ML20118C095

Technical Evaluation of Duane Arnold Energy Ctr Emergency Svc Water

ML20118C095
Person / Time
Site:	Duane Arnold
Issue date:	08/01/1992
From:	Valente J EG&G IDAHO, INC., IDAHO NATIONAL ENGINEERING & ENVIRONMENTAL LABORATORY
To:	NRC
Shared Package
ML20118C093	List: ML20118C095
References
JUV-92-011, JUV-92-11, NUDOCS 9210090243
Download: ML20118C095 (19)

Text

..

JUV.92 011 Technical Evaluation of the Duane Arnold Energy Center Emergency Service Water Techincal Specification Change Request J.U Valente August 1992 921oo9024[gj$$$$31 PDR ADOC PDR P

Table of Con:ents Introduction 1

2 Emergency Senice Water Component Review and Report Layout

.........,.2 3

Diesel Cooling Rcquirements.................................. 3 11 Calculation Methodology,.

.........4 3.2 Assumptions...

..... 4 3.3 Calculation DetC*..

.....................5 3.4 S u m m a ry of Fin dings.................................... 6 4

Control Room Chiller Cooling Requirements......................... 6 4.1 Calculation Methodology....

.........................7 4.2 Assumptions 7

4.3 Calculation Detai:s.....

.8 4.4 Summary of Findings

.. 9 5

Pump Room Coolers: HPCI, RCIC, CS/RHR Pump Rooms..

....... 9 5.1 Calculation Methodology

............................. 10 5.2 Assumptions.

11 5.3 Calculation Details 12.

5.4 Summary of Findings 13 6

RIIR Pump Seal Cooling Requirements 13 et l Calculation Methodology 13 6.2 Assumptions

.. 14 6.3 Calculation Details 14 M

Summary of Findings

. 15 -

7 Centrol Room Habitability Considerations 15 8

Conclusions 16 9

Reterences 17

T 1 introduction At the request of NRC/NRR, Brookhaven National Laboratory (BNL) staff have performed a review of a proposed plant technical specification change submitted by the Iowa Electric Light

& Power Company [1]. The submittal is for the Duane Arnold Energy Center (DAEC) and involves the modification of Emergency Senice Water (ESW) System flow requirements as they presently exist in DAEC's plant Technical Specifications (TS). These require weekly surveillance testing of the ESW system against strict flow limits which are a function of water temperature, once the river water e ceeds 80*F.

!n 19W a Senice Water Safety System Functional Inspection determined that the original basis for the TS limits could not be retrieved. As a consequence, DAEC commissioned a re-cvaluation of the heat loads on the ESW system. The resultir.g series of calculations were performed by Bechtel Corp. These are Bechtel designated Task 466 calculations M-001 to M.

u l 1.

These calculations provide a basis for the DAEC proposal of reducing the flow requirements of the ESW sysicm and reducing the surveillance requirements to once every three months ne licensee pt ports that the reductions are now possible because of

• improved modeling methodology, current design information, and the reduction of component heat loads through the installation of additional insulatien."

The scope of DNL's review effort was to determine the validity of the stipporting calculations provided by Bechtcl. A sitt visit to the plant by NRC and BNL staff was conducted after a preliminary review of the calculations and requests for supporting documentation (primarily references used in the calculations). The intent was to verify the assumptions that went into the calculations, and familiarize the review team with pertinent as built conditions. A further benefit of the visit was the interaction with the licensee's engineers.

Before leaving the site, the licensce's personnt. were informed by tne NRC representative that the submittal was not acceptable in its present form. Key to this finding was the assumption governing the river water fouling factors for heat exchangers and cooling coils. Other concerns invoked discrepancies in the assumptions for piping insulation, room temperature stratification, et,Jpment qualification and air cooling tiow short-circuiting.

1

4 Subsequent to' this finding, the review of the calculations continued so that the licensee could -

be made aware of all BNUNRC concerns in the event DAEC decided to resubmit their 73 change request. What follows is a discussion of these calculations with insights gained from the site visit. This is followed by oui conclusions.

2 Emergency Senice Water Component Review and Report Layout The discussion on the ESW components is divided into four sections. In addition, there is a section dealing with control room habitability. Three syste.ns seniced by the_ESW represent, accordmg to Bechtel s newly performed calculations, nearly 93% of -the total ESW flow requirements at 95'F river water temperature (see lable 1). At this temperature, the three heat exchangcis on an e m e rge ncy diese1 Table 1 ESW Mow Requirements (gpm)

(combc. tion

air, jadet wat e r, and lube oil)

Equipment River Water Temperature represent 49G of the 95'F 80*F projected required ESW Diesel Generator Coolers 310 310 flow,

The diesel cooling requirements are discussed

-Control Room Chiller 190 75 in Section 3.

RHR/CS Room Cooler 95 30 RHR Pump Seal Coolers 21 14 The control room eccler (two) senices a number of rooms RCIC Room Cooler 8

3 in the control room tailding.

This control HPCI Room Cooler 5

4 room cooling coil interfaces RHR Senice Water Pump 4

4 with the chiller (a

Motor Coolers (two) -

refrigeration loop). On a CS_ Pump Motor Cooler 3

3-LOCA trip signal the Control Building HVAC 2

2 chiller's condenser is instrument Air Cornpressor cooled by the ESW instead 638 445 of the normal well water 2

[

system. The projected ESW Dow requirement of this system is 30% of the total projected ESW Cow requirements at a 95'F river water temperature. De control room chillers are discussed in Section 4.

Section 5 of this report discusses ESW Oow requirements for the ;afety related pump romm cooUng i nits. These involve those unit coolers servicing the reactor building rooms housing the RMR and Core Spray (CS) Pumps (one room with 2 RHR and 1 CS is considered); the High Pressure Coolant injection (HPCI) pump; and the Reactor Core Isolation Cooling (RCIC) pump.

The RHR/CS pump room is projected to need 15% of the newly estimated ESW total flow requirement. He llPCI and RCIC coolers represent only about Wc each of the estimated ESW total th m.

In section 6 calculations on the ESW Cow requirements for the KHR pump seal cooler are discussed. These requiremer.ts are quite smallin comparison to those addressed in the preceding sections, accounting for abcut 3G of the total projected ESW Gow requirements at 95'F river water tempeiature.

Section 7 contains a discussion of control room habitability and single failure criteria. (Dese

i. sues are closely tied to the chiller system.) Finally, Section 8 contains our conclusions.

3 Diesel Cooling Requirements Each emergency diesel has three heat exchangers which rely on ESW cooling water. These are the combustion air. Iube oil, and jacket water heat exchangers. The ES% Cows on the tube side and cools the three heat exchangers in series. The shell Cow path is baffled and the design How varies from 400 to 500 gpm [2] at diesel rated kw output of 3250 kw. As can be seen in Table 1 the estimated ESW Oow requirement at 95'F river temperature is 310 gpm. Calcul ion #466-M 019 is the governing calculation for this effort. It is supported by calculation #466 M-001 and 466-M-003. The latter of these calculates the heat load on the heat exchangers for diesel operation adjusted for a design continuous load of 2S50 kw. The diesel vendor's supplied rated speci0 cations for the diesel is 3250 kw. The three emergency diesel heat exchangers are a major 3

load for the ESW system, requiring 49% of the projected ESW Dow at 95'F river water temperature.

3.1 Calculation Methodology One wishes to establish the unit conductance (U) between the tube and shell side fluids for vanous ESW Gows and river water temperatures. This involves the convective film heat transfer coefficients on the tube's inside and outside surfaces as well as the heat resistance afforded try the tube material and any fouling on the inside and outside tube surfaces. Since convective heat transfer correlations for forced Dow inside tubes is well established, as is the tube conductance, the approach in obtaining U concentrates on detern"'4g the convective heat transfer coefficient on the shell side. Data sheets are supplied by the vendor for clean (no fouling) conditions, giving the total heat transfer capacity of the heat exchangers w;th the related fluid conditions (tube's inside and outside Guid mass now and temperature).

Then using the Img Mean Temperaturc Difference (LMTD) approach for heat exchangers, the preparer is able to establish the shell side convective heat transfer coefficient for the given ESW Gow conditions. This is checked against the shelt side convective heat transfer correlation being employed. Once this is donc, the preparer is now able to determine the effective unit conductance for various ESW ths.md riser water temperatures, since the heat load on the heat exchangers remains fixed.

Estimates are rnade for the fouling factors under service conditions.

3.2 Msumptions Calculanon M-003 lists the vendor supplied heat loads on the heat exchangers for the rated (3250 kw) and design continuous rating (2850 kw) conditions. We take no issue with using the 2K50 kw v'ilue for the ESW study.

I Calcobtion M.009 assumes fouling factors that are not realistic for DAEC service. In particular, the use of a fouling factor of 0.0005 (tube and shell) to determine required ESW flow is non-i conse rvative. These values are more representative of sea water cooling than DAEC's river water. IndeeJ. based on the flandbook of Ileat Transfer [3], the fouling factor values for river uater should be 0.003 if the now is 3 ft/s or less and 0.0d2 at greater than 3 ft/s. The 4

i

i recommended values for engine lube oil and engine jacket water are 0.001-[3). Iowering the ESW flow to below about 400 gpm reduces ESW flow velocity through the tube to less than 3 ft/s, and hence raises its fouling factor from 0.002 to 0.003. Both of these are greater than the factor employed in the calculation. It is of interest to note that the heat exchanger vendor used fouling factors of.002.003 for shell side water and.006 007 for shell side oil in some of the specitication sheets PJ.

During the site visit, DAEC personnel stated that test results indicate: lower loads on the heat exchangers at 2850 kw operation then that supplied by the vendor. However, it should be understood that this could be greatly affected by the environmental conditions in which the diesel operates. At DAEC, the diesels are housed in rooms which could have relatively low ambient temperatures. The piping to the heat exchangers appeared to lack insulation. Hence, the lower than expected observed load on the heat exchangers may be more attributed to the environmental conditions at the time of the tests (especially if conducted at less than summer design temperature) than any conservatism in the design.

There are bypass lines for the shell side fluids (diesel fluids) around the heat exchangers. One would assume the bypass functions is used to assure the cooled diesel fluids return within a given temperature tolerance. Thus, increasing the bypass reduces the heat ternoved and allows the diesel tluids to return at a higher temperature. If in establishing the tube outside heat transfer coetticient (ht,) for the diesel heat exchangers, no bypass wts assumed but bypass existed, the velecit of the diesel fluid would be overestimated and so would ht, since ht, should be o

proportiona! to approximately the square root of the velocity. Since ht, is proportional to 'U",

the effect could be a lower predicted ESW tlow than required. Also of particular concern is the ettect on the baftling's ettleiency on the shell side due to any bypass.

13 Calculation Details The vendor's data sheets on the heat exchangers are used to establish a shell side heat transfer coefticient for clear tube conditions, For the air cooler heat exchange in calculation M401 version 1, an error is made in using the venJor's data. The calculation uses rev 1 of the data sheets Hj tar the e:can unit conductance of 622 Bla/hr.tt:fF but uses rev O's value for the heat 5

i

transfer of 2,769,000 Btu /hr instead of rev l's value of 2,795,000 Btu /hr. - Further, the value used for the shell side flow area (0.314") must be a derived value, and the review team could not determine-how it was attained since the data sheet does not supply the shell diameters. The preparer, in checking his results, found a convergence within about 10% on the shell side film heat transfer correlation against vendor data to be acceptable. De review team believes that this should have been translated into an uncertainty in the required ESW How.

3.4 Summary of Findings The touling factors employed on both surfaces of the tubes need to reflect the service conditions that exist. In establishing the design condition parameters, the use of the vendor supplied data sheets should be consistent. That is, the unit cont'uctance and heat transfer should be from the same sheet. If the preparer finds that his calculations show a convergence only within 10% on the shell side Him heat transfer correlation, this should be translated into a safety factor for the ESW 110w requirements.

4 Control. Room Chiller Cooling Rcquirements The Chi!!ct Water System (CWS) is a dual loop system with one loop normally operating. Each loop con:ains a separate condenser, compressor, evaporator, chiller watet heat exchanger, and control room buildmg cooling coil.

In normal operation, the freon of the chiller is cooled in its condenser by well water. However, on a LOCA signal the emergency diesels stan and ESW is supplied to cool the chiller condenser.

Both the chiller water heat exchanger and the control room air cooler remain on.line during a LOCA.

The chiller compressor is capable of supplying 200 hp (509,440 Bru/hr) to the retrigeration loop.

If oti. site powei is lost, chilled water flow into and out of the chiller water heat exchanger is kolated. The compressor is reduced to 75 hp (191,040 Btu /hr), and the control room coolin;;

coil remains on-line.

s 6

The calculations being reviewed for this task involve 466-M-003, M-007, and M 008. In terms of projected ESW flow requirements, the demands of the chiller system, servicing only essential

!oads, are second only to the diesels (see Table 1), requiring 30% of the projected ESW flow at a 95'F water temperature.

4.1 Calculation Methodology The basic approach of determining an overall unit conductance for the mndenser utilizing known data points is employed. Since the data points are not specifically for the DAEC condenser, the Wilson Method K] is usca to adjust the supplied data to the specific DAEC coil. 'nic preparer i

then goes on to determine the necessary ESW flow for different river water temperatures i

knowmg the condenser's unit conductance and the design load on the condenser. The design load is based on the Control Room cooling coil design capacity and the refrigerant compressor's work.

4.2 Assumptions Supporting heat load calculation M-003 assumes that 100cc of the compressor work must be removed by the condenser. This is a conservative assumption, since some of the work would appear as heat lost to the ambient air.

It should be noted that calculation M-007 is used to determine the seguired ESW flows to the chiller candenser when only the control room cooler and the refrigerant compressor are the loads to the,ondenser. Calculation M-008 determines the heat load to the control room cooler by examining the human, equipment, lighting, heat transmission, and solar loads to the room sersiced by this cooler, but no fresh air intake is assumed. Calculation M 003 simply assumes that the heat load to the cooler is that of the cooler's design which would include fresh air inteke loads. Calculation M-007 uses the data from M-003 and not M-008. It should also be noted that calculation M-O'J8 lacks administrative control since there is no reviewed sign-off.

The capacity ot' the control room cooler is based on an inlet air temperature of 84.5* dry bulb temperature (DBT) or 67" we: bulb temperature (WBT). This gives a relative humidity of about 7

409. One might expect that for operator comfort the control room may be maintained below this Dirl. The coil's heat removal capacity is a function of its condensation capability, and this does not appear to be considered in M-007.

As stated in 4.1. calculation M-007 employs the Wilson method [5]. This assumes the freon in the condenser has a condensing fluid tilm coefficient which is held constant over normal operating ranges.

His assumption translates into requiring the Log Mean Temperature Difference (LMTD) and the fluid flow rate (freon) to be held ':.anstant.

If used appropriately in the calculation. this would seem to be a reasonable assumption. The wndenser manuf acturer provided unit conductances [6] for operation of the chiller at various E5W tiow rates. Unfortunately, this was dor.e for a condenser different from the one installed at DAEC In:' cad of requesting a revised series of calculations from the vendor, the data was used in the calculation to establish a shell side (ftcon side) heat transfer coefficient for the meorrect condenser. This I ecame part of the overall condenser unit conductance in the Wilson method approach. Then, before applying it o the DAEC condenser, it was corrected for tube s

msiJe and outside surface areas.

Another auumption made was that the tin type of the DAEC condenwer was the same as for the cm inser for which information was received from the vendor. Again, the review team he icses this ;wrmation should have been verified.

2 The ESW required flows were determined using a fouling factor of 0.001. As discussed in Secuon 3.2 of this report, this is too low for thi. ESW water. We believe 0.003 to be morc realistic.

J3 Calculation Details For the Wikon methoJ to have a reduced uncertainty, it is important that the freon flow rate anJ condenser LN1TD be held nearly constant. The calculation lacks documentation to confirm thiv 8

With respect to the freon flow rate being constant, the vendor's supplied data and the DAEC condenser freon Dows should be compared to assure that the Wilson constraint on freon fl<w rate is satisfied. In doing this, one should assure the correct condenser load from the vendor is used with the appropriate freon Dow rate. The referer.ced data [7] supplied by the vendor assumes an evaporator load of nearly 120 tons. He heat load on the condenser for the DAEC specific case was 859,200 Btu /hr or only = 72 tons, and only 668,000 Ptu/hr or less than 60 tons is evaporator load. The true load on the condenser assumed in the vendor's calculations [6] used to generate the supplied normalizing data pints needs clarification.

4.4 Summary of Findings The calculation is piesently unacceptable. Confirmation through in-field inspection of actual condenser components may be required. The two constraints which need to be adhered to for use of the Wilson method (i.e., constant freon Gow and LMTD) have not been confirmed to be true. Control room coolirig coil capacity may be affected by its condensation capability for the air thermodynamic conditions assumed. We were not able to confirm if condensation capability was considered. The fouling factor employed is non-realistic for the ser ice conditions.

5 Pump Room Coolers: HPCI, RCIC, RHR/CS Pump Rooms The Bechtel task 466 calculations associated with these cooling coils are M-002, M/X)3, M-005, and-M 006. The Rl!R'CS pump room coolers are predicted to need more ESW Cow than the llPCI and RCIC room coolers combined. The proposed ESW flow requirements can be found in Table 1.

For the purposes of this review, thcre is suff.cient similarity in the pump room cooh.:alculations that we will combine our discussion of the three (four if one considers both RHR,CF rooms) air cooling systems The I!PCI, RCIC, and RilR'CS pump room cooling systems consist o' unit coolers in rooms l

which are presumed to be isolated frora their surroundings during a LOCA. There are some through wall penetrations with draft activated louvers. (Based on our site visit, these louvers appeared to be in need of maintenance.) he rooms in quesFon surround the torus room.

l l

Typically the rooms are about 20 - 30 feet in height with the h<.a: sources (steam turbines, l

9 l

l l

i electric motors, pumps, and most of the piping) located at the lower elevations, and the cooling-coils at the upper elevations. De cooling coils are er closed in unit coolers consisting of filters and air harr.flert with exhaust du ts. The cooled air is typically exhausted close to the unit coolers. nat is, there is a minimum of exhaust duct work. De Door areas of the RCIC and the two RilR/CS rooms are each about 600 to 800 ft2 while that of the HPCI toom is closer to 1500 ft2 5.1 Calculation Methodologv A computer code was employed to perform these calculations. It is Bechtel's desigreed PC version of ME2bl (DASIICC). Details regarding the code were r.ot provided. Based on the fimnert iatormation available, tne revio.rn have reconstrue.ed the methoJology as follows. The approach taken ny the preparer is to use a design point for which as much data as is available can be used. As indicated in Asuptions, Sectica $2, there sie four unknown parameters.

These.re: air flow rare, fouling factor, exit air wet bulb temperature, and the calculated number ai coil rows. The code user inputs the first two of these as initial guesses. The third is calculated by the code, once the code user stipulatn that the coil only performs sensible cooling of the air, and the inlet WBT and DBT along with the exiting DDT are supplied as code input f rom design criteria. With the total air to coil surface area furnished as code input from vendor supplied data, the code will output the number of tube rows, as well as the cooling capacity of the coil. The coil cooling capacity is a known value that the code user matches.

l l

In an iterative manner, the user then varies the air volume flow to match the coil's known at-l capacity the known conditions, while fixing the fouling factor and stipulating only sensible air cooling. Initially, the number of tube row allowed to vary. Once a match to the coil's upacity is obtained, the code's calculated number of tube rows is held constant for the ESW flow study.

l l

The number of tube rows has thus been establishca based on matching.a code's prediction of heat removal capacity at the investigated design point to the code user's estimate of the air llcw over the coi', the fouling factor, and the assumption of only sensible air cooling. De code user now goes about determining the required ESW flow through each coil for various ESW coilinlet 10

-~.-

s temperatures, each time assuring that he is matching the air cooling capacity previoudy determined in calculation 466-31 003, and employing a fou!!ng factor of 0.0005.

By this approach, the preparer has essentially established an active wi' nurface area with an air flow rate. These parameters are determined by the number of actual tube rows calculated by the code for the supplied air flow. Together with the stipulation of only air sensible cooling, these parameters help determine the air to coil Nat transfer rate. But the degree of true condensation on the ccus is a function of how cool the coils are, as well as the air flow red moisttere content of the air.

In determining the required ESW tiow for different ESW temperatures, the relationship of :he coils efficiency, based on the degree of condensation atforded by the differcut coil temperatures,is lost in this approach. The exhai' t air's V BT, and hence the degree of condensation vs. sensible cooling are not supplied ry the vendor.

The approach appears to lack sufficient design point characterization and it may be necessary to determine the true air flow rate over the coils by examining the air handler. The number of tube rows are known and should be used in the calculation. 'Ihis would leave only the exit WBT and fouhng factor to be matched. To establish the design point, a communication with the vendar may be necessary. If this is unsuccessfut, then in situ testing to establish a new design pint may be necessary.

12 Assumptions The cakulations assurne no condensation of the air vapor mixture since they state only sensible cooling is used. Reativtically the amount of condensing which will occur is a function of the true rel:.;ive humidity (RH). coil temperature, and volumm flow over the coils. if condensation did occur. it is likely that the coils would have better heat transfer coefficients on the air side of the colis. Not knowing whether condensatior was assumed by the vendor in establishing the coil's capacity could invalidate the design condition which the preparer uses to determine some missing design data, including the number of cooling coil rows.

11

E The number of cooling coil rows could have been made a parameter that was known, i.e., an input parameter. Instead,it appears as a derived parameter tied to establishing the design point for the coils. The air mass flow is also assumed because of lacomplete data.

Finally, the fouling tactors assumed for the ESW flow in the calculations (0.0005) is too low for the sersice condit ons which would demand a value of 0.003 for flow velocities less than 3 ft/s.

i 5.3 Calculation Detaile The computer code ME201 (DASHCC) was verified by showing that the results one would obtain by performing a sample problem oy hand match the results the code would supply.

Ilowever. since the hand check was done using the cc relations in the code [8] this method of verification only checks that the algorithms in & szm were correctly programmed and the code correctly installed. We do not know whethei e wrrelations therne'ves were validated. This sht uld have been done for safety related applications, During our tield visit, the HPCI coil wu examined and found to have 14 ron (vertical to air fbv) of 8 tubes each. In the computer code used by the preparer for the calculation,14 rows wers input, but the code determined the number of tubes. A design point value of 6.6,5 rows was calculated by the computer code. This discrepancy appears to have been known by the preparer, but it remains unclear why this remained an unmatched parameter for the calculations.

The heat loads determined in calculation 466-M.003 are not representative of as-built conditions in terms of piping and component insulation. There wa, observed temperature stratification in the equipment rooms cooled by these air coolers and this is not accounted for in the allowable maximum air inlet temperatures to the coolers. Arrangement of the unit coolers ini the rooms would appear to allow for air flow short circuiting octween the unit coolers' exhaust and intake.

These are two RHR/CS rooms and it is not clear why the proposed ESW requiremems do not include both rooms in determining total ESW flow requirements (9).

l 12

t ;-

a 5.4 Summary of Findings The calculations are not acceptable in their present form, ne calculation appears to have too many unknowns when establishing the design point. Further, the fouling factors used for the ESW in detcrmining the required flow are not in agreement with Reference 3.

Finally, the assumptions used in the calculation for piping insulation and Dow distributions are not representative of as. built conditions.

6 RilR Pump Seal Cooling Requirements The RHR pum;) seals are cooled under emergency conditions uej ESW water.' A small quantity of nuclear grade water from the RHR pump discharge is passed through a pressure reducing orifice, a cyclone separator, ari then to the tube side of a tube and shell heat exchanger before being injected as seal cooling water. ESW Dows through the shell side of the heat exchanger. The ESW Oow requirements, as shown in Table 1, are only 21 gpm at 95'F.

Calculation M404 is the governing documcat for establishing this flow requirement.

6.1 Calculation Methodology The basic approach employed uses pressure drop curves for the orifice and separator to establish the Dow to the heat exchanger. This appears acceptable. However, allowing the seal cooling water to return at the scal's equipment qualification tempenture limit is overly optimistic. It leaves no margin for the possibility that the seal may be elevated in temperature beyond the seal cooling water temperature due to contact with the NSSS water.

Once the NSSS dow and change in temperature through the heat exchanger are known the required energy removal by the ESW is established. The heat exchanger's unit conductance is then determined using correlations foi convective heat transfer on either side of the tubes.

Design data was made available giving the heat exchanger's ESW Oow and inlet temperature for a given change in temperature of the NSSS seal cooling water.

13

6.2 Assumptions Two Ri!R modes are considered. Operational mode J [10), which is the Minimum Flow Bypass Mode, and mode E {10), which is the normal shutdown mode after Reactor Pressure Vessel pressure tclief blowdown to the plant's main condenser, ne udeulation 466 M-0N discounts mode J because of its short time duratien. Ilowever, the analysis shows that J mode operation will result in a seal temperature exceeding the 150*F qualiScation temperature, his discrepancy has not been adequately addressed by the licensee, he mode J alignment [10) which tan suction from the RPV at 135 pt!3 and discharges to the suppression pool should be examined by the licensee to ascertain if the procedures call for RilR pump operation. An RHR pump head is necessary to drive the RHR pump seal water.

Calculation M.004 lacks appropriate administrative controls. All new revision 1 pages of this calculation hek checker's sign off. The revision is poorly incorporated into the original version ut this calculation.

It3 Calculation Details Sation '2 of calculation M.004 is identi6ed as beit6 'not required," but is referenced in Section 14 to supply the fouling resistance.

In deriving the shell side 61m heat transfer coefficient in Section 7.4, preliminary data of Section

' 1 was employed to verify the assumption that the correct ESW flow is 12.4 gom. The use of this preliminary data of Section 7.1 is inaccurate since Section 7.1 employs what appears to be an incorrect pressure drop assumption across the separator to determine the seal water cooler NSSS Dow of 1.5 gpm. Later esta empicyed in revision 1 of the calculation shows a 1.1 gpm seal water Oow for the same 147, si RilR pump head cot ditions. De problem aiises because a shell side Elm heat transfer coef0cient of 359 Btu /hr.tt *F is determined in the original version.

2 This value is carried forward into the revis sec. ion of the calculation to verify the newly determined unit conductance.

14

Il

l 1

i w.

.w a-Fouling factors of 0.0005 for ESW water and 0.0002 for NSSS water were used. -We take issue with :the ESW vahtes and even that of the nuclear grade water should be justified. De recommended p) value for distilled water is 0.0005, and this would appear to be acceptable for NSSS use. 'Ihe ESW water should be assigned a fouling factor of about 0.003 p)If its velocity.

r

-is less than 3 ft/s.

6.4 Summary of Firidings Overall this calculation appears to lack administrative controls and clarity. Use of data _to -

-determine the shell sid: film heat transfer coef0cient and overall RHR seal water caoler unit -

conductance was inconsistent.- The calcuhtlon employed overly optimistic fouling factors and may have incorrectly assumed a fully developed pump _ head across_ the RHR pump for rnode J (Minimum Flow Bypass Mode) operation. Also, failure to satisfy RHR pump seal qualification temperatures during mode J operation was not adequately addressed.

-7 Control Room Habitability Considerations DAEC had a series of calculations performed to determine the habitab!!ity of the control room-

.under :off-normal conditions. For the loss of the chilled water coolers, ts.s cases were examined.

The Orst involved 1007c makeup air from outside, and the second concerned using the Standby Filter Unit for makeup. In thir tatter case, outside air makeup is only1000 cfm compared to nearly 16.000 cfm fo' the former case. In both cases circulating air is maintained within the control building complex, thougn at a somewhat adjusted distribution.

Unlike the ESW flow reqtiirements for the control room' chiller cooling, the heat loads in the.

control building calculated in 466 M 008 are employed '(r.ee Section 4.2 -of this report).

~

Calculation M.010 gives the results of the two cases examined and M.011 supplies:a validation _

of the siniple computer code employed in M.010 4

This computer code is a transient code but lacks the effect of heat slabs. Instead it performs simple mass and energy balances throughout a multiple control volume arrangement. It uns the -

4.

15

l l

\\

capacitance and fic,w distribution of the air with the sources of energy in each control volusae to determine an air heat up rate.

All three of these calculations lacked administrative controls since they were not signedEby a reviewer.

The assumptions and approach appear to be on the conservative side, primarily because of the -

lack of heat slab capacitance consideration. %c results of the heh are N concern, however, since in the case of using only standby Siter unit makeup air, the control maan temperature would rise at a rate of about 2*F per minute. Based on the assumptions used in the calculation, this would continue unabated. This would not allow time for a controlled shutdown of the plant. Derefore, allowing one chiller system to be inoperable during plant operation may need to be restricted.

8 Conclusions The review team believes that the presert calculations do not support the proposed technical specification amendment. Problems were found in the following broad categories:

Fouling factors, The fouling factors used for the ESW flow requirements appear to be non-realistic for DAEC service use (Sections 3.2, 4.2, 5.2, 6.3).

Methodology.

Son e calculations employed an approach which involved too -many unknowns (Section 5.2), or did not sufficiently document the calculational results to assure-that the approach was adequate, as was the case in employing the Wilson Method (Section.t.2).

Assumptions. In many cases, assumptions ustd in the calculations could have been verified and hence removed as assumptions with a request for information to DAEC or vendors (Section 4.2).

16 I

As. built concitions Our field observations ga/c us the impression that the calmlerInn preparers did not perform walkdowns or request cordirming information from the Seld (Section 5.4).

Equipment qualification (EO). In at least one calculation (RHR pump seal cooling), an argument is made that the EQ Temperature envelope for the seal can be exceeded F.

a short time, Supporting justification for this condition was not addreseed (Section 6.2)

Administrative controls. A revision to one calculation, and all the calculations dealing with control room habitability, lacked checker signeff (Sections 4.2,6.2,7).

Computer code validation and verification. The documentation for the pump room cooler coil computer code was validated but the review team did not have sufficient documentation to assure verification (Section 5.3).

Within the constraints of this limited review, we ' nave tried to clarify all the problems associated with this calculational series. We have not discussed the distribution of the ESW total flow.

Clearly, the eppropriate flow to each ESW serviced component would need to be assured, including consideration for uncertainty of flow measurement and variations in intake structure water level 9 Re ferences 1.

Mineck, D.L. Letter to J. Murley of 6:28/91 on Docket 50 331.

2.

Dwg. I1905649, Colt Industries Heat Exchanger Asembly Vender Drawing #343M1508A3 of American Standard Ir.dustrial Division.

3.

Roshenow W.M., et aL Handbook of Heat Transler, McGraw-Hill,1973.

4.

Letter W. Lederhouse (ITT Standard) to J. Olson (Bechtel), dated 8/30/90. (Chron 34N)5) and Letter W. Lederhouse to J. Olson, dated 9/05/90. (Chron 34617).

i 17 h..

-i -