ML20136F038

Forwards Addl Info Re Reactor Coolant Inventory Trending Sys Design Responses,Per NUREG-0737,II.F.2.Info Submitted as Suppl to 840131 & 0706 Ltrs & in Response to NRC 830614 & 850912 Requests

ML20136F038
Person / Time
Site:	Three Mile Island
Issue date:	12/31/1985
From:	Hukill H GENERAL PUBLIC UTILITIES CORP.
To:	Stolz J Office of Nuclear Reactor Regulation
References
RTR-NUREG-0737, RTR-NUREG-737, TASK-2.F.2, TASK-TM 5211-85-2206, NUDOCS 8601070190
Download: ML20136F038 (26)

Text

-

GPU Nuclear Corporation Nuclear o a::erdie:r8o s

Middletown, Pennsylvania 17057 0191 717 944 7621 TELEX 84 2386 Writer's Direct Dial Number December 31, 1985 5211-85-2206 Office of Nuclear Reactor Regulation Attn: J. F. Stolz, Chief Operating Reactors Branch No. 4 U. S. Nuclear Regulatory Commission Washington, D. C. 20555

Dear Mr. Stolz:

Three Mile Island Nuclear Station Unit 1 (TMI-1)

Operating License No. DPR-50 Docket No. 50-289 RCITS Design Responses (NUREG 0737 II.F.2)

In response to your questions of June 14,1983 (Enclosure 1), and September 12,1985 (Enclosure 2), enclosed please find our conpleted response which references inforation provided in GPUN letters dated January 31, 1984, and July 6,1984. The Reactor Coolant Inventory Trending System (RCITS) is installed and has been tested.

Sincerely,

. D. ill Director, TMI-l HDH/LWH/gpa:0431 A cc: R. Conte J. Thom Attachments i

.' I -h0 8

8601070190 DR 851231 -

I ADOCK 05000289 PDR GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation

r Enclosure 1 Responses to NRC Questions of June 14, 1983

r Question 1. Provide an analysis of the expected errors in the hot leg and reactor vessel head level measurements. This analysis should include not only an overall estimate of the measurement uncertainty, but estimates of each contributing factor, i.e.,

temperature of the inpulse lines, common mode pressure effects on the differential pressure transducer, and uncertainties associated with the transducer. Explain how the individual errors are combined to give an estimate of the overall error.

Response: Attachment 1 entitled "TMI-l RCITS Loop Error Calculation Sumnary" provides a summry of GPUN Calculation C-1101-662-5350-023 which is available for review at the GPUN Corporate Offices.

The nornal and accident errors are based upon two separate error calculations: (1) Instrument loop errors and (2) errors due to temperature distribution in the RCS. The instrument loop error is random error, + 13.88%

for either hot leg or vessel, which will likely be reduced as RB condTtions become less severe. The temperature error -9.6% and -9.9% (hot leg + vessel respectively), which is reduced as time elapses from reactor trip.

An estinate of the overall error could be stated as +4.28% to -23.48% and

+3.98% to -23.78% (hot leg + vessel respectively).

Question 2. Suppose an impulse line on one hot leg was broken that would tend to drive the d/p transducer full scale. How would this condition be detected? When would a level be detected in the other leg?

Response

All four of the RCITS transmitters are set up as follows: If the reference leg to the top of the hot leg would fail, the corresponding transmitter would fail high. These effects can only be detected by indirect means (e.g.,

leakage, nakeup). There would be no effect on the other 3 transmitters.

Question 3. Provide specifications for the proposed d/p transducer.

Response: Please refer to Attachnent 2. Note that appropriate errors are described in Attachment 1.

Question 4. Provide an analysis to show the effects of flashing or dissolved gases in the impulse lines.

Response

A. SATURATION CONDITIONS IN THE IMPULSE LINES:

Under conditions of nornal plant operation, a simplified heat transfer analysis performed on the impulse lines of the reactor vessel and hot leg level instrunentation will show that the RCS temperature is dissipated over relatively short lengths of impulse piping. Tne result will suggest that the thermodynamic and transport properties of Water in the impulse legs are primrily functions of the local ambient temperature. At the pressure differential transducer, the hydrostatic pressure in the high pressure and reference ports can be shown to be related to local static pressure in the RCS, the ambient temperature, and the vertical distance between each instrunent port and its corresp,nding system tap.

Consequently, under the conditions ci a SB LOCA, there is the likelihood that the first several inches of water in the hot leg level reference leg my become saturated. Obviously, this depends on conditions in the RCS.

If the first several inches of water in that leg reach the vapor phase as the RCS pressure continues to drop, the hot leg level instrument my reflect a slightly higher level than actual.

Further saturation of water in the reference impulse leg my occur if ambient temperature reaches that value where a local value of hydrostatic pressure in the line corresponds to the saturation pressure at that temperature. Under these conditions, the instrument my report a reading which is biased high. Saturation conditions in the high pressure impulse line inplies large break LOCA. (Note: The RCITS is not designed to be operational for a LBLOCA.)

B. UNDISSOLVED / DISSOLVED GASES IN IMPULSE LINES:

In water, it can be shown that the solubility of gases are primrily a strong function of temperature. Should any undissolved gases appear in the impulse lines, the following will be trua:

1. The density of the entrained gas will be much less than the density of the liquid which surrounds it.

2. A buoyant force will be imposed on the gas bubble by the surrounding water. The force will be proportional to the bubble volume and the difference in density between the two fluids.

3. The unbalanced buoyant force '.4f11 cause the bubble to attempt to migrate toward the highest part of the inpulse line.

With regard to undissolved gas entrainment in water filled impulse lines:

1. Entrained gas will migrate to the highest point in the inpulse line as long as its motion is not impeded. It will not pass through horizontal legs and tends to distribute itself in the upper region along the horizontal.

2. Entrained gas tends to become trapped in the impulse line at small radius (tight) 90* or greater departures from the vertical. The gas will generally tend to distribute itself along the bend. Some gas nay be collected in the transmitter during a transient.

n

-3. The net effect of gas bubble entrainment on instrument reading during short lived pressure perturbation in the system would tena to reduce the system damping coefficient and produce an oscillatory reading around the actual due to compressibility effects of the gas. As soon as the pressure perturbations cease, the instrument should return to the correct reading '(if RCS pressure is returned to 2155 psig).

At TMI-1, the hot._ leg level and reactor vessel instrumentation will be installed with a nanual gas venting system. Instrumentation impulse lines will be installed per specifications which require a minimum slope. Fill and

-Vent on transmitters and sensing lines that have had mintenance performed on them.or' exhibit abnornal behavior will occur as part of' the instrument naintenance program. (Note that the transmitter sensing lines enter the bottom of.the transmitter and nake the transmitter a high point.)

Under a ' scenario of RCS LOCA where the reference impulse leg of the hot leg level instrunent is uncovered and a gas bubble is produced in that line, the undissolved gas is not expected to be a najor problem for the following reasons:.

1. ~ The ' system is designed and will be installed to eliminats undissolved gas as described above.

2. If snall amounts of gas do become lodged in the inpulse lines as mentioned previously, the instrumentation should produce a reasonably accurate reading (trending).

, -3. Redundant instrumentation-(one level instrument per 1oop and two level instrunents for the RV head) allows a means of comparison.

Question 5. Discuss the ability of the transmitters to' withstand a LOCA environment within the containment and be available for post-accident monitoring - consider the loss of the pressurizer transmitters in the TMI-2 accident in this' discussion.

Response: The pressure transmitters are Foxboro model N-E13DH which are qualified for LOCA environments.

The transmitters ability is to survive the LOCA scenario is amply demonstrated-by the transmitter qualification to the more severe accident per Attachment 2.

The pressurizer level transmitters (RCl-LT l&2-3ailey By 3B40X-and LT777

'Rosemount 11530D5 are environmentally qualified in accordance with 10CFR 50.49 as are the Foxboro- transmitters.

Question 6. On page 6, " Temperature element will be installed in the reference leg, ... to correct for water density as a function of water tempera ture. . . . ".

Provide an analysis of the error that would be expected both with and without' the temperature compensation.

Response: Since the TMI-l RCITS design uses temperature compensation, that is the only error analysis provided. (See the response to Question 1.)

Question 7. On page 7, "A ' caution' sign will be mounted by the inventory indication stating that readings are valid only when the reactor coolant pumps are idle."

Describe the location of an indication of the state of the reactor pumps with respect to the location of the inventory readouts in the control rooo. Does " idle" mean pumps are off?

Response: As indicated in the System Design Description (Attachment 1 to our January 31, 1984 submittal) the level display will be via the plant conputer on the computer screen in the control room.

The display will indicate null display if RCPs are operating.

Idle does mean that the RC pumps are off. Therefore a caution sign will not be put up.

Question 8. On page 8, "A ' caution' sign will be mounted by (the void fraction)... indicators stating that readings are valid only when the reactor coolant pumps are in operation."

Describe the location of an indication of the state of the reactor pumps with respect to the location of the void fraction readouts in the control room.

Response: As indicated in the System Design Description (Attachment 1 of our January 31, 1984 submittal) the void fraction display will be via the plant computer on the computer screen in the control room. The display will indicate null display when RC pumps are not operating.

Questfor.9. On page 9, the void fraction neasurement is stated to be more important for trending than absolute accuracy. In this case, is the trend available on a chart recorder for reference by the operator.

Response: Trending information is displayed via the plant computer as described in Attachment 1 of our letter of July 6,1984.

During nornal operation data points are plotted every 6 minutes and during accident conditions (post trip every 30 seconds).

The resultant plot displays infornation over a period of 150 minutes and is historically retrievable on tape for long term  !

storage.

Question 10. The analyses presented for SBLOCAs are for 0.01 square foot breaks. Describe the range of breaks fc which the RCITS may be useful to the operator. Specifically, what is the largest sized break for which the transient proceeds slowly enough for operator action? What is the area of a stuck open PORV?

r Response: This reply is based upon a 10 minute estinted response time interval for the operator:

1. To be alerted to the fact that a problem exists.

2. To diagnose the nature of the problem as a LOCA.

3. To decide and implement a course of action in support of the actions of plant safety systems in order to bring the plant to a safe shutdown.

RCITS my be useful to the operator for the following:

1. With the RCS pumps running, for assessment of the thermodynamic conditions of the RCS inventory.

2. With the RCS pumps tripped, for RCS inventory tracking; from any snall break size up to an estimted maximum break size of 0.15 f t.2, The largest estimted break size for which the transient proceeds slowly enough for operator action is 0.15 f t2 ,

In addition, the operator nay use the mixture height vs. time history for example, to calculate an order of magnitude of system leak rate.

The bore diameter of a PORY is 1.094 inches (.006525 f t.2),

Operation 11. On page ll, " ... fuel temperature rise to th6 point of fuel da nage". Discuss the range of possible values of clad temperatures, which you consider to be indicative of fuel danage.

Response: Fuel damage is considered to occur when the fuel assembly is subjected to temperatures at which the naterial properties of the fuel could be altered so that the fuel's perfornance characteristics are degraded beyond original design parameters and thus, rendered unusable. The onset of fuel danage nay be considered to occur when the fuel cladding reaches temperatures of about 800*F. For all temperatures above this value, the fuel would have to be re-evaluated to determined usability.

Question 12. ' Discuss the ta; in the decay heat drop line from the standpoint of a single failure of a line leading to this tap.

Response: Redundancy is not compromised by having a shared tap since it is highly unlikely that the tap will fail either from plugging or breaking. Freedom from plugging is enhanced by, (1) use of stainless steel connections which preclude corrosion products

~

and, (2) absence of mechanism, such as chemical solidification of boric acid which precludes buildup of entrained material.

It is also unlikely that the tap will break because it is in a area which is physically isolated from physical damage. It should also be pointed out that in other cases where sharing of a tap occurs in the RCS, we know of no prior experience reporting deleterious malfunctions of the shared tap. In the unlikely event the shared tap does fail, it should be recognized that RCITS is not a Protection System initiating automatic action, but a monitoring system with adequate and redundant backup monitoring such as by core exit thermocouples for operator correlation.

Question 13. How many of the level d/p transducers will be conducted to a single tap?

Response: As shown in Figure 1 of our July 6,1984 submittal, d/p transmitters are connected to single taps as follows:

Head Tap 2 level d/p transmitters Loop Tap 1 level d/p transmitters per loop

. Decay Heat Drop Line Tap 4 level d/p transmitter Attachment 1 TMI-1 RCITS LOOP ERR 0" CALCULATION

SUMMARY

' 0BJECTIVE To determine the instrument loop errors for the RCITS reactor vessel level, and hot leg level that are output to the plant computer. Find the error for normal and accident conditions.

REFERENCES

1. Nonlinear neasurerent of water and steam in a pressurized vessel. Koska &

Rednond, the Foxboro Company, Foxboro, Mass.

2. USNRC Meno - Peter S. Kapotow 8/23/82 page A6 NUREG-0737 analytical solutions to two problems pertinent to Items II.F.1.4,5,6, a statistical treatment of deadband and hystersis errors.

ASSUMPTIONS

1. Unless othervise stated, vendor published accuracy data includes the combined effects of linearity, hystersis, deadband, and repeatability as stated in Staniard ISA-551.1959.

2. Unless otherwise stated vendor published accuracy data represents 3 Signa 30 values and can be converted to 2a values by multiplying by 2/3 as stated in Reference 2.

~3. a. Hornal Conditions are:

RC Temp. at 560*F, Ref. Leg temperatures at 100*F

b. Accident conditions are:

RC Tenp. at 650*F Ref. Leg temperatures at 250*F Radiation at 2x107 RAD TID CONCLUSIONS Reactor Vessel Level Instr. Loop Error Non accident = + 6.76% or + 9.46" level Accident = 113788% or + 17.44" level Hot Leg Level Instr. Loop Error Non accident = + 6.16% or + 40 56" level Accident = + 13788% or + 87.28% level

1 DISCUSSION

- The Reactor Coolant Inventory Trending System (RCITS) provides a neans for the control room operator to monitor the water inventory in the hot leg of each

primary loop and in the reactor vessel when the RC pumps are not operating.

Hot leg water. inventory is measured by a differential pressure transmitter

'with taps at the decay heat drop line and a reference leg from the vent

, connection at the top of the hog leg. Reactor vessel inventory is neasured by a separate differential pressure transmitter with taps at the same decay heat

' drop line and its reference leg from the Reactor Vessel head vent line. See t figure 1.

l Process -instrumentation is provided in an' environmentally controlled area to

. correct 'the water and steam density from pressure. This provides a more

. . accurate level reasurenent than straight line approximation. See Reference 1 for full explanation of .the nethod.

For.each loop of level indication, the process instrumentation is designed to solve the following equation:

7 .

.A P = hGw + (H-h) Gs-HGo

= H(Gw-Gs) + H(Gs-Go)

Solving for h from the above 4

h= A P-H(Gs-Go)

Gw-Gs where:

A P = Dif. pressure at transmitter h = Level height above bottom tap Gs = Specific gravity of steam in the vessel Gw = Specific gravity of water in the vessel Go = Specific gravity of water in ref. leg & sensing lines H = Reference leg height above bottom tap where H = 600" for-hot legs H =.140" for Rx vessel Figure 2 provides a flow diagram of the process instrumentation for one of the two redundant instrunent loops. Error calculations are referenced to Figure 2. The loop error is derived as the two signa (2a) standard deviation, and is calculated using the statistical technique of finding the square root of. the sum of the . square of each of the random variable errors, each expressed at the 2 ovalue.

The error of.each module changes with the gain of the next module in the instrument loop. Therefore, the gain of each module is included in the calculation. Error values for each nodule are related to the span of the module output, 'therefore some method is required to relate all individual errors to the final loop output which is in ' inches of level'. This is acconplished by expressing all input & output values in voltage, normalized

1. . between 0 & 10V a; O to 10.

RCOT RCOT JL JL

[RCV488 [RCV48A RCV5IS RCV518 RCV51A RCV52A

@ RCV400 RCV48A@

N Dic ' Vani VanT Dic WLO. LO.

LO. LO. ty RCV158 LO. LO. LO. LO. ARCV15A LO. W >4C40 LO.

[h RCv4 - -

me .co- ,.' l RCV58 ( '

I I

LC.

f \

i LO.YR'.V58 I I ll

"" ' l

~ LO.fRCV55 \

~

W 1P W I

1 Y ,

y W . / r 7 y o

. , ,s 1e u .-

,SEL LEG

,0, iI VES OHVI RANGES LO.

' ~

LO.

\ a i N N INN [LO. )

l t

I

%/  %/

Figure 1 Reactor Coolant Inventory Tracking System

9

& 1 7 +W ~

+  !>

ec<)$,

+ w hr a

4 -

4 -4)

• &5 8@

b $

$}

u J -

lto 9 ii 9

i usa .a . >

. e3 - -

$ v 8 ., b- a .s j

I 5;

3]l, 7 e;

'4 i.

21o

1. g# 4 13 $@1 a eEE

N-E13 SERIES NUCLEAR - -

ELECTRONIC PRESSURE TRANSMITTERS These transmitters measure absolute pressure, gauge pressure, or differential 2 wire transmitters that have been type tested for applica ,

Plants per IEEE 3231974 and 3441975 requirements. radiation OPERATING CONDITIONS influence Reference Operating Conditions Normal Operating Operative Limits Amt ent Temcerature Conditions Limits (OsE) (DsE) 24 :2*C (Teoworks with AMDhfer) (75 2 3*F) 0 ana 80*C {

(32 and 180*F) -18 and - 216*C o Process Terr :erature (0 and 420'F) l 24 22*C r 75 : 3*F) 0 arc 80*C (32 ano 180*F) -18 anc + 120*C

• Relative Hum @ty <0 anc 250'F) 50 = 10 %

Amorent Pressure NoLet 101 *Pa (t4 7 csi) No L m t iArmesenenc; 94 and 108 APa 42CApa Veerat,on anc Seismic (13 7 and 15 7 csi) i60 csa Negng cie

("g" lead in any clane) Per CBE Test Profile )

(see Figure 1) Per SSE Test P*ofde Poston- (see F:qure 2) vor gi 1 0 to 90* tilt in any clar's Rad!ation Rate fGamma) Negng ble No Let Total integratec Dose (710) 0 2 rad!h Negng;o,e 3 x '06 racin e>

(Gamma Air Ecurvaient) 3 5 x 108 rad 2 0 x tca raam u

+LOCA Figure 5 peax is 216'C (420*F) for a 3-min te curar:on. u and ? 76'C (350'F) for a 10-m'n t test crofde.

• Acc cent cose rate rs 3 x too rad'n 'or 2 nours tnen ' x 'Ce racin

'c'TIO r ciudes 15 x tca rad acc cent cosage clus 0.3 x Margin also meiuced in 710. 'Oe rad gamma ecurva.ent for a cera T10 of 15 x 109 rad.

I 10 20 ' ' loo

' ' - ,  ;  ;;jjj 10 j , ,

8 j_9 7 , yg 40RizGNTAL OSE

, ,, , 10 ---

8 j, h, , ~i "("i[IYI Y

'4 j , ,

5 ' '

M i 7I II ' #

4 '

ill  ! I I 3 f f -

i (2 I ! i[ !i T ' '

3 P' i i!!  !  !

i ,

VERTICAL OBE i , i< s i 5' 2 - F

- l I I I I I[ \ II!I g g l '#

f I lf

• t i ! l  %- w g i

! !  !! ' i '

i ,e i.f  !

r U 6 ' ' ' I I ' ' 8  !  ! ' '

M 5 ' '

6 'I ' '

)  ! ' ' ' '

w 4  !  ! I I ' ' ' 5- '

w  ! '  ! '

l ' ' ' '

2 i

6 i i4 . > 4-  !

3-- > >

2 ' ' '

2 N-E11 ANO N-E13 QUALIFICATION PROGRAM- 2 3-

$ 2-- REQUIRED RESPONSE SPECTRUMS FOR $ N E11 AND N E13 QUALIFICATION PROGR AM-C OPERATING 8 ASIS EARTHQUAKE (08El 2' REQUIRED RESPONSE SPECTRUMS FOR DAMPING 1% l

' SAFE SHUTOOWN EARTHQUAKE ISSE)

(  ! '

! OAMP1NG,1%l 0.1 -

g 2 34 568 g 2 3 4568 100 O.1 -

1 2 3 4 568,'O 3 FREQUENCY, Hz 4568'00 3 FREQUENCY, Hz '

Figure 1.

Figure 2.

SUPPLY VELTACE AND EXTERNAL LOAD LOOP RE The Supply voltage limits are Shown in Figure 3 for both LOCA and non LOCA ac i voltage. the hmits of the transmitter external 1000 resistance Series resistance Of each Component

• in the 1000. . are Set a ions. However, for aTO givenCetermine Sucoly th e actual 1000 resistance. add the

• Transmittert without a junction bb.t is consicered as 0 onm:

traasm;tter eth uncton box has 10 onm resistarace for tre to to 50 mA vers:en. or 25 onm for resistance version rnA for Me 4 to 20 4 TO 20 mA OUTPUT 1500 ,

to TO 50 mA OUTPUT

- MAXIMUM LOOP RESISTANCE 1475 0 1250 -

MAXIMUM LOOP RESISTANCE

1. g .i 1250Q la f

T :;<:

]t. re-ITI

$ 750-ys g _y# Ysv w , '- .

E -

ulNJMUAALOOP a 500 - [,,*.*. RESISTANCE a

, asinit" Uni LOOP v .O~. 600 Q (AREAS A AND 81 RESISTANCE

- 600 Q

,,5. :

250 - -

VINIVUM LOOP RESISTANCE 250 Q (AREAS C AND 01 0 i 6

0 ii 1+

20l' 30l 40 l 6

50 i t I i 22 8 60 l I l 31 5 42 70 75 80 90 95 700 SUPPLY VOLTAGE. V cc Area A - Coerstmg Area reouire* eats for aon LOCA/MEL8 acohcat ters powered by a separate power tsociy and not from a svstemions of transmitters up to 80*C (180*F) emo'en u

Area 8 - Coerstmg Area reouiremeats for non LOCA/HEL8 module %mm m loco andresistance LOCA/HEL8 600 G.

Area C ame.ent. Transm tte s oo*ereo av a secarate power sucoiy and not frorn a system moaconcations u

- Ooerstme Area requiremeats ter non LOCA,HEL8 ters powered tnru a svitem moo +e usmg a f ete Dus supply er otP cute. %mm accocattons m loco resistance of800 transmitters G. un to (180*F) amoient. Transmit-Output Components N 2 Al 12V ro N-2Al 13V Converters. Loco resistances set ber system I/O component.

y 40o0 configuration intamally an the Area D - Operating Ares roouirements for ron LOCA/MEL8 and LOC ambient. Transmitters oc*erto te AreeE SPEC 200 System inout:Ourout u a systemComoonentsmodule using

- N 2Al a field 12V and OutNsupply. 2Al 43V e erence orCo other system scurce R v

n erters as en Area C.

Area F - Opetatmg Area for transmitte's .n non-t.OCA,HELS soo0 cations uo to 80'C (180*F) m ammum.

- Ooerstmg ancheations. Area for transmitters in LOCA,HEL8 anchcations u up to 176'C (350'F) ma nim m and non-LOCA/HEL8 Figure 3-l s

. . FUNCTIONAL SPECIFICATISNS N-E130L, DE and DH Series Differential Pressure Transmitters Spen. R nge, and Maalmum Static Pressure Sensor Transmitter Code Maximum Series Sean Umits Static (Caosule) Ranae Limits'ai kPa AP l inh 03 aP kPa AP ' Presaure N-E f 3DH M 5 ana 51 mH3 0 AP MPs psia H 20 anc 205 -51 anc + 51 50 aac 210 200 arc 850 -205 arc + 205 21 0 3000

-210 arc + 210 -850 aac -850 2' O 3000

'a Nonzero caseo ranges reawre cct oral zero e evation suceression nceec mt Uccer and lowe rarge ornts PERFORMANCE SPECIFICATIONS Sensor Span Between Transmitter Code Performancem at Reference Operating Conditions kPa mmHg, in % of Calibrated Span Series (Capsule) Hopeat.

or MPs inh 20. or pal Hyster. Dead Accuracy ability Reprodu- Dr rt/ ~

N-EtJOH M esis Band 5 arc 20 arc cibility'c Year 5' <Da AP =05 01 Note o' 205 rH 2O AP O C5 0 t5 0 25 M

50 aac 200 a*c

'30 <Da AP =05 01 0 10 525 *H:0 AP 0 05 0 15 M 3 25

• 30 arc 525 aac 2'O <Pa AP =0 75 01 350 r*2O AP 0 10 0C5 0 15 0 25 e Hysteresis 4 0 'C'i c' scan er 0 012 <Pa 'O C5 rHa Performaace scec:frea!ces Ostec OL *n crever s 'arger

.'c Aecrocuc:cmty 'ectoces e"ects of rysteres s receatacihty. ceac DBE PERFORMANCE SPECIFICATION *

(Output snifts are in cercent of scan)

Max. Output Shift,

% of Calibrated Span Normal Radiation at USL _See Note A TID 3.5 x 10' rac (gamma)

Accident Radiation =05 2 0.5 TID 2.0 x 10* rac # gamma).

(inc:uces 0.3 x 104 rac gamma ecurvalent 'or teta TIO :25 6 of 15 x 10' rad otus 0 2 x 10' rac for margin).

Seismic (Refer to Figures t and* 2. OBE anc SSE Resocnse Scectrums Dunng OBE's and SSE s After OBE's and SSE's 25 25 LOCAlHEL8 ISee Figure 5 arc Note 8) :1 :1 First 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> of Test Profile -

From 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of Test Profile 23 :S From 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to 30 days 22 :6 (30 cays at 176'F = 1 year at 120*F) :t :3 Notes:

A. The 085 cerformarce scecif cat cas ae'e cerivec from tyce test x-ng transmrtters at accre rrat greater The facie cercw hsts tre ar"um accectacte scan for !Pe extremes of tre OBE conc t:orsely 25 Sensor Model Coce Minimum Acceptable Span for N-Et30M cr N-E'3CM Extremos of D8E Conditions M

H 50 inh2O cr 12 kPa '

200 inh2O or 50 kPa A more accurate analysis. casec ucot'scecific acchcations (or for scans wn aceve aac fess tn n tae USL.

a

8. The outcut snift for gauge pressure trarsm.tters coes not inctuce ameient pressure effectssnown

50 i i s. I t 4 i 40 6

.-) >-

, QuauPIE3 UPE vs SEnveCE TEasFE AfumE FQm N-Et t (N 2 4-Et3 SE DES TKANsade TTEms 30 SAS40 0N ACfivAfiON ENERGY E4

• 0 70 ev iELECTRON VOLTSJ 20 l  ! l l TESTED ' I h\ l  ! I l

y

. ', iSOUD UNEl l

; A.

2 3

g 6

EXTmapOLATED

/ '

x A i

< DCTTED LINEl i 3 3 f l j i l

2 il 3,6ALiFsEO LtFE way SE ExTENDE0 y yo $3% Sv 8'EL3 REPLACEWENT

! CF AGE-SEN51T8vt COV80NENTs I i

, i

i l

t ,

' F -* % 60 70 80 90 t 00 110 120 130 140 150 160

' C -* 10 16 28 27 32 38 41 49 54 60 66 ft SEnvlCE TEveEmATumE Figure 4.

232 450 '

216

• SS os.

4 20 === - _. -- - -

- r 3 i  !

"* 25- ...s.

-A-  !. .s. wea-E's "r,,LE

,60 no_ _ _ _ _ ,

3

_ _ _.l!_i, a-

'Il l i l /

! I I 1 '

l 116 240a - -- - -

< ,/

s ii 5 i ' li i ,

t

( _.I__ _ _ g'  !

,3 ,, ,,, - l 3 ss.

2 66 150 f l

' l l

-l ., ,20 + _;-

O l l ll l I Meef viC A L $7m A y **,

i I

10 s ( l

,) y_ f I 4 4 i. 3*a .

'O 8

, l ' l

_'0** l_10 m.a j d wyg g _!

3, 2183 l

24 n -

t 30 a t a f YEAR AT 120 FI -__

TIME 2Emo T1VE 1% SECONOS ist. WiNuTES imias. Moum5 im. ANO D A v$ f as

• Cumvt AT 2'6 C 420 8 'sLOCA * *CMEMICAL SPR AY CONSISTS op SJ VL L A T CN TO E s 'EN3 #E A< T5 9pEmafumE SCaiC ACIO i0 28 90LAmi SO3IUM TMiOSULF ATE so 064 MOLAmi ANO $O0iuu MvomOxsOE . A$ MEOLimED TO CONTMOL om LEVEL TO 8ETWEEN .

8 $ ANO 1101 sen Av m ATE 70 at 0 2s U s w*et' Figure S.

c 1

PHYSICAL SPECIFICATIONS ' -

~

.the Enclosure Classification Meets IEC (P65 and orovides environmental protection of NEMA Type 4

• With Cast iron l Topworks Cover and Base Material Topworks(e.ca With AluminumR Transmitter -

Topworg,,,ei Code I Cast iron per ASTM A48-64. Class 30. with Amercoat 66 finish.

ga gg -

Ib ( kg l lb Code finish. A Low copper aluminum ahoy with vinyt[ paint N-E130H { 25 55 f 18 f 39 Application The aluminum cover and base meet the same requirements as tne nuclear transmitter with cast iron cover and base for seismic, radiation. aging, and LOCA/HELB.

present except where chemical sprays are Instrument Mounting Mountmg is on a vertical or hor-izontal DN 50 or 2 in cice. or en a surface A set of carts 1ctiunally available for this curDose. The. N-E11GH is h

ow, 'aac12.3 Wnen arg (5oct'onal cast iron. nonindicating sect:en

01 to total mas tures section for mounting selection.ever. can not be surface-mounted Aefer to

""U'"#'"*'"'*'""s aca 0 9 kg (1810) to total mass'"O'catng snet.on tax .s used.

Bottomworks Materials Refer to table below Approximate Mass Refer to adjacent tacle. ;c.4,so aca 3 0 kg(6 7ib)racket (3 0 :t) witn N-Et tGH. and aco 3 4 kg (7 5 ib) ween cot with N-E t 1 AL s

N-E t TAM. N-E f 3OL N-E13DM. and N-E t30H 71 Sottomworks i Part Description N-E13DH Boav Bolts ,

Bocy .

316 ss ~

Process Connector 3t6 ss sensor 3 6 ss Lower Force Bar l'6 ss Force Bar Oiacnragme Cuh ailev

_ Process Conrec!cr Gasset3*6 ss Sensor Gasnet 3*.6 ss -

Force Bar Gasket SJcc*e a Eiastcmer J n

.a ciTre Allforce asteriskea car c,acreagm mater aistrem s r= ace area coca acacrocess t.niewe! enrom r als are crocess wettec wettec All otrers hStec raate ium ailey STANDARD PROCEDURES, TESTS. AND CERTIFICATES Standard Procedures Design Control in accordance witn ANSI N45 2 and Tests and Certificates Standard Code of Federal Regulation 10CFP50. Accendix B.

Nuclear Cleaning To 1 com chtcrrde content Certif .

Quality Assurance Program in accoroance with cate (Form 7221)is crovided.

i ANSIN45.2andCodeof Apoendix 8. FeceralRegu'ation 10CFR50

, Hydrostatic Test At 150% of rated cressure. Cert.fi-Qualification Program cate (Form 7220)is provided.

Class IE Per IEEE 3231974 andIEEE Calibration Certificate (Form 3759)is crovided Certificate of Compliance Nuclear Classification Coce is CS-N/ SAC.344 1975.

Class il Per IEEE Class lE (Safety rela ed: to lEEE 3231974. lEEE 344-only). Nuclear Classification Ccce is 1975, CS-N/SRO-3441975 and NUREG 0588 (Cat.1] (Structural

. Dec 1979) integnty Form 4129CS-N/SRC is crovided.

Class 11 (Structuralintegnty: to IEEE 344 1975.) Form .

4129CS-N-SRD is proviced.

4

Type Test Reports Type test recorts (user documents), '

supporting ouahtication of thes0 transmittIrs to IEEE Trmw -

Standards 323 1974 and 344 1975 (aging, seismic. LOCA, aseest Nummer series user Documents and radiation), are available on an as needed basis. They N-Ef 30H-HA. Hl. IA. and il 1. COAACf o.

are not routinely orovided with every transmitter. Tne ad-jacent table hsts transmitter models covered Dy the ao- fr'e' m$.E to's orcoriate user documents- 2 OOAAC11 Test Recort 3 GCAACt2.

S.mriarity Occament I

DESIGN QUALIFICATION These transmitters nave ceen designed to meet the re-cuirements of acchcacie Product Quahfication Stan-cards The acchcacie standarcs crovice tr e methodology for proving device cerformance uncer generic environ-ments as estachshed in the Foxcoro Transmitter Quahfi-cation Program (OOAAC10). This includes Design Basis Event (DBE) environments. User documents. supporting the cuahflCat)Cn tests. are avallacle as recuired.

MODEL CODES Model Code: N-E13DN Series Transmitters

'N-E13DH = Nuclear Efectrorhc Differen'hal Pressure i nge Transmitter-Outout Signai

-H = 10 to 50 mA. 60 to 95 V de

-1 = 4 to 20 mA. 25 to 42 V de Base and Cover Materral

. I = Cast Iron IFor caustic scray environment)

A = Aluminum Senser Code-Scan L e its M = 5 anc 51 <Pa AP (20 anc2 205 inh O AP)

H = 5 and 210 kPa AP (200 and 2 850 inh O AP)

Process Ccnrect: ens 0 = None 180cy taccea for 1/4 NPT) 1 = 1/4 NPT 2 = 1/2 NPT 3 = R 1/4 4 = R 1/2 5 = Machinea to accect 9/1618 Amirc3 fitting

Enclosure 2 Response to NRC Questions of September 12, 1985

Question 1: Provide an analysis of the expected errors in the hog leg and reactor vessel head level measurements. This analysis should include not only an overall estimate of the measurement uncertainty, but estimates of each contributing factor, i.e.,

temperature of the inpulse lines, common mode pressure effects on the differential pressure transducer, and uncertainties associated with the transducer. Explain how the individual errors are combined to give an estimate of the overall error.

(This is a repeat question described in the June 14,1983 NRC to GPUN letter).

Response: Attachment 1 of the enclosure entitled "TMI-l RCITS Loop Error Calculation Sumery" provides a summary of GPUN Calculation C-1101 -662-5350-023 which is available for review at the GPUN Corporate Offices.

The norm 1 and accident errors are based upon two separate error calculations: (1) Instrument loop errors and (?) errors due to temperature distribution in the RCS. The instrument loop error is random error, + 13.88%

for either hot leg or vessel, which will likely be reduced as RB conditions become less severe. The temperature error -9.6% and -9.9% (hot leg + vessel respectively), which is reduced as time elapses from reactor trip.

An estimte of the overall error could be stated as +4.28% to -23.48% and

+3.98% to -23.78% (hot leg + vessel respectively).

Question 2: The Reactor Coolant Inventory Trending System with no pump running as described appears to satisfy the requirements of NUREG-0737 II.F.2 with regard to qualification, redundant trains, etc., including redundant displays on two computers, except for a common mitiplexer between the instrument trains and the two computers. Earlier submittals suggest that the reliability of the mltiplexer was being investigated, but we have seen no data or justification for this common link. Since Technical Specifications will require that at least one complete train be operable, the mitiplexer could become a critical component requiring plant shutdown if it should fail.

Therefore, provide the results of the investigation on the reliability of the multiplexer and justify this common link or provide an upgrading plan for the multiplexer to meet the single failure criterion specified in Appendix B of NUREG-0737.

Response: In order to clarify som apparent misunderstandings concerning the display system for the RCITS see Figure I for details. As noted in Figure I the void fraction signal from the RCP's is fed into one of a number of cabinets which mkes up the transient monitor portion of the multiplexer. Void fraction can only be calculated in the Mod Comp Computer and not the Bailey. Therefore, this portion of the system is not redundant nor could it be mde such with the existing computer system.

-u Although new computers for TMI-l are under consideration, there is no justification for installing new computer simply for this instrument which is used for a beyond-design-basis-event and is only used when RCP's are operating. Since saturation margin below 25*F is the trigger for turning off the RCP's, and once the RCP's are' turned off the void fraction is useless; redundancy is not required. This same argument holds true for the 855 multiplexer interface switch which is also not redunda nt. - Note too that signals in the Mod Comp Computer cannot- be displayed through the Bailey Computer displays. For the RV head and the hot legs level -indication signals enter a separate section of the Bailey 855 multiplexer and are displayed through the Bailey or Mod Comp Computers. In this case, the computer acts as merely a display device. Again the 855 multiplexer interface rennins the common couponent.

However, no significant problems in the operation of this component has been experienced at TMI-1. Although recorders could be installed for level in the control room, they could

- not be displayed on the main console and would be relegated to a position on a back panel which could not be easily. read from the control room operators work station.

GPUN has investigated the reliability of the Bailey and the Mod Corp Computers. Because TMI-l has not operated in 6-1/2 years and the data prior to that was incomplete, no neaningful reliability data was obtained.

Concerning Technical Specifications, GPUN does not intend to use the model standardized Tech Specs but rather an integrated Tech Spec for the inadequate core cooling instrumentation. In this way failure of a single conponent would not, of necessity, result in the inability to monitor an inadequate core cooling

- situa tion.

Question 3: Provide.a block diagram of the final ICCI system from sensors to the display system.

Response: See Figures 2, 3 and 4 attached.

Question 4: Describe the current. status of the final ICCI system with respect to conformance with the design requirenents of NUREG-0737. Item II.F.2, identify any as-built deviations of the system from your previous design descriptions, and provide the-implementation Letter Report (described in Enclosure 2).

Response: The RCITS is fully described in GPUN submittals dated March 10, 1983 (5211-83-071), Ja nuary 31,1984 (5211-83-379), July 6, 1984 (5211-84-2173) and February 6, 1985 (5211-85-2026).

Additionally, conpliance with the requirements of this system with respect to RG 1.97 is described in GPUN letter dated r--

October 1, 1984 (5211-84-2252). The principle deviations are the common taps for RV Head and the DH drop line; RCP motor power is not 1E and the display system which cannot be proven to have 99% availability for the liquid level display due to lack of recent operational data. These deviations have been addressed in the reference submittals.

The Implenentation Letter Report will be provided in February, 1986 The RCITS has been installed, tested, and calibrated.

Test results are available at the site for review. The system, in general, performs as designed. Procedure changes have been made which incorporate the RCITS, and operators have been trained in the operation of this system. ICC Tech Specs are currently under developnent.

Question 5: Provide a detailed description of the upgraded CET system and SMM.

Response: The extent of the upgrade for the Saturation Margin Monitor (Siti) included replacenent of the existing temperature sensors with qualified sensors and assuring independence from the ICS.

Additionally, the saturation margin digital indicator was seismically qualified by test.

The extent of the upgrade for the Backup Insure Readout System was to demonstrate environnental and seisnne qualification of the system. Additionally, a seismically qualified digital indicator has replaced the original indicator.

GPUN Letters:

02/02/82 (82-007) 12/22/83 (83-369) 05/31/84 (84-2125) 04/12/65 (85-2072) 05/21/85 (85-2100) 06/28/85 (85-2121) l i FIGURE-1 i D FUNCTIONAL SYSTEM OVERVIEW 4

I mina=$igMr e TC81' CC aCf 2m 'tm Iy l

t ,

Mu%1s9Lissa ,y,Em C tm

{

• M AC ,,M S

\

ANALOG ggg impur MUL TTPt tX t2 a

e MM i CISa )

enoCAAMMEAS atACfuwtTER '

m w'omos 3

Cat

~ _ i l

TMI 1 -

e L U

=co coup cussIc "

3 K% " ***

C nft atiot) -

nxto M.

0 53 SOLID STATE Mt % EOR Y i M wono 512s womos

-, ECF ass CRT atamu F4 TNT (4 m.L w 200LPw p<rtawace -

FAR HQ CRT I vf0$av o'

0 .

inr w aca UTILITY N C"'

i

- PRI!rTER

200 LPM --

J b ess meuf gte m tutas 1 COLOR UTILITY cpy COI4R

' 8 "0 "

a OG cisc CRTS(2) g Imma E i come MEJWO4Y wtyomy anatoc RV KM >

l32KWORDS \

monum

{,

8ECO*088 TRENO 09tRATC9s AWGI ' ## CONSCLE

• M { 4 nato'o numme amm.m coNas

. Lea {* =

"' , ojpg w

? ""

moon,w "

1 conscos oc

~-

" P'"

b

~

l .m Lu, ,

~

PAPf 4 faPE PApga Tapg AAActA Py4CM PROGRAM "N )m ALAmu I MA t

m e.. _ . _ _ - ,

O o

e HOT LEG LEVEL A

RCP POWER DDESEL-8ACKED 1E POWER 1A1 PM1 -

u MMR RMM RV HEAD LVL

> CRT DISPuy A 8 l m

> COMP 1A2 PM1 -

^ COMPUTER

> PRINTER VOID INPUT a T COMPUTER ROOM MULTIPLEXER >

181 pm1 -

FRACTION f CONTROL ROOM 182 PM1 _

Y RECDRDER 855

> BAILEY COMPUTER

> PRINTER

" CONTROL ROOM VITAL POWER (BATTERY BACKUP)

HOT LEG LEVEL 8

Figure 2 RCITS Display System

~'- -

t g

s

)

A nmf 7i rA

ga(

' rl aA M

4 r iy l y e aa tl?

, t u

p aao tl i p(

i p gs(A m gs i i o ti uD DD C .

)

Y f, A

<' L P

S I

D

/

N I

G R

A 3

l M l a e a i N r i t O u t n I g ll n) 1 e) T i eA rB A F r( e( R e f f

U f T f i A i D S D

e e tt )g tt ) au sp(

auA s

sp( T mo _

1 T, om C C _

. I I[

IJ I k- /

)

) B )

6 ) ( B

( A (

( g g

i Ti P T P I

/

,s

'. i . ,' j ,

~T Fit:11Rit i .

k I E l'ower .

,_ _ _ _ _ _ _ _ .. _9F _ _ _ _ _ _ _ _ _ _ _ _

8

- 8 Con t ainmenit

)

, luput Sinnn! 4 lilgital k ronillt lon Dinplay.

't s'

• E 8

• IiointIon Devices Ilackup Syntem, 4 RTD'n 16 T/C'n 3 - s 4 VI.tal Power i

36 T/C's tio:1 )l CRT Dispiny l I- 8l_. 8 4 RTD's i

-4 comp C -

lf ses R'l.D e i Computer sk Il Printer l 8 I

e i T C s, Input l i, I

, HuttIplex -

I 8

' . i t 855 8 m -

e

-) Computer r m

Printer 8 I

, ah Digital ,

, Display ,

e Non IF, Power 8

Battery llackup 8

_ _ _ _ _ _ _ _ _ _ _ _Pr.i.m_arl Syst_em_._ _ _ _ _ _ _ _ _ _ _ _ _ _ - 8 INCORE TilERHOCOlli'i E SYSTDI

.