ML20214P844

Forwards Nonproprietary & Proprietary Responses to NRC Questions Re Util 851029 Proposal to Eliminate Resistance Temp Detector Bypass Sys.Proprietary Response Withheld (Ref 10CFR2.790)

ML20214P844
Person / Time
Site:	Mcguire, Catawba, McGuire, 05000000
Issue date:	05/26/1987
From:	Tucker H DUKE POWER CO.
To:	NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM)
Shared Package
ML19292H326	List: ML20214P844
References
NUDOCS 8706040108
Download: ML20214P844 (76)

Text

-

DUKE POWER GOMRWY P.O. BOX 33180 CHARLO1"rE, N.C. 28242 IIAL B. TUCKER ren.ernoxe exisreessons, (704) 373-4531 MtT4 EAR PRODUCTION May 26, 1987 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C.

20555

Subject:

McGuire Nuclear Station Docket Numbers 50-369, -370 Catawba Nuclear Station Docket Numbers 50-413, -414 RTD Bypass Elimination Gentlemen:

On October 29, 1985, Duke Power Company submitted a proposal to eliminate the RTD Bypass System from McGuire Nuclear Station. On April 14, 1987, a meeting was held among representatives of Duke, Westinghouse Electric Corporation, and NRC Staff to discuss the proposed modification. Enclosed are the NRC questions and the re-sponses to them, which were prepared by Duke and Westinghouse. Please note that contains information which is proprietary to Westinghouse Electric Corporation and should be withheld from public disclosure. The appropriate affidavit is included at Attachment 5 of Enclosure 1. contains non-proprietary versions of the information presented in Enclosure 1.

Very truly yours,

/

/

N Hal B. Tucker SAG /73/jgm 4

h Attachment Dett &

• fog

/

I A*I J

Woe t iAlf EP 2858" Rggg6

'$g,p a

PDR a

i Document Control Desk May 26,1987 Page 2 xc:

(w/o Enclosure 1)

Dr. J. Nelson Grace, Regional Administrator U.S. Nuclear Regulatory Commission - Region II 101 Marietta Street, Suite 2900 Atlanta, Georgia 30323 Mr. Darl Hood, Project Manager Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C.

20555 Dr. K.N. Jabbour, Project Manager Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C.

20555 Mr. W.T. Orders NRC Resident Inspector McGuire Nuclear Station Mr. P.K. Van Doorn NRC Resident Inspector Catawba Nuclear Station

s.

(<;*.'

AW-76-60 r,.

If AFFIDAVIT-t t

~,

COMMONWEALTH OF PENNSYLVANIA:

ss COUNTY OF ALLEGHENY:

Before me, the undersigned authority, personally appeared 1

. Robert A. Wiesemann, who, being by me duly sworn according to law, de-i

{

poses and says that he is authorized to execute this Affidavit on ' behalf p

of W'estinghouse Electric Corporation (" Westinghouse") and that the aver-ments of fact set forth in this Affidavit are true and correct.to the best of his knowledge, information, and belief:

__/$

r tbfL 01&tM.44 Robert A. Wiesemann, Manager Licensing Programs r

Swora to and subscribed before,methis8-r.

day i

of Ni*hublU 1976.

l N/L NM es

~/

Notary Public,,,

. i:'

~

i I.

v

)

)

i l

+

i..

?.

' AW-76-60 A.

F

~

(1) I am Manager, Licensing Programs, in the Pressurized Water Reactor Systems Division, of Westinghouse Electric Corporation and ai'such,

,1 have been specifically delegated the function of reviewing the proprietary information sought to be withheld from public dis-closure in connection with nuclear power plant licensing or rule-I making proceedings, and am authorized to apply for its withholding I

'on behalf of the Westinghouse Water Reactor Divisions.

t (2) I am making this Affidavit in conformance with the provisions'of 10 CFR Section 2.790 of the Commission's regulations and in con-e

[

junction with the Westinghouse application for withholding.ac-companying this Affidavit.

I I

(3)

I have personal knowledge of the criteria and procedures utilized

~

~

' ~,

by Westinghouse Huclear Energy Systems in designating information as a trade secret, privileged or as confidential commercial or L

financial information.

~

(4) Pursuant to the provisions of paragraph (b)(4) of Section 2.790

~

of the Commission's regulations, the following is furnished for i

consideration by the Commission in determining whether the in-f formation sought to be withheld from public disclosure should be withheld.

~

(i) The information sought to be withheld'from public disclosure l

is owned and has been held in confidence by Westinghouse.

lI s

~

~

~.

I

' AW-76-60 (ii) The information i.s of a type customarily held in confidence by r

blestinghouse and no't customarily disclosed to the pub'lic.

~

Westinghouse has a rational basis for determining the types of j

information customarily held in confidence by it and, in that connection, utilizes a system to determine when and whether to hold certain t;ypes of information in confidence. The ap-plication of that system and the substance of that system,

constitutes Westinghouse policy and provides the rational; basis required.

r Under that system, information is held in confidence if it

-[

falls in one or more of several types', the release of which

~

T might result in the loss of an existing or potential com-

~

j petitive advantage, as follows:

a (a) The information reveals the distinguishing aspects of a h

process (or component, structure, tool, method, etc.)-

where prevention of its use by any of Westinghouse's

,{

competitors without license from Westinghouse constitutes

[

' a competitive economic advantage over other. companies.

Y l

,I (b)

It consists of supporting data, including test data, relative to a process (or component, structure, tool, i

method, etc.), the application of which data secures a competitive econom'ic advantage, e.g., by optimization or

-- improved marketability.

~

e

N*i.

f'i.

j-AW-76-60

's It' W

j (c)

Its use by,a competitor would reduce his expenditure

'r l

of resources or improve his competitive positio'n.in the

~

l' design, manufacture, shipment, installation, assurance of quality, or ifcensing a similar product.

(d)

It reveals cost or price information, production cap-acities, budget levels, or commercial strategies of, Westinghouse, its customers or suppliers.

/

{

(e) It reveals aspects of past, present, or future West-i inghouse or customer funded development plans and pro-grams of potential commercial value to Westinghouse.

~

/

(f)

It contains patentable ideas, for which patent pro-i tection may be desirable.

~

(g)

It is not the property of Westinghouse, but must bc treated as proprietary by Westinghouse according to agreements with the owner.

I There are sound policy reasons behind the Westinghouse f

system which include the following:

(a) The use of such information by Westinghouse gives

~

Westinghouse a competitive advantage over its com-petitors.

It is,' therefore, withheld from disclosure to protect the Westinghouse competitive position.

r i

i

l*'

'l AW-76-60 e

~

a.

i i '

(b)

It is information which is marketable in many ways.

e The, extent to 'which such infomation is available. to.

V competitors diminishes the Westinghouse ability to sell products and services involving the'use of the information.

l (c) Use by our competitor would put Westinghouse at a fi v

competitive disadvantage by reducing his expenditure G

I of resources at our expense.

i (d) Each component of proprietary infomation pertinent to a particular competitive advantage is potentially i

4 as valuable as the total competitive advantage.

If

_.I competitors acquire components of proprietary infor-j mation, any one compone. t may be the key to the entire n

]o!

puzzle. thereby depriving Westinghouse of a corpetitive 9

]4 advantage.

lf (e) Unrestricted disclosure would jeopardize the pos.ition

' of prominence of Westinghouse in the world market, and thereby give a market advantage to the competition j

l i.

in those countries.

J I f 4

j (f) The Westinghouse capacity to invest corporate assets in research and de'velopment depends upon the success

~

]t in obtaining and maintaining a competitive advantage.

i 1

I.

e l.

l:

j.

i

=_

h, *.

i,

c. i.

. AW-76-60

? I (iii)

The information.is being transmitted to the Comission in e

E confidence and, under the provisions of 10 CFR Section 2.790, i

'L it is to be received in confidence by the Commission.

1 (iv)

The information is not available in public sources to the best of our knowledge and belief.

}i,.

(v)

The proprietary information sought to be withheld in this,sub-mittal is that which is appropriately marked in the attach-E.

ment to Westinghouse letter number NS-CE-1298, Eiche1dinger to Stol'z, dated December 1,1976, concerning information relating

?

to NRC review of WCAP-0567-P and WCAP-8568 entitled, " Improved J r Thermal Design Procedure," defining the sensitivity of DNB l

I ratio to various core parame. ers. The letter and attachment t

are being submitted in response to the NRC request at the

.g

'l October 29,1976 NRC/ Westinghouse meeting.

I i!

This information enables Westinghouse to:

I (a) Justify the Westinghouse design.

L

~

(b) Assist its customers to obtain licenses.

(c) Meet warranties.

if, (d) Provide greater operational flexibility to customers

! I assuring them of safe and reliable operation.

ll l6 (e) Justify increased power capability or operating margin' for plants while assuring safe and reliable operation.

t b

i O

h__,._,

y

I

(., :1 4

a,e.*7',

L../ ',

, AW-76-60 j;'

Lo a (f) Optimize rea'ctor design and performance while maintaining a high level of fuel integrity.

~

Further, the information gained from the improved thermal design procedure is of significant commercial value as follows:

(a) Westinghouse u'ses the information to perform and justify analyses which are sold to customers.

t (b) Westinghouse sells analysis services based upon the experience gained and the methods developed..

E Public disclosure of this information concerning design pro-I I

cedures is likely to cause substantial harm to the competitive -

^

position of Westinghouse because competitors could utilize I

this information to assess and justify their own designs without commensurate expense.

The parametric analyses performed and their evaluation represent a considerable amount of highly qualified development effort.

This work was contingent upon a design method development pro-gram which has been undeniay during the past two years.

I Altogether, a substantial amount of money and effort has been <

expended by Westinghouse which could only be duplicated by a competitor if he were to invest similar sums of money and pro-vided he had the appropriate talent availab'7.

i Further the deponent sayeth not.

e e

.g*

b

_= = - _.

4 i

l i

PROPRIETARY INFONTION NOTICE

\\

I TRANSM:TTED HREWITH ARE PROPRIETARY AND/0R DOCUMENTS FURNISHD TO ME NRC IN CONNECTION j

PLANT SPECIFIC REVIEW AND APPRWAL.

IN OR ER 10 CONFORM 10 THE REQUIRDEN75 W 10 RIDULATIONS CONCERNING THE PRMECTION W P TO THE NRC, THE INFORMATION WHICH IS PROPRIETARY IN CONTAIND WITHIN BRACKET 5 AND WHRE THE PROP l

DELETC IN THE NON-PROPRIETARY YESIONS GC.T THE BRACKET l

i INFOMTION 1 HAT WAS CONTAINED WITHIN ME B 4

THE JUSTIFICATION FOR CLAIMINO THE INFORMA HAVIE BEDI DELETE.

DESIGNATED AS PROPRIETARY 15 ICICATE IN BO i

(.

LCTERS (s) EROUGH (g) CONTAING WITHIN PARENTHISD LOCA 4

l DE.DIATILY FOLLCWIE THE BRACKETS INCI.3ING E i

THDE j

DENTIFIED As PROPRIETARY OR IN THE HARGIN OPP 2

LOWD CASE LETTERS REFER TO THE TYPD E IN l

H15 IN CONFIDENCE DENTIFIED IN SECTIONS (4)(11)(a) throug l

AFTIDAVIT ACCOMPANYIM MIS WAN5MITTAL PUR5JA k

i I

k I

~-------,ge,yww,-e.w,we,,~,--

,.m-9%,-

q,---_pp,-,---.-e--*w e-ywy,w

--g-,--y,

--e--+-w--e-w m --

4 ENCLOSURE 2 r

(NON-PROPRIETARY)

CONTENTS:

l i

ATTACHMENT 1 RESPONSE TO NRC QUESTIONS 1

ATTACHMENT 2 HOT LEC TEMPERATURE STREAMING i

MEASUREMENTS AT McGUIRE UNIT 1 ATTACHMENT 3 RCS N0ZZLE AND THERM 0WELL LOCATIONS ATTACHMENT 4 RTD BYPASS ELIMINATION UNCERTAINTY CALCULATIONS I

i i

)

1 l

i i

l I

i

1 WESTINGHOUSE CLASS 3 l

ATTACHfiENT 1 RESPONSE TO NRC QUESTIONS 1

1

1 I

1 Cover letter Your cover letter of October 29, 1985, mentions j

f current testing being conducted at McGuire to support conclusions of the Westinghouse report. Provide the results of this testing.

l 1

RESPONSE: A description of the testing and the results are provided in

.. " Hot Leg Temperature Streaming Measurements of McGuire Unit 1".

2 4 (1.2.2(a) to (c))

The existing cold les RTD bypass penetration nozzle is i

to be modified to accept the RTD thernowell and a new penetration is to be made to each cold les to accept an additional well-mounted narrow range RTD.

Indicate the relative locations of these two RTDs by a i

i description which includes dimensions of their proximity to each other.

RESPONSE: Drawings showing the RTD locations are provided in Attachment 3.

j The drawings provided are for the Catawba Nuclear Station, McGuire cold leg RTD placements will be similar.

3 8 (1.3.3)

For RTD failures, a spare RTD will be available in the cold leg. You state that a failure of an RTD in the,,

hot leg will require manual action to defeat the failed signal and that a manual rescaling will be made of the electronics to average the remaining signals.

1 What is the time interval to defeat the failed signal?

What is the time interval to rescale the electronics j

to average the remaining signals? Describe the steps involved in this process.

RESPONSE: The operators will be alerted to the failure of a RTD by the T ** or t.T deviation alarms. Upon recognition of the failed RTD, the

• affected chsunel will be declared inoperable and placed in the tripped condition within 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> per Technical Specification 3.3.1.

The rescaling and return of the channel to service will not be required within a specified time period since the tripped channel J

results in a conservative 1/3 protection system logic.

The rescaling process is as follows:

A C.

l

T A,c.

t l

i 4

10 (2.1)

Westinghouse has performed RTD response time tests for the fast response RTD manufacturerd by the RdT Corporation in a configuration modelling actual in-plant installations. The results (Table 2.1) i indicated a certain mean response time for the RTD, I

l thereowell, and scoop. Mas this test been duplicated in the current McGuire testo? If so, are the results j

the same. Table 2.1 compares the obtained response r

j time for the RdT RTD with the response time for the current system which uses Rosemount RTDs. Ws the response time for the Rosemount RTDs obtained from tests similar to those for the Rdr RTDai Wy is the RTD response time of the " fast response RTD" slower than existing Rosemount RTD? Also, explain the reasons for the differences in RTD filter time constants in Table 2.1 for the two systems. Does this i

filter time difference relate to the temperature q

oscillation probles referred to in request 11 3 below?

l RESPONSE: The testing performed at McGuire and discussed in Attachment 2 was **

1 linited to data collection associated with temperature streasing.

4 The difference in the response time characteristics between the RTD Bypass Systes and the proposed thernovell arrangement is due in large part to the introduction of the thernovells. The existing bypass system involves direct immersion of the Rosemount RTDs into i

i the coolant in the bypass manifold.

In addition to the thersowell, j

dif ferences in the construction, design, and/or materials used by RdF and Rosemount RTDs also impact the response times.

I j

The RTD filter time constants were changed from 2 seconds to 6 seconds (November 16, 1984 subeittal for McGuire 2 Cycle 2 0FA Reload) in response to hot les temperature oscillations which were l

j causing spurious OTAT/0 PAT channel trips. The proposed filter time j

constant for the thernovell system (3 seconds) saintains the overall j

response time of to seconds. Since the RTD/thernowell systes also l

acts as a las function, the electronic filter time constant can be reduced without increasing the risk of spurious trips due to the hot j

les temperature oscillations.

5 11 (2.2)

For the three objectives for the stressing tests for l

McGuire Unit 1, only the first is clearly stated.

4 W ere are the other two objectives?

RESPONSE: The objectives of the stressing tests for McGuire Unit I weres a) Determine the magnitude of the differences between branch line l

temperatures i

l

b) Confire the short tere and long ters stability of the tempera-ture stressing patterns c) Evaluate the impact on the indicated temperature if only 2 out of.the 3 branch line temperatures are used to determine an average temperature 6

11 (2.2)

The* operator can review temperatures recorded prior to RTD failure in order to take further steps to correct the "two RTD" average to obtain the "three RTD" espected reading. Now such time does this procedure take and how accurate is the new value?

RESPONSE: The procedure to correct the two RTD average will be performed with the channel declared inoperable and in the tripped condition and can therefore be completed without a required schedule.

The increased uncertainty is discussed in Attachment 2 and, Appendix to the RTD Bypass Elimination Licensing Report for McGuire Units 1 and 2.

7 12 (2.2)

The McGuire Unit 1 tests are indicated to have provided information not obtained before on tempera-tures from the pipe interior, and this is indicated

)

to have greatly strengthened the assumptions and inferences made with previous test data which investigated temperature gradients near the pipe I

surface. What did the results show to reinforce the accuracy of the average temperature of the hot les?

Provide these results.

RESPONSE: Additional information regarding the streaming tests are provided in.

3 12 & 13 You state that reactor coolant flow is verified with (3.1) a calorimetric measurement and that two of the most important parameters are the narrow range hot leg and the cold les coolant temperatures as measured by the RTDs. The uncertainty of the proposed fast response RTD is given and is indicated to be somewhat higher than that of the RTD being replaced. Now much higher is it? A new flow measurement uncertainty analysis is needed for verification of the new values.

Provide results of this analysis for our review.

RESPONSE: For McGuire, the original RTDs installed were manufactured by Rosemount and had a total uncertainty, as used by Westinghouse, of l

(

)+a, c.

These RTDs are being replaced by RfDs manufactured by j

RdF, which have a total uncertainty of (

]+a,c.

The accuracy used in the RCS T1ow calorimetric measurement is [

]+a c which reflects the accuracy of the Westinghouse RfD cross-calibration procedure. Therefore, while the RdT has a larger uncertainty for I

protection functions, it has a smaller uncertainty for the RCS Flow calorimetric. Catawba's original RTDs were manufactured 'oy RdF.

i thus there is no uncertainty change for this plant. Revised RCS Flow calorimetric measurement uncertainties have been provided for each plant. The net result is that the RCS Flow uncertainty is the same for both plants, [

]+a, c without venturi fouling allowances (See Attachment 4).

9 13 (3.1)

A procedure is described for obtaining a more accurate temperature measurement in the calorimetric flow measurement procedure. Does this method require a standard temperature to measure against?

If so, what is this standard? Has this procedure been demonstrated in an actual test? You state that this method should give an overall flow measurement uncertainty about the same as the existing value of 11.7 percent flow (not including the 0.1 percent for feedwater venturi fouling allowance). This will require verification in a revised flow measurement analysis as per request 8 above. Provide this analysis.

RESPONSE: The October 29, 1985 submittal described an RCS Flow calorimetric measurement technique which relied on the measurement of a Delta-T and T to from an isothermal condition (0% RTP) for both TH establishthefullpowervalueswithahighdegreeofpfecision.

I This technique was considered to be possibly necessary due to some

large potential calibration uncertainties for the RTDs. Since that submittal was made, Westinghouse has resolved the calibration uncertainties by the use of a multiple point RTD cross-calibration

)

procedure. This allows the determination of the absolute tempera-ture for the hot and cold legs in the standard manner i.e., direct measurement. The cross-calibration procedure has been described in other submittals and discussions with the NRC staff. The RCS Flow calorimetric uncertainty calculations noted in the response to (8) are based on use of RTDs that have been through the cross-calibration process and thus have an accuracy of (

j+a,c.

10 14 (3.1)

In paragraph b, last sentence, what RTDs are used?

Does this include hot and cold less7 RESPONSE: The procedure described in 3.1 has been revised as described in Response (9). Rdf RTDs are used for narrow range hot les and cold les temperatures.

11 15 (3.2)

A value is presented for the stressing temperature uncertainty from recent calculations for the thereo-well RTD system in the hot leg. You state that the overall temperature stressing uncertainty applied to 4

the calorimetric is slightly higher than in previous analyses. How much higher is it? The temperature j

measurement uncertainty affects the flow measurement uncertainty. A new flow seasurement analysis is needed for verification as per request 8 above.

Provide the analysis for our review.

_ = _ -.

t O

RESPONSE: As noted in the RCS T1ow calorimetric uncertainty calculations provided as part of the response to (8), hot les streasing values of

[

.]+a, e are used. While the arithmetic sua of these values is less than the [

)+a, e that Westinghouse has used in the past, the mathematical impact of having a random component (which can be divided by the number of primary i

loops) and a systematic component (which is applicable for all loops) is slightly greater than the original value.

In actual fact, the overall impact of the splitting of two components is lost in the 7

round-off, and thus does not have any significant effect.

12 15 (3.2)

The new method of measuring hot les temperature with thernowell RTDs located in the three scoops is stated to be at least as effective as the existing RTD bypass I

systes, even though the new method measures tempera-ture at only one point within the thernowell. Discuss what data exists to support the maximum inferred i

temperature gradient value presented.

RESPONSE

(See last page of Attachment 1) 13 16 (3.2)

The last paragraph states that test data have been I

collected from McGuire Unit 1 to provide a plus or j

minus value for a bias to be applied if one of the three RTDs is out of service. How is the bias value *

• l determined for a particular out-of-service RTD since the temperature value of each RTD may dif fer depending on its circumferential position?

RESPONSE: The temperature indication of each hot les RTD will be periodically recorded. Upon the failure of an RTD and the switch to 2 out of 3 averaging, the bias value will be input such that the average of the

~

2 operable RTD's is equal to the most ecent record of the average of the 3 RID's criar to the failure.

] *.4 I

14 17 & 18 The RTD response time of the new Rdr RTDs as compared (4.1) to the currently used Rosesount RTDs is shown in Table 4.1 to differ.

Because of the elimination of the RTD bypass piping and thermal las and a change in I

the RTD electronic filter time constant, the total l

response times of both the existing and proposed systems are shown to be 10.0 seconds. Explain why the value specified in Table 4.1 for the RTD electronic i

filter time constant was previously needed and why the specified value for the filter time constant in the new system can be used. Does this relate to the l

temperature oscillation probles referred to in request

!! 3 below?

I j

RESPONSE: The value for the electronic filter time constant was changed from 2 seconds to 6 seconds during the refueling outage at the end of McGuire Unit 2 Cycle 1.

Temperature oscillations in the McGuire i

i l

t i'

Unit 2 kot legs resulted in spurious OTAT/0 PAT channel trips. The i

oscillations are shown in Figure 1.

The oscillations eststed in i

each loop but magnitudes varied. The cause of the variations in hot l

1es teePeratures is believed to be gaps in the hot les aosale/down-l comer interface resulting in streams of cold water in the het legs.

I i

These streams are periodically picked up by the RfD scoops and cause the oscillations.

J i

j Justification for the change was included in the McGuire Unit 2 l

}

Cycle 2 0FA reload related submittal to the NRC (Nov. 16,1984).

[

The safety analysis assumed a total 0747 response time of 10

!l seconds. The analysis for the elleimation of RfD bypass also assumed a 10 second total response time but the allocation of delays l

was changed due to the las introduced by the RTD/theseowell design.

1 I

The reduction in the electroste filter time coastsat is compensated I

for by the increased RTD/thernovell las and thus the risk of spurious trips should not be increased.

15 18 (4.2)

You have provided the total uncertainty of the RTDs aanufactured by the RdF Corporation and you state that j

the Rosemount RTDs have a different uncertainty. What is the total uncertainty of the Rosemount RTDs?

f RESPONSE: The allowance for uncert@ nty sed for the Rosemount RfDs for the "8'

j RCS flow calorimetric is 16 19 (4.2)

You state the results of system uncertainty calcula-tions verify that sufficient allowance has been made in the reactor protection systes setpoints to account j

for the increased RTD error for the new RdF RTDs as j

i compared to the posemount RTDs. From this, you conclude that the current values of the nominal i

setpoints (given in Section 4.2 of Reference 1) as defined by the McGuire Technical Specifications remain valid. Provide the results of the uncertainty calculations and show how they verify that there is sufficient allowance in the reactor protection set-points to account for the increased RTD temperature Show what the RTD error is now and what it i

error.

l would be after the proposed design change.

RESPONSE: Westiaphouse used the NRC reviewed and approved Westinghouse 8etpoint Methodology to determine the acceptability of the McGuire and Catawba protection functions which are impacted by a change in the susbar or type of RTDs.

In all cases, the Total A11ovence is greater than the Chamael Statistical Allowance resultias in a positive margia, thus indicating that sufficient allowance was made for the lastrument uncertainties.

In these calculations, Westing-j house used a hot les streening value of [

1+a, c, as RfD calibration accuracy of I

]+a, e and a drift value of

(

J+a,c.

These are consistent with calculations performed j

for the following plants, Vogtle (setpoint study), Shearon Narris 1

I (RCS Flow), South Teams (setpoint study and RTD Sypass elimination),

)

an3 Byron /Braidwood (RTD Bypass elimination).

Pre RfD Bypass l

i.

59 i

I i

W

~

58 l

LOOP A i

AT (F) i I

57 -

56 l

l 55 i.

0 1

Time (minutes)

FIGURE lt McGUIRE UNIT 2 OSCILLATION IN AT (T}{0T C0!.D'

~

(Data recorded on Feb. 10, 1984)

O

Elimination calculations and uncertainties are noted in the setpoint studies submitted for each of the two sites (McGuire and Catawba).

17 19 (4.2)

You state that "The Chapter 15 non-LOCA safety analyses were performed assuming that, at steady state full power, the average RCS temperature was equal to the nominal value plus 5.5'T for non DNB events.

For DNB events, at steady state full power, the average RCS temperature is assumed to be at its nominal value; the uncertainties are convoluted into the design limit DNBR value." Explain the significance of the 5.5'F value and how it was obtained.

RESPONSE: For the non-DNB events a 4.0 degree F uncertainty is applied to the j

average RCS temperature. An additional 1.5 degrees F is assumed to address steam generator fouling allowance. For the DNBR events, the temperature uncertainty is convoluted into the DNBR margin and the 1.5 degree steam generator fouling allowance is applied to the average RCS temperature.

18 19 8, 20 You state that the following RCS flow values are used in the Chapter 15 Safety Analyses:

(a) 388,880 spm for DNB transients (b) 382,000 spe for non-DNB transients (c) 377,000 sps for loss-of-coolant accidents The 388,880 appears to be from the design flow (4 pumps x 97,220 spe/ pump). Please indicate how the flows for (b) and (c) were determined. The value for (b), 382,000 spa, appears to use the 1.8% flow measurement accuracy and the value for (c), 377,00 spm, is approximately 97% of the design flow.

Although the flow measurement uncertainty is currently stated to be 1.7% (not including the 0.1% for feed-water venturi fouling), it could possibly be changed because of the increased inaccuracy of the new Rdf RTDs. However, you state that for DNB events you have used a conservative flow sessurement uncertainty of 2.2% (not including the 0.1% for feedwater fouling).

Is the difference between the 2.2% and 1.7%

uncertainty for flow your main reason for stating that there is no need for re, analysis for DNB transients shich employ the Improved Thermal Design Procedure?

RESPONSE: The value 342000 sps for non DNB transients is the design flow (388,880 spe) sinus the 1.8% flow uncertainty allowance. The value assumed in the LOCA analyses, 377000 spe, was chosen conservatively low in order to preclude the need for future reanalysis resulting from RCS flow related issues. The discussion on page 20 refers to an evaluation of an increased flow uncertainty (2.2%) which was performed because the uncertainty analysis and transient analysis were required to be done concurrently to meet the original schedules. The uncertainty analysis results (Attachment 4)

WESTINGHOUSE PROPRIETARY CL, ASS II demonstrate the existing 1.7% allowance remains adequate. However, the transients reanalyzed assumed a flow rate of 377000 spm in order to allow for an increased flow uncertainty. The impact of the 2.2%

uncertainty on the ITDP calculation of the design limit DNBR was evaluated and found to be minimal.

19 20 (4.2)

For.the uncontrolled RCCA withdrawal from suberitical condition, you state that "This event has been reanalyzed with a reactor coolant flow consistent with the full flow of 377,000 spe (only two reactor coolant pumps are assumed to be operating)." For this assuarption, shouldn't the flow be oc1y about one-half of 377,000 spe? Please explain.

RESPONSE: The uncontrolled RCCA withdrawal from suberitical event was analyzed assuming a flow rate corresponding to the operation of two reactor coolant pumps (377,000 GPH x.46 = 173420).

20 23 (4.2)

For the control rod ejection transient, you state that all the safety criteria are set including the criteria for peak clad temperature less than 2700'F.

Your results tabulated in Table 4.2 appear to be close to the limit for hot standby at the end of-cycle with a value o: 2690'F.

Has a new flow measurement uncertainty resulting from the reduced accuracy of tht' Rdf RTDs been accounted for in the reanalysis 7

RESPONSE

For the control rod ejection events, RCS flows (377000 spm for full power and 377000 gpm x 0.46 = 173420 spe for initiation from suberitical) conservative with respect to the anticipated maximum flow messarement uncertainty (2.2%) were assumed.

21 24 (4.3)

Item 3 in the list of changes in the instrumentation and control of the RTDs states "The modification will signals.

include means to manually reject failed T Identificationoffailedsignalswillbebhgthe same means as before the modification, i.e., existing control board alarms and indications." When there is an indication of a failure of one of the three RfDs in a given loop, will it be corrected within a given period of time?

If so, within what period?

Is there a Technical Specification regarding this?

If two of the three RTDs fail, what steps will be taken?

RESPONSE

See response to question 3.

22 29 30 (5.0)

You state "The need to modify control system set-points will be detensined during the plant startup following installation of the new RTD system by observing control systes behavior." Are these changes expected to be small? Will they affect the Technical Specification requirements?

L--

RESPONSE: Any changes required to optimize control systems are expected to be small. No changes to Technical Specifications are anticipated.

23 B 2-5 In the proposed change in the Bases for TS 2.2.1 for Overtemperature AT you refer to " thermal delays associated with RTDs sounted in thernowells (about 5 seconds)." Is this value of 5 seconds to be checked by testing? Why does this value differ from the 5.5 l

seconds which you proposed to add to footnote (2) of TS Table 3.3-2 on TS page 3/4 3-97 RESPONSE: The value of 5.5 seconds in the footnote and the about 5 seconds given in the Bases refer to the same delay (5.5 seconds). Section 2.1 of the report describes response time testing performed by Westinghouse.

!!.1 It was noted that stressing tests were to be conducted during the second half of 1985 and would provide plant specific data on the nature and stability of the streasing patterns during steady state and load change operations.

Provide the results of these tests.

RESPONSE: The test results are provided in Attachment 2.

!!.2 It was stated that a new calorimetric procedure developed by Westingouse will be used which reduces flow measurement uncertainty because RTDs can sessure temperature difference more accurately than I

1 absolute temperature. Explain the significance of this more fully.

RESPONSE: As noted in the response to question 9, a Delta T measurement to arrive at the precise values of T and T is not necessary.

AbsolutemeasurementsofsufficieUtaccufacyareavailableafter utilization of a multiple point cross calibration procedure for RTD and R/E calibration.

!!.3 The filter time constant is noted to have increased due to hot leg temperature oscillations experienced at McGuire Unit 2.

Explain i

more fully the McGuire Unit 2 oscillations and how the oscillations are considered in the change in filter time constant for the fast response thernowell system. Are these oscillations expected with the proposed RTD system and what testing has been done in this regard?

RESPONSE: The oscillations were discussed in the response to question 1.14.

The despaning provided by the RTD/thernovell and the electronic filter time constant of 3 seconds will continue to provent the oscillations from causing spurious AT channel trips. The cause of the oscillation is believed to be gaps in the hot les nostle/down-comer interface resulting in streams of coolant at cold les tempera-tures within the hot legs.

This phenomenon is not expected to change as a result of the RTD bypass elimination modifications but its impact on indicated hot les temperature may change slightly due to the differences in temperature sampling (3 single points per hot I

les versus the scoop arrangement).

I I

l Response to NRC Question 12 on Catawba RTDBE QUESTION:

The new method of measuring hot leg temperature with thermowell RTDs located in the three scoops is stated to be at least as effective as the existing RTD bypass system, even though the new method measures temperature at only one point within the thermowell.

Discuss what data exists to support the maximtan inferred temperature gradient value presented.

RESPONSE

Ee inferred temperature gradient of (

)+b,c.e was established from evaluation of data from hot leg temperaturc streaming measurements at a 2-loop and a 3-loop plant. We data originated from thennocouples strapped to the outside of the pipe, which provided an indication of the circumferential temperature distribution around the pipo. Demeasurementsindicatedthatgg'g1ximtzn l

l temperature difference across the pipe was betwen [

}

, which for t

thg'g, inch ID pipe is an apparent gradient of (

l

]+

It was concluded from this evaluation that the temperature gradient l

t I

within the pipe would be no more than the gradient soon on the cire m ference. The analysis th.it used thg' gradient to establish the temperature streaming uncertainty incorporated at (

)

margin for the final uncertainty.

l 3e hot leg temperature streaming mear 2rements obtained at McGuire Unit 1, where i

the scoop b'anchline temperatures wort: mgasggggdicatedthatthemaximum Since the [

difference netween any two scoops was ]"'J, the, temperature grg nt measured in this tout la within the maximum gradielt of (

)+

Similar hot leg temerature stre.gggmeasu 'emtnts obtained from (

)

indicate that the internal gradient is smaller.

I i

6-WEST!NGH0VSE CLASS 3 ATTACHMENT 2 HOT LEG TEMPERATURE STREAMING MEASUREMENTS AT MCGUIRE UNIT 1 e

f

I.

l i

NOT LES TEMPERATURE STREAMING MASUREMENTS AT K 6UIRE UNIT 1 4

1.0 1MMI i

l lecluire Units 1 and 2, like most other Westinghouse 4esigned plants, are equipped with the RTO bypass system for the measurement of reactor coolant temperatures. This system was introduced in ig6g to reduce the errors in l

temperature measurement caused by the temrature streaming phenomenon. For

{

the het leg measurement, temperature streaming results from the incomplete mixing of the coolant leaving regions of the reacter core at difforent

{

temperatures, producing a,significant temperature gradient across the hot leg pipe. The RTO bypass system provided a means of determining the approximate i

average temperature in the het les by collecting samples of the coolant from l

three probes, er scoops, inserted into the pipe, and then combining and measuring the total sample flew with an RTO installed in a manifold downstream from the point where the seamles were combined.

l Although the RTO bypass system has been offective in reducing temperature l

measurement errors, the system has introduced significant maintenance and l

availability problems at many plants, including the McGuire units. The piping system has experienced numerous flange and valve leaks and valve failures, and the equipment has become highly radioactive, complicating maintenance on the system and other nearby equipment. To eliminate these problems at the McGuire plants, the RTO bypass system will be removed and replaced with thermswell RTDs installed in the coolant piping. Three therunwell RfDs will be installed j

in each het leg within the sangling scoops employed in the RTO bypass system.

l The average of the three RTO measurements will provide essentially the same i

measurement as the RTS measuring the combined sample flows in the original system.

)

A het leg temperature measurement test program was performed jointly by l

Westinghouse and Duke power personnel at McGuire Unit I to obtain data en l

l l

1 4424e:1d/030386 1

i a

8 A

temperature streaming distributions and stability, and to obtain an indication of the temperatures that will be measured by the thernowell RTDs that will replace the existing RTO bypass measurement system. Since the thernowell RTos l

will be installed within the RTO bypass scoops, the desired temperature measurements were obtained from thermocouples installed on the individual j

l branch lines connecting the scoops to the hot leg RTO manifold. Branch line temperature measurements were[

a., c.

i The test data collection was initiated on June 28, 1985 and continued through November 20, 1985. During this period, the reactor operated for 18 of 22 l

weeks at full power, and more than 40 sets of data at various operating conditions were obtained and analyzed.

Although some drif t in the thermocouple measurements complicated the analysis I

ofthedata,ithasbeenconcludedthat[

e 4

. o.,c,Specific results from l

the program are described below:

c 1

0. s.

l 1

i i

-l l

l

.2 i

4424e:1d/030386 2

6',

A

(-

1 1

1 4424e:1d/030386 I

s,a l

2.0 TEST PROGRAM DESCRIPTION i

2.1 Instrumentation System 1

The instrumentation system consisted of 16 thermocouples and the appropriate mounting fixtures, signal conditioning electronics, data acquistion system (OAS) and the interconnecting cabling. The system was installed in June 1985 in connection with a refueling shutdown. The following paragraphs provide I

additional details on the instrumentation.

t A.

Thermocouples and Mounting Fixtures To provide the desired measurements of the temperature streaming pattern and j

to minimize the impact of a thermocouple channel failure, a total of 16 thermocouples were installed on two adjacent loops (Loop A and Loop 0) in i

pairs at 8 locations, on the outside of each RTO bypass branch line and on the common header. The thermocouples are type K (chromel-elumel) premium grade :

ungrounded thermocouples encased in a 1/4 inch diameter stainless steel sheath. The thermocouple sheath extends about 12 inches beyond the piping insulation and terminates in a standard plug connector. Each thermocouple pair was held tightly against the pipe by two straps wrapped around the pipe.

]

The thermocouples and mounting straps, which are the only parts contacting the j

pipe, are stainless steel. Figure 2-1 illustrates the thermocouple l

installation locations and identifies the instrument channel numbers assigned to the thermocouples.

B.

Reference Junction / Amplifier /DAS l

The instrumentation system is illustrated schematically on Figure 2-2.

The

{

output signals from the 16 thermocouples are transmitted through solid alloy wires of the same thermocouple material. Each pair of wires are twisted and 1

individually shielded to minimize electromagnetic interference and channel l

cross-talk.

4424e:1d/030386 4

,_._.__-.-.,__.-.--_.~.-___-.,,_.v__,,-...----_-mm__--.

.,m___.._ _ _ _ _,, _, _

-__,....-x-

s-The extension wires are input to a thermocouple reference junction control unit located within the reactor containment. This control unit is an Acromag flodel 354 which is a 0*C reference junction that employs a reverse compensating double oven arrangement. The principle of operation of this unit is illustrated in the following schematic diagram.

g..______...

IIsasuring Chromeli A Alueel A BCopper

_ Referenced i

T/C Junction (

Alumel A

Chromel Abooper Output T1 T2 l #J The thermocouple wires are input to the first oven which is precisely controlled at a temperature T significantly above ambient (typically around j

50'C). The thermocouple materials are then reversed, and the reference junction of the second thermocouple is held in an oven which is precisely controlled to a temperature T such that this thermocouple produces an est e 2

that compensates the omf produced by the thermocouple in the first oven to yield a not output equivalent to a 0*C reference.

The millivolt-level output of the reference junction control unit is then input to an amplifier located in the same instrument rack within containment.

There are ten amplifiers grouped together in five pairs of amplifiers within l

an amplifier unit. Two 10-amplifier units are provided, with channels 1-10 connected to one unit and channels 11-16 connected to the other unit.

Channels 1 and 2 share an amplifier pair, channels 3 and 4, etc. The amplifier is a Sensotec SA which is calibrated so that for the range of millivolt inputs corresponding to a Type K thermocou)1e with 0*C reference junction frea 0 - 1000*F the output of the amplifier varies linearly from zero to 1 volt. This high level voltage output is connected to field cables 4

that run through the containment penetrations to the OAS room. The field cables are twisted shielded pairs that minimize electromagnetic interference and channel cross-talk.

4

\\

4424e:1d/030386 5

l l

g.

~

The field cables in the OAS room are connected to devices that facilitate the interface of the field signals and the data acquisition equipment. The primary piece of data acquisition equipment is the data logger, a Monitor Labs Model g302 digital data acquisition system. The data logger acquires the data by a sequential multiplex sampling process. At periodic time intervals set by the user the data logger scans all the data channels sequentially. The sample process takes slightly less than one second per channel, thus the minimum scan interval that can be set is about 15 seconds. Most data were obtained at a 30-second sampling interval. When the data logger scans the channels the value on each channel is presented on a front panel display and printed on a hard copy paper tape. The primary means of data storage used during the tests was with a cassette tape recorder (Columbia Products Model 300 C), which provided an efficient means of data analysis (averaging, statistical analysis, calibration corrections, etc) on an off-line computer system (in this case, at the Westinghouse Monroeville Nuclear Center).

2.2 Data Collection Data were collected at appropriate intervals and at appropriate plant conditions from June to November 1g85. Data typically collected for evaluation included the 16 bypass line thermocouples, the hot and cold leg RTDs and a set of core exit thermocouple measurements.

In general, the procedure for collecting data was to operate the bypass line thermocouple data logger for a period of 30 minutes and concurrently obtain the RTO and core exit thermocouple data. Not and cold leg RTD data were recorded at one minute intervals by the plant data logger. Data were always obtained while plant conditions were stable to minimize the impact of small temperature perturbations.

The majority of the test data was obtained while the plant was operating at full power, to obtain full power temperature streaming data and to determine how the streaming distributions change with time. Although the thermocouples were strapped tightly to the RTD bypass piping, the temperatures measured by 4424e:1d/030386 6

-, _ = _

s, the thermocouples are somewhat lower than the temperature of the water in the pipe due to heat losses through the piping, contact resistance and temperature gradients across the thermocouple itself, and errors in the thermocouple channels. Therefore, the thermocouple measurements were normalized with tests at hot zero power when no temperature streaming is present. Additional test measurements were obtained at 305, 525 and 185 power to determine if 7,c During the first two weeks at full power, several sets of data were obtained to determine if streaming distributions change in the short term.

Subsequently, the data collection interval was increased, and eventually data were obtained on a weekly basis. Af ter about 17 weeks of full power operation, the plant had an unscheduled shutdown, and additional zero power and full power data were obtained subsequently. Data collection was completed on November 20 with an electronics recalibration.

1 4

4424e:1d/030386 7

~

T

=

.r E

W g

i s

a j

l Q

b

~

a 1

g 5

g

(

g )_

E C

E a

i w

l m

N i

a c

s c

E A

a

_J d

R g

p

....n a

w a

C a

i C

3

~

12 2

l1 i

4424e:1d/030386 8

i

?

f

e, MAGNETIC CASSETTE a

i j

TAPE RECORDER C

DATA

~

PAPER ACQUISITION TAPE RECORDER COMPUTER O

01 I m V0LTMETERS o

.i i

DUTSIDE CONTAINMENT YMHHMHMMHHHH/MHHHi INSIDE CONTAINMENT p __ _

__q l

l AMPLIFIERS g

l 1

INSTRLMENT i

CABINET l

3 INSIDE l

C M AI N NT l REFERENCE l

1 JUNCTION I

L _ _.

_ _._l 16 THERM 0 COUPLE SIGNAL CABLES FIGURE 2-2 THERM 0 COUPLE INSTRUMENTATION SYSTEM 4424e:1d/030386 9

~

3.0 TEST RESULTS AND CONCLUSIONS l

l 3.1 Data Reduction 1

As stated in Section 2.2, the thermocouple measurements were lower than the water temperature by varying amounts depending on heat losses and thermocouple i

channel errors, and normalized with measurements obtained at zero power.

The normalization at zero power eliminates all absolute errors in the thermocouple instrument channels. Uncertainties in the thermocouple j

measurements at full power are therefore due only to the uncertainty in the change in temperature from zero to full power. Both the thermocouples and the amplifiers were calibrated over their expected operating range from about 500*F to 650*F and their combined errors were considered in the data analysis.

As reactor power and hot leg temperature increase, the normalization corrections are adjusted to account for increased heat losses as water temperature increases. This adjustment is illustrated on Figure 3-1, and is i

based on'the ratio of the differences between hot leg temperature and containment ambient temperature at power versus at zero power. For the full power cases, this ratio is 1.12, so all of the normalization corrections are multiplied by 1.12 to obtain the appropriate correction at full power. After corrections have been applied, the thermocouple temperatures are compared with the RTO measurements to determine the differences due to the temperature streaming distributions.

3.2 Data Evaluation Based on various comparisons and corrections for drift in the data, the measurements defined a reasonable'and consistent temperature streaming distribution. All of the measurements exhibited some drift over the test period, probably caused by long-term exposure to the containment ambient temperature. Corrections based on the zero power renormalization and recalibration at the end of the program confirmed that the observed temperature changes were due to instrument drif t.

4424e:1d/030386 10

---

r-

-r-v

---

w w-

---

,------ee-,-,r---.w

--w-

.m,ew n

c-

--w--

-. ---ewv-----wwwv,,i-w-

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

= -.

3.3 Igggrature Streamine Distribution l

I Figures 3-2, 3-3 and 3-4 provide an indication of the temperature strsaming distributions during the test period. As mentioned in Section 3.2, the changes occurring after the initial measurements are mostly due _to instrument j

drift. The initial measurements are similar for both loops and

[Ihe final measurements based on the calibration checks indicate that the distribution has changed slightly but is essentially the same as at the beginning of the test period. A best estimate of the average for the two loops is suunarized below:

Branch Line Location Initial Data Final Data

()Jm i

- o. c.'

Top scoops Adjacent scoops Opposite scoops These measurcaents suggest thath

]a..cThe gradient for McGuire is within the limits assumed for the analysis of possible temperature streaming measurement errors.

4 3.4 Core Exit Teneerature j)istribution Sincehetlegtemperaturestreamingresultsfron(

i

}"kortexittemperaturenorswere obtair.ed when the branch line thermocouple and RTO data was obtained to determine if temperature distributions changed during the tect period.

In 4

evaluating this data, it was found that only 24 of the 65 thermocouples i

4A24e:1d/030386 11 s

,-,-n

..,n,, -,, -, -. - -, -., -,., _,.,w,-,

l appeared to provide valid measurements. Many of the thermocouples indicated tesperatures near cold leg temperature, possibly caused by leakage flows through the thermoccuple guide tubes from the reactor vessel upper head. A few thermocouples were not consistent with other thermocouples in other symmetrical locations in the core, or were not consistent with the temperature distribution defined by thermocouples in adjacent locations. The remaining 1

thermocouples did describe a reasonable exit temperature profile across the core. All of the thermocouple indications changed over the 4Hoonth period by varying amounts. The magnitude of the change is summarized on Figure 3-5, which illustrates the core exit temperature distributions measured at the beginning and end of the test period. Of the areas of significant change, h

}"-F 3.5 Bias correction The data was also reviewed to determine the magnitude of the bias correction that will be used if one RTO fails, and the change in the bias with time. For i

this review, the average of the three possible combinations of two thermocouples was compared with the 3-thermocouple averages on Figures 3-6 and 3-7, using the channel 2-4-5 and 10-11-13 combinations.[

i

]"$1thoughthe thermocouple measurements were subject to instrument drift, as discussed in

[

]. A review of the initial and final temperature streaming differences presented in Section 3.3 indicates that the bias corrections changed less than[

.J

.c.

a Based on these testresults,[

- a.,c.

~

l 4424e:1d/030386 12

3.6 Part-Load Doeration During the plant startup in July, data was obtained at 305, 525 and les power to determine the change in temperature streaming distribution versus power 1evel. The branch line temperature streaming differences based on this data l

and the first set of data at full power are illustrated on Figure 3-4. (

} * >c r

1 4424e:1d/030386 13

---

,--,,,---w-

,n--,--4e,--~,

v w-r-v-w--~'

4 e

m

-,._,3_

.4a.

.12i-+.... _,,

as._

y.-

m 2A,.

_ a m

A

_e.

G 6

O., bi l

l i

THIS TABLE IS CONSIDERED PROPRIETARY IN ITS ENTIRITY l

l 1

I I

a i

l TMLE 3-1 m per a N M MTA 14 i

1

-.,--,.y-

_,m__.-_,,,

,,,,_.y.,.,_,,,m,,___..

,,,,.....w, m.-

,_m___.

I

A,C

)

i THIS FIGURE IS CONSIDE8ED P'0PRIETARY IN ITS ENTIPITY e

G e

I e

O M

e.

FIGURE 3-1 THERMOCOUPLE NORMALIZATION ADJUSTMENT e

15

l

~

A,C THIS FIGURE IS CONSInEnED P90PRIETARY IN ITS ENTIPITY e

McGUIRE STREAMING DATA FIGURE 3-2 DIFFERENCES BETWEEN BRANCH LINE T/Cs AND RTD l

16

I A,C l

THIS FIGURE IS CONSInE8ED P90PRIETARY J

IN ITS ENTIPITY l

i i

McGUIRE STREAMING DATA FIGURE 3-3 DIFFERENCES BETWEEN BRANCH LINE T/Cs AND RTD l

17

~

A,C I

THIS FIGURE IS CONSInE8ED P'OP81 UARY IN ITS ENTIPITY i

McGUIRE STREAMING DATA FIGURE 3-4 DIFFERENCES BETWEEN BRANCH LINE T/Cs AND RTD l

l l

i A,C s

THIS FIGURE IS CON 51nEsEn PMP81ETARY I

IN ITS ENTIPITY 1

i

.i i

I FIGURE 3 5 CORE EXIT TEMPERATURE dis m g f

I 19 t

b

~

A,C THIS FIGURE IS CONSIngsED P90PRIETARY IN ITS ENTIPITY l

McGUIRE STREAMING DATA FIGURE 3 6 LOOP A B1AS CORRECTIONS

~

A,C TNIS FIGURE IS CONSInE8En PMPRIETARY IN ITS ENTIPITY I

I l

l

)

McGUIRE STREAMING DA' A FIGURE 3-7 LOOP O BIAS CORRECTIONS 1

4 21

\\

t I

I 1

J,b c THIS FIGURE IS CONSInEDEp ponP91ETARY IN ITS ENTIPITY l

i McGuire Streaming Data FIGURE 3-8 VMgAMON W POWUt M 22 I

I Q'

s WESTINGHOUSE CLASS 3 9

ATTACILMENT 3 RCS N0ZZLE AND THERM 0WELL LOCATIONS 1

f a

i' 9

This attachment consists of a copy of Duke Power Company Drawing Number CN-2680-1. The drawing shows the RCS penetrations (nozzles and thermowells) fer Catawba L' nit 2.

The drawing revisions associated with the installation of the RTD/thermowell system are marked BA.

The locations of the thermowells at the McGuire Nuclear Station will be similar.

The identification numbers for the thermovells are:

Loop A Loop B Loop C Loop D Hot Leg RTDs 1-7 2-6 3-6 4-6 l

(Narrow Range) 1-8 2-8 3-7 4-7 1-9 2-9 3-8 4-8 Cold Leg RTDs 1-6 2-5 3-5 4-5 (Narrow Range) 1-20 2-18 3-19 4-9 Wide Range HL 1-11 2-7 3-9 4-10 Wide Range CL l-5 2-4 3-4 4-4 Note: Wide range RTDs used for indication only.

Protection System and calorimetric procedure utilize narrow range RTDs.

QA CONDITION 1 wm,,, u.

?

::xa.::ct:~s::".*'ND EAR sArtii RtLATit 1 mMntsont d:"US.*,"N='Ell>i:I s-Tl!CY.

",*.***ea.toaesioo set as saipqwons p..,,g, 5,!L ATION for' rents.o t.t ette,n ow.tessen sv omg penui=rs as N m

-- em soen mat.ar one.camioi.oi oo*

emm t.s.n.n se cas.e== us,6 4 NUCLEAR SAFETY RELATED

. s.w.m.,, s 4.-,,,,

c, arv.ssscastssonattu c.mpeam.No :-4.aue<-==. 46>

ovae nowsa sempasse d"* 6

1. 1 CATA*BA NUCLEAR STATCN UNIT 2 l

a sous Lew 6 asuu H cm. n e Jam,.u 4 ers c. -_.masseup p es.umme M

aa n. sa#s. is( M 5

E2TC C n w C.-

ft. w

.i. *.i e s're& _,-

NO2Zl.E 10ENTIFlCATON &

3.se to s.enw weena ame.str - ~- m

-ere et ma A

REACTOR C001. ANT svSTEM

.T

n. +s sa== *=..==s sa.

/

r==

e

A.

. n. v v.,... u

_ - J 2 ""'tfi'"' 'O" - M ene.a oroe x.,ivacri+ e mnavw esvesiens as==,ea{'I'"""

bDWG NO CN 26801 I"el

i 6

9 9

yn e

b d

C C

f h ~h k

,. so.,.

s

.m YN.mN... t

.e..

,, e _.

..w IM'S'H l

MX.

j.

V{lE g

4

, K::'.\\t~'

I

-i '."'".."t,1 g-

'~~

{

f.

t/

f Oi).a%< A

,..b.

".' N.jiy E r m.,s.:

t a.

n 1zy.x

+Wyo >

\\

--e y k~w, ny l -[.y-Mc,gg o M' 'S~,e

!*Fra i

kI/

PT 7

.,, Q

[lu o<<

f e

j f'

Il Up '.'. A, \\

=/8 e

es

jua,

\\. t ;.

.w y =a-.

= -

4 r.

2a s

c,...... s n e ws. @ %,

~

e

/. s.=

w@w 3#

~~-

>,~-

-n g

e-

[n - [,

m'.

.v%;Il 1. u h'..sf tt

( e ; h,*,I,h

! []

N g 'w=,l N

te 9

A4 493-s f

.)

? l

%x a, -

y

1. ~~g$ ?, %,

. 'p a b..x.,s. N. $: $...ft*

.%g,/

r4

/

o

...e 1.

2 'N A _

t'..c. e <h-u

~

• K-+

tetY e

i

~-

',u 1: C rn.-

r nr-

%., s

..i.

<,..%.x* r. g.,

s*.e n*

q,- Ys s;

c,_. /

^

i

,, ')*

i

,Ft", "+,t*-l/

a;.::: T -'

.,t_ 3 gr-l9

.v n q.,

i

-ir i

V de J;

/

" P.

f.

i c....r e*j m..

ne e ne e-1 2,0*

/, ?

Sb A

,-_e_a--

4 A

a-

-a 4-,-

- - as---r

- N,n 7

4 4_5 f \\

sq

+

,.A. 2-o,i., }

i!. ~ M, x N

p-g I

C 3

Zes j

\\' _

~

s h

ox ;

c e N v'OW o

6 e' @--)

(D i'

4,2[.

V'jQ@

/

\\

6if ".' ?

1 I

q y

^"'

.&t f

' _')/

\\ Qy)~e \\.

qs -%g

' pgm tr n W. i f > A,t h W.$ O%1 Q d(('xxgM\\

q-h,r s

C, G 4s4 '

uz 23 rg7w,gt~

~hh b*

W a

, n,, '

y*Jr..g'x

.. i 1

4, Jv N

t vg fb, >d

'W e

• r i

y 7

H\\

"~

5 = [ b-(\\h.

pN

1. f 9

g; v"

4T mi<*

N," "%

e,l<fc 1 -

j

hg&

os W

11

' c.~ p,

/

k.

a s

i S

1 T

a A,) Q

- / #4'

. @ - f. p! l h, 4 e4 ef,tf.g,yg,._j, w%

. s 1i&

x

/G g %. _

r

+n.

y s

e s

N w.

3..

)

2 3'

m y _d,..

7p e. _'_.

s yc e

.c f,

.., r i- -

_ _y W"

s.

,,, \\ p _.f',;*t. i 'y ' g 4

J W

0

-0,Y M

ww 4

k s Y_.. s n. ~ J '

1 1

-' gg/j$

g n

p o

/

L_-m h--

-e

.f 0 009 N f

\\

._ p.p. _

^

c A

.z.-

3

[

.ans.

3 2

I h

(

i

/ r, 'i.

~

.l 5 di

~^

k

8 V3 sf:l

'.s-^

& 0 ces n 1_L}'p 9

q e

s i

y A

C-1

%.g k

uk'n. _\\ N o

r I,

~

x-

,e_,K p

vr 1

N

==

N,k' du Wy s

y

/

W^

7' y

[

'Y f go.

Y b'.h k,

3 y

.u + w s e

il^ M

/,,

w s

~

j-} l&'Q

,.h

'_s?

m UN

(

_l's qk &, k >$

Sh{

'Y s

4

-e 3p !:

i

• M r @3d M,.i g-

.~

m. ~.

p,1-

m... t, f

r f,i=.?,,r--

~.

~

N ~~

{

I III-.

b

$L

.__-,p O./.6, s Y

n N

g8 4,-

a e,

g 3*+

i s

1.

\\@

T O

q

u--

w-i g

3, e

7 e

L J

!!!r -

G 1

2 i

as

< et

g. 3.

e a

6 y

f,[ f o'

k

'\\

N

'~ s, h W +

3g '<S_./ X,'31 1 a,(N

~

i

~.

&'#xA

\\

d [x

~W $b

, t(t g

.TM f

• 3,

G.\\

%.g.+,r,f~,j,%>%f;n/>

n c

~,

9,,@,f Q, e i

/d e

3 A

es i.

i, f.$

Jikma A,*

.. ~g

.x s

4n I

. v..

..?

e j-4 c.

~

s [.

T tj g*

n/

b\\

xy G

x u

e

.! ysy ps :e

n..n

,...y s.e,

.s 0

?"$_

y Q,1

p. ll pf 8

• h lio N

~~ ---

E!'

N y&Tf' c

. _ ~ _ _ _

i P

i 0

M/ TOA? CCx;;t.pC 3Y3 EM MT.LZ IDENTftCAT/ON

4. / 5 7' FMFIMARY LOOP 2-A anec A Tw mgsg roc sogg OeANC_** D99Cnen>h0%

g ry ppy

e.,

ty t geeom ite4CTiese 70 LAO 8 94 TOP 9,08 *CQ.0 &#8 e.g ee*

1 tortTT eau m "MAfot 2.a em T98 ts et - Coka e.St,,,,,,,

e.g 3*

t sectesAe. 4asagtsens apr ygeg 04*CALOL.E&

e..

s' I met SSwaitre SPf as poc for sies.cebo LE C I

e. $

• at tenOwS LL. te c 7w s aes vge g.pg.cgi,e Lgo _

J

e. 6 l'

THE88'sowtLL = geeCAD9e*O Lbc TcP SeOf -CD60 680 F

e.1 s'

TMSEb40vetwL E M SO S* Le I an L EAGf SsOS esOf=6Me e.g i-Teetsadowsww a suggggeto 3 e r.

rcP 5685 *seOf.Lg6 I

s T....

-s c.o ss w,..

.. L a.

e.66 W

. ee? LEG SAes#LE esas witTSs0E* sect tSG fep teOt-no07. LEG r ug gesew tgg,. g nect n este s it e.

sart?v esw to $4FC77 pass.fBuespt GS et*ftes 4 81EJtti *-" -

e-e s 3-E f D Ov sett er?veev stC T6p SsDE *CE8tgels e.e.

F*

Loop Denise 70 waart era es e s.v eses-__T g g see 9 LP Flew teamb T49. tess M toes

'usfrSiest; C E ELL e =e8 W

6 # FLOW TGaast rap te 4syse,

tyggt geg.cengyggg ggg,

[ trit? 90t eGetgett GA '

e s?

4' L P F6ew 78tes ? AP a tasc 8P? teto (Jitt sect <esseesse sea.

e.e 4g*

w e etew reness. f ao see.. t_

a et a.-

utena sens.. sac _. _ _ esem zu

.ts"'a +st.is' ae

_s.no s

so.samM>ssu.. euseena sa ses ner nessa'a r 1

i AN?tMAR** LC09 d=D

&#4NC 8' Of &cAiMsOw d.OCAT/CW aest&4 eso S/fd

'Pe* S.iJs *C&O 686 ~

2-e ag-eseone sasa 70 Lee #

es!

ir...=.C, E g,_,.

j p

u

$aren.~,o. aecue. ore se

.a 34 e

seastuestte great esc

??op tage Ce68 LAEe TWu' 6:05 -eALA LES f

i 7,eg a sac e,gsg. gast Tsw $809 34 3-6 r

TseERG.^4L4. Pu..-^*a

!,8pt TOP SeM COLA LES 55 l'

TMEEDASWE64.* ENLEW44 hee.

'LDG E55T WWE. teor 636

,- -. TOP 8489 &SW el6 TmEWatputta.6 ressTm6590 i

a 51 3-5 f*

T9MWWu.-Ep. Esp 944 0 EBC Tuur 584W a8'WT &#G pd e

v...ese*ww se.c sac wrs r s....e.Or r.

i 8.se ae" pu t 35 ut il.gE egges tr8t r 584N ap4'W 4.#4 lu rkv TO tteE WWespet see st

=

em aprest g6 3 **

e7 E.st E I e

WTO

r. SW tt BrNa es WC w Sist = n _7 a 3 sE 20

F LeeP I P94em 4 twe8TB art.

1. 4

a. mas Geo espa g lide, add we esa, 3.s*

g-n.8 ^< bow vesseg rap. Sanc e?tese

=

tSRF 9am- - M aiaL 3**5 t'

L'F60w Yeagg re e-tasCW'" tees

=.-

+

hSSt tsOS M 36h f

8*$

J 844w T99 eft fpp fear 89toep m

e=

u v

e,w vs.. s v.*..,......

$AV t e4 t*

Tut;- O; SLL. geggggg4=te tes ess vver sesr *1 t

• i 1

A a

l l

l i

PRIM / RY LCCF 2-C

,,,, ;,,,, g s es ? +a j vs.

e < =...

oa m.

4..

e io aceow

.N fo Loop 6

=2 Top SiOE -cotD tro

3. t

=

recs es 6traoww ow foo lio8-co6e we j

s.3 6 as=

saffh sans Ptom ActumWL A r#4 Fm.2 f.

Ist Te8 Sets *Cete e.88 sesf tact-case 4.de_

s..

rwe assewtLA - a=c t= s ee r

s.s t'

Twtowo**w-twcaossso tac foe sect. see.a Ass 3.s e.

Tweeuewsw.- tucao stot tac ig ast gets - seer 644

/

s.,

e-Thatsuowtu aa.cee44eo t&c irop sees - seet see 34 e*

T eet rado ed ELL-sw.umib o s tec.wf 5r 484E - NOT &JIo 900 soet not LEC rnsteoangss.tascs.ueeee s.s

3. o e--

e a ousp6r 70 asse puneet see ms es sues asse.eW3r $s548, S.,

3*

sta tv sess ervuese met 1909 3er.Csesserge 3 it

?*

t0c* CEmiN

• Lf rOOwed set

• 8e f ISOT?oss Saf C80escott east sa0F tMe E4 6 e roow fa nat'. Fe*.264evsome 343 43

• stestsam m esea66 3..e es*

6e#6h t e masg em e. t =c e w,o igests.EN t h etu

.s.o r 4*

L* F 60 w temma 7mO an.ceTsoeo

's,ts? Sst.:G,sst sw g e4 es a psos, fea. sos twensee s s a_ sit ot Lac g am ste som aest,s.5 act 4Jo 3.. e s.e W

emme m a essess'u'en serm 14e re tteae 6 se-ae' 6ee 3 is t-Tweesmowsu. En.scaossio 1 ab t ase *NP 444sr UC

{

e j.

4 PRIMARY LCCP 2-D mess.s m.e l

sigd ennene De s tsin ore

, y,,,,

LO arnord atvan ne rr c.ec..

=v voe nios -co6o a.se i

a 3-a-2 a4*

totow neu vo Loop sel WSr !Lagt tee.g 446 4.s se

• gaatvy isu paoen accuasWaaroe ?t2n as 700 303e - Cc60 L84_

7098 00 - to6 D 6.t e retenscwrts resev ssee 4.s a-vusaueess6L 1meumoescso ac 48 SiSS-te6e 4.d 4

{

4.s twaammeestw. 36 aces stoo Iec top 3:08=esov&s4 i

47 f*

thetaMOwWEi.L the6ED%tet t est e4SF W -esprLES 44 pa teettedowst.L =tWCSOS546 1&&

EstS m.eestLes g

49 t*

tusses,apertu. s,aacso gggo t et, me e w inest"J J" i

vo# S Og*seOf4.84 I

f eeggestwitL facfast9ro e e@

4 40 4*

geogTV n6s asees teryTv ena svenp3 geg v*ap Tp gegy Ams. st l

4 *e2 eF S $7* PASS 48tW4 8e set D8 Sagar f 89988'84 4 63 E'

LOOP OG#itste WAS?t Ft.

toc ader C8DStoest Apr ser CfeW8888 S&L 1

LP Flow fEngel ?ap tiscrtseep

+.e6 Mit amar sier egeneessa sAs 4.s s en*

La Flew teassa Tor.awce'seco hAbt See testafete nu

+.e s gn.

Le atow renset vos.tescet sino wer sier-casosaura su 4.s 7 in~

ea

• Ftow raeus vne roe..

.t E

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

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

---m

4 9

9 31

+

3 3

h.,

{D

.Jh.i l V i.

p

._ w) t :

i@

2

. s s 't i_

(

E i

{

iEd hi d'

,5 h _J> $

]

W

4) [

\\

a go I

I 5

I T

k kp h

giMJ mi

,).% x 3

i

• l

-i i 9{$

y cq i

i 5 's) g$\\

-B i t

t s

j e

O$

s I

k s

g

,E 48-k'/ 5 i[

f YNf k$

E

$}

.}m o't i!@,

e4.,

a!

'4 ej k

-~

1 5 2 1

2 h

.) _ $

~

i.'

3 y

w i

e y

.a n.

i 1

I i

s's.s W

\\

g N,

s a

w

/s 8

j

\\

6

.! 4 l

6 [ j, g l

.j___t-

=.q.r.w 3..

1I g.sc 3...

s g I

j o;

$1 E gjU[j:

e

_I E

jh!

.l i,/ / -i-T._,E_E

', / i_.))/n l

n% "s,1 Q' )

vl

+

\\

~.

/

c 1

- ~

~

s e

!t

\\

e/ ~j d T

8 E

d O

n d

D y

s g

.t CE Nil

(*~~

s g,

'g $[j2 I[5@o, A

i. i n r. c E

S3 I

h k:

! U, U,

a I

5,il

\\

g

~

~

g:, 5 it 5

B

\\

7)@.

-h. O N

6 s\\

M N

'(

... _..._Q,

,b)

>(' (

' s.,

,y s.

7 s,

Y

'\\

b g'

s v._.,..,

1. "a*n g

.s..v.,

g..

j

=

t u

/

Z e e

n.

s s

D b

la t

g

. r y

/

.2 J

\\

n e

n

.s a

Eo3 g H-b e s4

! i.r g:

3; E'1 r

gj r

a..

u,

.;g a

\\ gg.

I

P WESTINGHOUSE CLASS 3 TTACHMENT 4 MCGUIRE RTD BYPASS ELIMINATION UNCERTAINTY CALCULATIONS

Westinghouse has been using an NRC approved approach to the calculation of RCS Flow Calorimetric uncertainties for some time.

The original submittal of the current approach was NS-EPR-2577, 3/31/82, a letter from E. P. Rahe, Jr.,

Westinghouse, to C. H. Berlinger, NRC. This document provided information I

concerning the Westinghouse methodology for the determination of the instrument or measurement uncertainties for Pressurizer Pressure control, Rod j

Control T input, the Daily Power Calorimetric and the Precision RCS Flow AVG Calorimetric. The methodology outlined is applicable to two, three and four loop plants that measure T and T with the use of scoops and the RTD H

C Bypass Manifolds.

In the case of the McGuire plant, it has been decided to 1

remove the RTD Bypass Manifolds and replace them with RTD/Thermowells placed in the modified scoops.

This results in different accuracies for the measurement of temperature, specifically T. With the changes in the H

electronics there is also an opportunity to modify the specific methods for the determination of the uncertainties.

First it should be noted that the RTD Bypass Elimination does not impact the i

calculation of the uncertainties for Pressurizer Pressure control or the Daily Power Calorimetric. Second, more than just the Rod Control system and the i

Precision Flow Calorimetric uncertainties are impacted. Overtemperature Delta-T, Overpower Delta-T and the Low Flow Reactor Trip are also affected.

This document will describe the changes to the calculations necessary to reflect the changes in the plant and the hardware.

4 i

As noted in the basic description of the RTD Bypass Elimination, three RTD/Thermowells will be inserted into the modified scoops in the hot leg, and one RTD/Thermowell will be inserted into the modified scoop in the cold leg.

The three T RTDs feed signals to R/Es which then feed an averaging H

function. The average TH signal is then fed to the protection system for combining with T to determine Delta-T and T Use of all three T C

AVG.

H RTDs allows some reduction in the instrument uncertainty for T f r control H

and protection functions, but use of the R/Es results in an additional uncertainty to be accounted for in the precision measurement (since all three RTDs are used for control and protection purposes it is not possible to use a bridge to measure the RTD resistance value when at power).

4 1

0509v:1o/051187 1

Three equations are noted in NS-EPR-2577 which calculate the uncertainties for the ITDP parameters based on use of a DVM for precision measurement, the process computer for daily parameter measurement and the accuracy of a control system. The descriptions that follow will provide the changes to these

)

equations and their impact on the various ITDP parameters.

Rod Control (TAVG)

The basic equation for a control system is:

CSA = ?(PMA) + (PEA) + (SCA + SD) + (SPE) + (STE) +

(RCA + RD + CA) + (RTE)

(Eq. 1) i For the Rod Control system this equation is modified in two areas,1) multiple

~

RTDs and R/Es for the measurement of T, and 2) the recognized statistical H

independence between the measurement of T and T. This results in the H

C SCA + SD term being modified to:

(SCA + SD) = (?(?SCA + SD ) / N + ?SCA + SD ) ! / 2.0 where N is the number of hot leg RTDs used in the averaging fu.tction.

The RCS term is modified to:

RCA = [

j+"'"

where N is the number of hot leg R/Es used in the averaging, [

]+a c g, the calibration accuracy of the R/E and [

',+"'" is the calibratica

' accuracy of the remainder of the temperature channel process racks.

All other terms of Eq. 1 are unchanged.

The total uncertainty for this systen is noted in Table 1 (identified as Rod Control System Accuracy).

.f i

0609v:1D/051187 2

.,,m

l j

1 l-Precision RCS Flow Calorimetric l

As noted previously the Flow Calorimetric uncertainty is impacted by several areas:

4 1.

multiple RTDs to read T '

H

]

2.

inclusion of the R/E in the reading of T '

H l

3.

RTD cross calibration each refueling on heatup, and

)

j 4.

the determination that the hot leg streaming uncertainty has random and systematic components when used with multiple loops.

The secondary side of the calculation remains unchanged, therefore it will simply be noted as the Secondary Side uncertainty. The basic equation in NS-EPR-2577 for the flow Calorimetric is a

Flow = ({(SEC. SIDE)2 + (T -RTD) +(TH-STRM)

+

H

+ (T -PRESS + T -PRESS)

/ N) U + BIAS (Eq. 2)

(T -RTD)

C H

C where N is the number of loops, T -RTD = SCA + SD + DVM, and H

TC-RTD = SCA + SD + DVM i

The impact of these changes on the inst *ument uncertainties and the Flow uncertainty can be seen on Tables 2, 3 and 4, identified as Flow Calorimetric

. Instrumentation Uncertainties, Flow Calorimetric Sensitivities, and Calorimetric RCS Flow Measurement Uncertainties. As a result of changing the Flow Calorimetric uncertainty, the indicated uncertainty for the Cold Leg Elbow Tap flow also changes due to normalization of the Elbow Tap to the Flow Calorimetric.

The impact is noted on Table 5, identified as Cold Leg Elbow Tap Flow Uncertainty. Finally, the Flow Calorimetric could impact the RCS Low Flow Reactor Trip, this is noted on Table 6, identified as Low Flow Reactor Trip.

i i

i l

l OSO9v:1D/0511s7 3

O i

Several other reactor trips were impacted by the RTD Bypass Elimination, i

Overtemperature Delta-T and Overpower Delta-T. For Overtemperature Delta-T an effort was made to remove conservatism from the calculations, e.g., the convolution of Tu and T RTD errors and the convolution of the R/E errors; c

instead of the uTtra conservative arithmetic summation for the two areas.

When this decision was made it was felt necessary to revalidate the model of the function with regard to what were the driving functions for temperature.

A careful examination of the uncertainty calculation revealed the following' areas of conservatism:

i 1.

the temperature s$gming uncertainty was set at I

]+a,c for the

]

for the cold leg for an average of hot le ng[However,basedoncurrentthinking,thestreaming i

[

uncertainty has been reduced to random and systematic components.

l 2.

the arithmetic summation of the uncertainties for Ty and TC f" SCA and SD, instead of treating Tg and TC as independent j

1 quantities.

e 3.

the arithmetic summation of the R/E uncertainties for TH and T '

C

~

instead of treating T and TC as independent quantities 1

H 4.

The double treatment of R/E uncertainties (one set for T and one setforDelta-T)eventhoughthereisonlyonesetofR/Ngin the r

j circuit.

A logic argument was devised which determined that the model of the function should be changed to reflect T AftersatisfyhonlyorDelta-Tonlyforerrorsforthe RTDs and R/Es.

that premise, a sensitivity study was j

performed to verify the conclusions of the logic argument.

l A review of the standard process block diagrams for the OTDT and OPDT 1

protection functions results in the verification that only one Ty and one i

T RTD feed through corresponding R/Es to two function generatory, one for C

1 T

and one for Delta-T.

Therefore, only one set of uncertainties needs to be$ccountedforintheuncertaintyanalysisifitcanbedeterminedthata y

i limiting set exists. Evaluating the impact of an error in TH or Tc, or 4

both, on T and Delta-T results in the conclusion that, at the extremes fortheune#tainties,onlyoneofthetwoparameters(T or Delta-T) can experience a change. Thetablebelowdemonstratesthishnt:

DEVIATION IMPACT ON i

T T

T Delta-T H

C AVG Max Pos Max Pos Max Delta No Impact j

Mas Neg Max Neg Max Delta No Impact j

Max Pos Max Neg No Impact Max Delta l

Max Neg Max Pos No Impact Max Delta i

Some Pos Some Neg Some Delta Some Delta i

Some Neg Some Pos Some Delta Some Delta i

0509v:10/051187 4

It should be obvious from the table that if both RTDs are off by X'F in the same direction, that Delta-T is not impacted (because the relative difference between the two RTDs remains the same), but that the T value is off by AVG half the error (because of the change in the average of the absolute temperature).

It should also be obvious that if both RTDs are off by X'F in opposite directions, the indicated Delta-T would be in error but the average of the two temperatures remains the same.

This would allow the conclusion that the worst uncertainty for the two conditions would be for the parameter that had the largest total uncertainty. For the intermediate case where the RTD errors are in opposite directions but not at the extreme values (and probably not of the same magnitude) a sensitivity study or an evaluation of the gain factors is necessary. Assuming that all other uncertainties remained constant, the primary differences between Delta-T and T are the T AVG AVG gain factor (K2) and the division by 2.0 to reflect the averaging function.

This then indicates that Delta-T is more significant as long as the value for K2 times the vessel Delta-T remains less than 2.0.

Since K2 times vessel Delta-T is typically in the range of [

]+a,c this conclusion should always be valid. A sensitivity study to verify this point was performed as an additional degree of conservatism. The table below summarizes the results of the sensitivity study where:

Delta-T is the error for a single RTD for Delta-T, TAVERAGE is the error for a single RTD for TAVG, DT is the convoluted sum of the TH and TC RTD errors, T is the convoluted sum AVG of the RTD errors multiplied by K2 times vessel Delta-T and divided by 2.0 to arrive at an average value and CSA is the Channel Statistical Allowance for the function.

0509v:1D/051187 5

i-DELTA-T TAVERAGE DT T

CSA AVG

.(*F)

(*F)

(% span)

(% span)

(% span)

+a,c j

f 1

4 i

1 The first and last rows are where the RTD errors are at the extremes, Delta-T is at it's max and T is 0.0 or vice versa.

Basically, this confirms the AVG logic argument for the first four cases of the previous table. As can easily I

be deduced from the above, the Delta-T error has a larger contribution to the uncertainty of the function than the TAVG error. The same argument can be mado with respect to the errors due to the R/E. Also, from the above table, it can be seen that the worst case condition for the errors for Delta-T results in the largest CSA value which indicates that use of the extreme value for the uncertainty for the RTD in a Delta-T function will be conservative.

I Based on these conclusions, a new model for the OTDT uncertainty analysis was devised with the following features:

1 1.

a PMA for Delta-T based on the hot leg streaming uncertainty instead of the Daily Power Calorimetric, j

1 l

2.

SCA and SD terms for an RTD to be used in a convolution of RTD errors I

for TH and T, assuming N RTDs for the measurement of T I""

C H

Delta-T I

1 j-3.

deletion of RTD and R/E uncertainties for Tgyg, i

l 4

0509v:1D/051187 6

i

4.

convolution of R/E errors for T and T, assuming N RTDs for,the H

C measurement of T for Delta-T, g

5.

addition of seismic allowance terms for the function generator cards for the Delta-I penalty function and Pressurizer Pressure, 6.

addition of a bias term for Pressurizer Pressure to account for thermal non-repeatability if a Barton transmitter is used, 7.

calculation of Delta-T span as a function of power (since that is the way the channel is scaled for a majority of the plants)

Overpower Delta-T

~

The basic methodology of error combination is the same as that noted for Overtemperature Delta-T, i.e., it is based on the assumption that T is AVG the driving temperature function. However the revised version is based on the determination, through the Overtemperature Delta-T logic and sensitivity study, that Delta-T is the driving function. A separate sensitivity study for Overpower Delta-T was not determined to be necessary because the value for the temperature gain for the TAVG channel is even smaller for Overpower Delta-T than it is for Overtemperature Delta-T [

3+a,c It was therefore concluded that a similar treatment of the instrument uncertainties war '; eptable. The following refinements and changes were made to the calculational model:

I 1.

a PMA for Delta-T based on the hot leg streaming uncertainty instead of the Daily Power Calorimetric, 2.

SCA and SD terms for an RTD to be used in a convolution of RTD errors for T and T, assuming N RTDs for the measurement of T I#

H C

H Delta-T 3.

deletion of RTD and R/E uncertainties for TAVG 0509v:10/051187 7

I 3

4 4.

convolution cf R/E errors for T and T, assuming N RTDs for the H

C

]

measurement of T f r Delta-T, H

I 5.

calculation of Delta-T span as a function of power (since that is the way most plants have the channel scaled).

r l

Finally, for completeness, McGuire has changed some of the transmitters since i

i the first calculation of the ITDP instrument uncertainties. The Daily Power Calorimetric uncertainty was redetermined based on the information supplied by l

the plant. The results of the calculations are noted on Tables 9, 10 and 11, identified as: Power Calorimetric Instrumentation Uncertainties, Power I

Calorimetric Sensitivities, and Secondary Side Power Calorimetric Measurement Uncertainties. The net result of all these changes in plant hardware and analysis methodology are revised trip response times and Allowable Values for the affected reactor trips in the plant's Technical Specifications.

l I

~

i i

j i

A l

i 1

)

i i

i l

4 1

)

osoev:to/os11s7 8

i 1

... ~.

-. ~

TABLE 1 ROD CONTROL SYSTEM ACCURACY T

TURB PRES j

AVG

+a,c PMA

=

SCA

=

M&TE =

j STE

=

1 SD

=

BIAS =

p RCA

=

M&TE =

i M&TE =

~

i RTE

=

RD

=

CA

=

l BIAS =

+a,c ELECTRONICS CSA

=

I ELECTRONICS SIGMA =

CONTROLLER SIGMA

=

CONTROLLER CSA

=

i l

i

'i i

0509v:10/051187 9

4

,,__,____,,,.,-,.-.___n,-,,..,

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

7.

TASLE 2 FLOW CALOR 1 METRIC INSTRUMENTATION UNCERTAINTIES 1

(5 SPAN)

FW TEMP FW PRES FW s/p STM PRESS Tee TE PRZ PRgSS I

+s$

SCA e

~

M& tee i

gpE e STE e SD e i

n/E =

RDOT=

I SIAS=

i CSA e w

e CF INST USED 3

1 4**

I eF pela 5 W/P pela

• F

• F psia i

INST SPAN = S00.

2000.

120.

1900.

100.

100.

800.

IN$T UNC.

gc.

I

( RAPCOM ) =

INST UNC.

(SIAS) e m

NOMINAL

=

440.

1100.

1000.

618.0 899.2 3250.

[ Thermal non-repeatabl18ty of Sorten transmitter, treated as a b l e s.1**A Number of Het Les and Cold Les RfDs weed fee measurement in each leep and the number of Pressweiser Pressure transmittees usad j

evere11e I.e., one per,leep.

1 8 9 l

i e'

m f

i l

10 4

---,,---,,-.,---w-m a-e,,n-v,w-.,,--,,---__---,-,-,--,.-._,-_-,---,-.,w-w-e---- - - - - ~ ~ - -.. ~ ~ -,

i TABLE 3 FLOW CALORIMETRIC SENSITIVITIES FEEDWATER FLOW Fe

+.,c MATERIAL e

4 j

DENSITY TEMPERATURE

=

l PRESSURE

=

=

4 DELTA P FEEDWATER ENTHALPY TEMPERATURE e

PRESSURE

=

• 1192.9 STU/LS~M he he e

419.5 STU/LSM DetSG)

=

773.4 9TU/LOM STEAM ENTHALPY

+,c PRESSURE

=

MCISTURE e

NOT LEG ENTHALPY TEMPERATURE o

PRESSURE

=

689.9 BTU /LSM hw e

he e

588.7 BTU /LSM 81.2 STU/L9M DntVES$)

e 1.546 BTU /LBM-sF CetTw) e COLD LEG ENTHALPY g

(

TEMPERATURE

=

PRESSURE a

~

CetTa) e 1.36T STU/LEM-*F COLD LEG SPECIFIC VOLUE

+*,C

~

TEMPERATURE e

PRESSURE e

I 4

l j

J i

f 11

\\

{

,}

i,.

l I.

TABLE 4 i

CALOR 3 METRIC RCS FLOW MASUREMENT WCERTAINTIES t

?

COMPONENT INSTRLMENT ERROR FLOW UNCERTAINTY 3

A

I FEE 0 WATER FLOW VENTURI T W RMAL EMPANSION COEFFICIENT TEMPERAT WE MATERIAL i

DENSITY TEMPERATURE PRESSURE l

DELTA P

]

FEEOWATER ENTMALPY TEMPERATURE PRESSURE STEAM ENTMALPY PRESSURE i

MOISTURE KT PLMP MAT ADDITION j

MOT LEG ENTMALPY i

TEMPERATURE STREAMING. RANDOM l

STREAMING. SYSTEMATIC PRESSURE I

COLO LEG ENTNALPY j

TEMPERATURE l

PRESSURE COLD LEG SPECIFIC VOLLME.

TEMPERATWE PRESSURE i

RfD CROSS-CAL SYSTEMATIC ALLOWANCE

}

SIAS VALUES l

FEEOWATER PRESSURE DENSITY ENTMALPY l

l STEAM PMESSuRE ENTMALPY PRESSUR3xER PRESSURE ENTHALPY = MOT LEs ENTMALPY = COLD LEO i

SPECIFIC VOLUME = COLD LEG I

FLOW BI AS TOTAL VALLE

)

• • , *, ** 19CICATE SETS OF DEPENDENT PARAPgTERS i

a +m,t l:

SINGLE LOOP UNCERTAINTY (W3TNOUT SI AS VALUES)

}

N LOOP UNCERTAINTY (WITMOUT SI AS VALUES)

N f.00P UNCERTAINTY (WITN SIAS WALW S)

!l 12

.=_-

TABLE 5 COLO LEG ELS0W TAP FLOW UNCERTAINTY INSTRUMENT UNCERTAINTIES J

j N W/p SPAN 5 FLOW f

PMA e PEA =

SCA =

SPE a STE

• SO e RCA =

M& tee RTE =

l RO e i

30 m A/O =

ROOT =

SIAS=

FLOW CALORIM. BIAS =

FLOW CALOR! METRIC

=

i l

'l INSTRLMENT SPAN

=

• m d.

SINGLE LOOP EL90W TAP FLOW UNC =

S FLOW

~

j N LOO

• EL90W TAP FLOW UNC e

l N LOOP RCS FLOW UNCERTAINTY (WITHout DI AS VALUES)

=

N LOOP RCS FLOW UNCERTAINTY (WITH BIAS VALUES) m e

I i

I e

i 13 i

l'

1

,I TAtti 6 LOW FLOW RCACTOR TRIP 5 DP SPAN 5 FLOW SPAN 1

q.c PMA1

=

PMA2

=

l PEA

=

SCA

=

SPI

=

STE

=

50

=

BIA5F=

s:Asi=

BIA52=

i f

RCA

=

MTE =

l RC$A =

RTE

=

RD

=

S!AS =

w t'

j FLOW SPAN

= 120.05 FLOW SAFETY ANALYS!$ LIMIT = [

)" ' '

ALLOWAllt VALUE

= 88.05 FLOW f

, EXIMUM VALUE

=L

]"

MOMINAL TRIP SETPOINT = 90.05 FLOW 1" '"

Aa[

)" '"

5 = 0.60 1 = 1.37

(

. ]" ' "

TA = 2.5

[

I 1

l f

i 14 3453e:1d/110185

_=.

TABLE 7 i

OVERTEMPERATURE DELTA-T TRIP Tgg NtSS WA-1 StLTA-T g

M.c PHA =

SCA =

i MTE=

STE =

50

=

l BIAS =

1

]

RCA =

MTE=

M,t=

i RCSA=

I RTt =

RD

=

j sA

=

l INSTRUMENT SPAN

= 88.2 OtGF 1

SAFtTV ANALYS!$ LIMIT = [

j" ' "

t 1

ALLOWABLE VALUE

= 2.435 DELTA-T SPAN i

MKIMUM VALut

=[

}" '"

NOMINAL TRIP SETPOINTS El = 1.2000 K3 = 0.001095 1

VESSEL DELTA-T = 58.8 9tGF DELTA-1 SAIN = 1.50 n essUtt u iN = t 1" ' '

]" ' "

1 A=[

)" '"

5 = 2.58 1 = 5.43

[

)" ' "

i TA = 8.4

(

\\

e I

15 36sse:1d/1101 5

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

!s S

I TASLE 8 1

l OVERPOWER DELTA-T TRIP l

SELTA-T Tgy;

+e.c peta =

SCA =

50

=

SIAS=

i I

RCA =

MTE=

MTE=

i l

RCSA=

RTE =

I.

RD

=

=

INSTRUMENT SPAN

= 88.2 DEGF SAFETY ANALYSIS LIMIT = (

]" ' "

ALLOWABLE VALUE

= 2.T85 DELTA-T SPAN IIARIMUM VALUE

=[

}" ' "

WOMINAL TRIP SETP01NT K4 = 1.0900 I

VESSEL DELTA-T = St.O DEGF 3" ' "

A=(

)**'"

5 = 1.T1 I = 1.24

(

)

TA = 4.9

(

l l

1 l

1 36$3e:1d/110105 16 i

TABLE 9 POWER CALOR! METRIC INSTRUMENTATION UNCERTAINTIES

\\

(5 SPAN)

FW TEMP FW PRES FW DP STM PRE 55 e,c j

SCA

=

MTE =

SPE

=

STE

=

50

=

SIAS =

RCA

=

MTL =

RTE

=

80

=

10

=

A/D

=

CSA

=

INST SPAN =

l'.if a NC (RANDON) =

INST UNC (BIAS)

=

l i

NOMINAL

=

f i

J!

3653e:1d/103185 17 i

,,,,,..,..-.--__.,_.---,,--,--,-,---n-..-.--.--,..--.-----.------w

.)

1 TABLE 10 i

POWER CALORIMITRIC SENSI11VITits l

FEEOWATER FLOW

,,,e l

FA TEMPERATURE

=

MATERIAL

=

SEN517Y

]

TEMPERATURE

=

PRES 5URE

=

SELTA P

=

I FEE 0 WATER ENTHALPY TEMPERATURE

=

PRESSURE

=

l j

j H5

=

MF

=

DH5G

=

STEAM ENTHALPY PRES 5URE

=

M01STURE

=,

i 4

J 1

I I

18

.i 3453e:1d/1031tl

,e O

TABLE 11 SECONDARY SIDE POWit CALORIMEft!C NEASUttMENT W CERTAINTits CORP 0NENT INSTRUMENT ERROR POWit W CERTAINTY l

+4.t PLE0WAftt FLOW 5 POWit VENTull TNttMAL ERPAN510N COEFFICitWT TEMPitATutt MATERIAL DEN 51TY TEMPERATutt l

PRES $Utt I

DELTA P Fit 0 WATER ENTHALPY TEMPERAfutt PRES $URE F

STEAM ENTHALPT Ptt$$Utt N015Tutt Ntf PUNP MEAT ADDITION S!A5 VALUES Fit 0WATit DELTA P Ftt0 WATER Pats 50tt St#51TY ENTMALPY STEAN PRE 55Utt ENTMALPY POWtt t!AS TOTAL VALUE INDICATE SETS OF StPENDENT PARAMtitt!

f 5 POWit 51#4LE LOOP WCitTAINTY WITNOUT O!AS VALMS) 5 POWER u LOOP WCitTAINTY WITNOUT 01A5 VALUES) 5 POWit N LOOP UNCERTA!NTY (WITN SIAS VALutl) 19 34Sle 1d/1031tl 6