Reciprocating Notes:

(1)

VREF is the reference velocity det~rmined when the pump is known to be running acceptable.

(2)

All upper limits are listed in inches/second (peak).

(3)

All centrifugal pumps in the IST Program operate at a *Spee<:}

greater than 600 rpm.

Attachment B RELIEF REQUEST NO. PR-lA PUMP NUMBER:

2(3)-2302 - High Pressure Coolant Injection (HPCI).

SECTION XI REQUIREMENTS:

IWP-3100 Inservice Test Procedure; !WP 4500 Vibration BASIS FOR RELIEF:

As stated in PR-1, the measurement of vibration velocity in lieu of vibration displacement and bearing temperature is the pref erred method of detecting mechanical change in pumps:' However, vibrafions have been observed on the Dresden HPCI pumps, as well as other similar units throughout the country, which are greater than the maximum values specified in the ANSI/ASME OM-6 (Draft -8) document.

The pump manufacturer attributes the high vibration to a vane passing frequ~ncy vibration created by the booster pump and transmitted hydraulically to the main pump where the high vibrations are observed.

As a corrective action, Dresden plans to replace the present booster pump impeller with a staggered-vane.impeller which, as seen at other plants in the cotintry, will reduce the vibration to an acceptable level.

This work is.Planrted on *-uhit 3 during the Spring, 1988 refueling outage and on Unit 2*duri:ng the Fall, 1988 refueling outage.

As stated in PR-1, measurement of bearing temperatures has many inherent problems arid can produce very misleading results.

In the case of HPCI, the pumps are driven by a steam turbine which exhausts to the suppression pool.

Extended run times to stabilize bearing temperatures would create problems in maintaining the suppression pool temperature below the Technical Specification limit of 95°F and the suppression pool level within its limits.

ALTERNATIVE TESTING:

For the HPCI pumps only, vibration amplitude will be measured in units of displacement.

Dresden will continue to use ASME Section XI, Table !WP 3100-2 as the basis for establishing vibration limits.

Bearing temperatures will not be measured for reasons stated above.

After the booster pump impellers are replaced and data is obtained regarding the vibration characteristics of the HPCI pumps, Dresden will consider monitoring HPCI vibration amplitude in units of velocity and, if possible, establish the vibration limits based on ANSI/ASME OM-6 (Draft 8).

.1

. ~

I

• *~

.~*

-::t. -* :~ *-

1

.7.

-*,r;::

, t:,

.)

.j t_.

'l'

,--;;J

r

,r

• J r-'\\*

t..:..

~*

~:

Attachment c RELIEF REQUEST NO. VR-22 SYSTEM:

Control Rod Drive (CRD)

COMPONENT:

2(3)-305-115 CATEGORY:

C FUNCTION:

Located in the charging water line to the CRD accumulator TEST REQUIREMENTS:

CT-1 exercise valve every 3 months.

BASIS FOR RELIEF:

In order to test this valve during operation, the CRD pump would need to be shutoff.

In doing this, a 250°F temperature limit recommended by General Electric could be exceeded causing damage to the seals on the control rod*

drives.

Also, shutting off the CRD pump would stop seal purge water flow to recirculation pump seals.

ALTERNATE TESTING:

Bach refueling outage, the CRD pump will be shutoff to test the integrity of the valves.

The integrity of the valves will be verified by the ability of each CRD accumulator to hold its pressure above the low pressure alarm setpoint.

A time limit will be selected (based on vendor recommendations, operating/maintenance experience, and technical input from other sources) for which the accumulators must maintain their pressure above the low pressure setpoint.

Attachment D RELIEF REQUEST NO. VR-16 SYSTEM:

All Systems COMPONENT:

All power operated valves requiring full stroke testing CATEGORY:

A and B FUNCTION:

Power operated valves TEST REQUIREMENTS:

Stroke time evaluation per IWV-3413(c)

BASIS FOR RELIEF:

Paragraph IWV-3413(c) requires that valve stroke times be evaluated against the previous stroke time to determine if corrective action is required.

To establish consistency in evaluating stroke times and make program implementation more practical, Dresden Station proposes to establish a reference stroke time for each valve which will be used for evaluating performance.

This reference value will be determined by averaging stroke times.

This actually results in a tighter band of acceptable stroke times but is much easier to administer.

The limiting value of full stroke time for each valve will remain as listed in the IST Program tables.

In addition, it is impractical to apply the requirements of IWV~3413(c) to valves with very short stroke times (i.e., < 2 seconds), particularly solenoid valves which typically have full stroke times under one second.

For these short stroke time valves, variances of 50 percent or more can occur in the measured times for reasons that are in no way related to valve performance, for example, operator reaction times.

In these specific cases, verifying that the valve stroke-- times do not exceed 2 seconds would be sufficient to evaluate valve performance..

ALTERNATE TESTJ:NG':

Each stroke time will be compared to an Alert Stroke Time.

The Alert Stroke Time will be established as follows:

1.

For Valves with referenc~ stroke times greater than 10 seconds, the Alert Stroke Time will be 25 *percent ~greate'r -than the reference time.

2.

For valves with reference stroke times between 2 seconds and 10 seconds, the Alert Stroke Time will be set at 50 percent greater than the reference time.

3.

For valves with reference stroke times between 1.5 and 2 seconds, the Alert Stroke Time will be set at 3 seconds.

4.

For valves with reference stroke times less 1.5 *seconds, the Alert Stroke Time will be set at 2 seconds.

If a measured stroke exceeds the Alert Stroke Time, the test frequency shall be increased to once each month until corrective action is taken, at which time the original test frequency shall be resumed.

3841K