ML20006C463

Flow Measurements,Seabrook Unit 1,Seabrook,NH,Circulating Water Pumps.

ML20006C463
Person / Time
Site:	Seabrook
Issue date:	09/30/1985
From:	Hecker G ALDEN RESEARCH LABORATORY
To:
Shared Package
ML20006C460	List: ML20006C459 ML20006C463
References
NUDOCS 9002080068
Download: ML20006C463 (26)

Text

, gg, tga w > \- #

..u ..

,.y.g .;;ip,, f~ s ,

,m -

M- '

p,.:.a',

v c 4 ,

... .u. y y . e -- , ,

-ngng1 s t 4 + -

-.t r pr y s.-gg:--

4

7. r ..

n

_'1 .m J , 1 FLOW .wIASUREMENTSz y sp u + -

o ,

, SEABROOK STATION - UNIT'1 e I-'

• 1SEABROOK, NEW' HAMPSHIRE'

m. ,

CIRCULATING WATER PUMPS N!"g 7

. l

(

• 5

+ ;g ,

.n

, ., 9. ~

glh, .

1 1 4

1- 4 ff-k i_ , ij 4.

, ( TJ.i , ,

_p ,

yQ:

• t, #, ,'

Q .4 , ,

> ) [ ' ~, . . .

~y6 2, :2,

/ .

• ~ 4 4 ,e 5 , ,-

Sub:.'itte'd to -

x United Engineers & Constructors r ., .

s , ,

3cr e.- s s rt-%g,-. ; , , . . .

New Hampshire Yankee ' '

c '.

.o. ,, , .

4 F l1- y

_v, n; g x >

y,) .m ,

{

5 v t 4 I-

+

1 yem Septemkor 1985'

&, \

7 4. : 1, . .

D

.o., .

qy' ' '

e'-

E. i.,,..

s ,

4 6 ;"

qj,t[ , , George-E. Me:ker, Director .

w d y

'ALDEN.RESEARCH LABORATORY

.,,, , -WORCESTER PC'YTI;MNIC INSTITUTE

.. HOLDEN, *A!! ACE *'!E""5 n-M 9 . , ' t' . ..

3[ E' c9002090068 900131 .

• l '

PDR: ADOCK105000443 n o; ,t -'p:

. PDC A ew-Ql k. e%

b

_.7.i r t j

'(-

4 I 1

y 7 .

, ,q 1 ,

pa).. q ; '

Vj j ,,~ L

)f: $

AESTFACT

's: .

I rlow measurements were . conducted .by the dye dilution method at .the Seabrook

~

Nuclear Pever Station to evaluate the head-flowrate characteristics of. the

'three Unit'l circulating water pumps. . rive test points were taken at varying

-total; pump heads for each of the three pumps and for all three pumps cperating Q .

simultaneously. Pumps A and C had ., essentially equal performance, ,which 7 exceeded the manufacturer's specifications. Pump B. equaled the manufacturer's' specifications at low heads and exceeded the specifications at high heads.

The-maximum measured flowrate with three pumps cperating was 469,900 gym at a i total head'of 70.9 feet. The test data for three pump operation confirmed the-tests of the individual pumps. .6 uncertainty analysis. determined- the Iflowrate uncertainty to be 1.8 percent at the 95 pe ant con'idence' level.

t .- , . .

t

.I'

- \i 1..

1.

4 s

~. ! ,, M

_h g ..

e

J;

. o 3!I L

i sF 4

V :x .

3 m :7 _

y ,.

i;f W -

i

'm

'l t INTRODUCTION m.

E p riow measurements were conducted at the Seabrook Nuclear Power Station to evaluate the - head-flowrate characteristics of 'the three Unit 1 circulating water pumps. yive test points were taken at varying total pump heads for each of the three pumps and for all three pumps operating simultaneously. The tracer dilution method was chosen as the most suitable method of-flow measure-ment. The tracer chosen was a fluerescent dye, Rhodamine WT, and concen-s trati >ns were measured by . a fluoreneter capable of detecting concentrations Lless than 0.001 ppb. The concentration of the mixed water flow and dye was i

i+

less than 2 ppb, which provided sufficient measurement accuracy while main-taining a concentratien suf ficiently low to be undetectable by eye.

Dye injection was ace:mplished at the centerline of the exit of condenser C, and the injected flewrate was measured by the volumetric method. The dilution method requires that the trar:er be fully rdxed with the flow-to be measured to minimice uncertainty.- The discharge transition structure was the sampling location'since the mixing length available from the condenser (about 820 feet) .

was sufficient for essentially cceplete mixing. A Turner Designs Model l'O

. fluorometer was used in the'threugh-flow mede to evaluate the time required to achieve. steady state conditiens 'after the start of dye injection and a strip chart recorder documented the time histo.mf of the tests. Af ter steady state

- conditions were achieved, the fluoremeter output was recorded by averaging the voltage read by a~ digital voltmeter.

The flowrote was calculated, based on continuity of dye, as the dye injection flowrate time's'iihe injected cencentration, divided by the final concentration ,

minus any . initial- background concentration. Field plots of the pump head--

versus flowrate were made to check the outeeme of the measurements.

a This report describes the test measurement procedores and instrumentation,-

L describes the calculation of flowrate, lists the test results, and provides an l ' .' uncertainty analysis.

4

i'

]

i p,

I

~:

r

:.11 . <>

byg*

_SN -

2 s ;g}

F T1,0W Mr.ASUREMENT Prine'iples of the Dye Dilution Method The dye dilution method is . based on the mass - balance technique. A t small g g quant $ty of fluorescent dye at high concentration is continuously injected at-

/

a' measured constant rate into the test flow. The concentration of the fully mixed flow is determined by measurement of the intensity of the-fluorescence..

The. ratio of the injected concentratien to the final concentration, minus any background cencentration in the inconing flow, multiplied by the injection flowrate equals the fully mixed test flow.

I '

E l I 9

• 0 (C -C ) III o 3 where

^

Q, = :njected flowrate (gpm)

C.

• Injected concentration (ppb)-

CO = Mixed concentration (ppb)

C3 = Background concentration (ppb)

Qg = T1owrate to be measured (gpm) 1

.The tracer may be any conservative substance detectable in small concen-o .. e trations. A Tonvenient' tracer is a fluorescent dye, Rhedamine WT, which is -

h '

. detectable in concentrations as low as 0.001 ppb 'using standard techniques. .

Rhodamine WT h'as low adsorption characteristics and is supplied at nominal- 20 percent concentration by weight. For the flowrate to be measured the full strength dye was diluted - to a - 2 percent concentration. The mixed concen-tration at the discharge - transition structure, about 2 ppb, was chosen to assure ' sufficient - measurement accuracy while maintaining a concentration o

essentially undetectable by eye.

l' '

i fj' _

~j

'< y

"W i KW; pg:g' b;nJ W

> +. >

,- e ,

, .i. -

< \ F.  : ^ ~ ,

• IffeU -9' O!\[i 'l 4 ba f: ._I~ ,;;V{-]N - [, ' E , aj: i j

n

'-g ,

v e. , , ;; ( bm .: p ~ - ^ -

c f;jp 4:' . a:.,

( ,

3

,; v.

d

.).

1 m y >; , - -

1

+ i

, -- gF- '-

j; n

T , , .

, riuorescence ils.. a 7 function' 'of J sample ' temperature c andG sample < temperature 9 3.^ t"- . "

svariatiens?frem'.thefcalibration temperature.are accounted for by: L

, o 7, . -

3 v

, im 8 .' \ F.- 1

.c e "  !

1. ' K 4

, x '

c = CR - e (TR-Tc) -

(2)

, , + '

m g,,

yt J.where-n

' , '1

??

.,4 m

1 I

' I[ I ,

r n ,

, C =; concentration . (ppb),* 1 i' . .<

f

7 , . c.; Cf=apparentconcentrationat-temperatureTg (ppb) .-

N, .o T-e = calibration' temperature .('r) -

e

• n; a , K =. teinperature correcti,on coef ficient (1/'r)  !

f:

r i

s J. .fhe temperature coefficient, K, used was 0.0144/*T, which isla. standard value.

.fcr Rhodamine'WT.

3.

' p'. ; 4 i t e

/ 1 4 3

%f.

rp,' o k'..,'  ; e s'

., i- K 'r s _' ,y

- f 'i Q f,

,ao bi-l lk' t , )ei f I

,%. \ -'

trL m

.,o

3 ;

i

. 4 4 $

' ; k, ,4 :n v

d

.4 ,;,

'c-'

e

'c h ; "

'.7-)  ;

5 l-d14

+

d ..p l 1 2

7 432 [ i* #.. .

i e .. - r ij3

7 C3; 3

J]O,i' . 3 s- 4 c jr: i fSl }r

- rw ,

y ,, ..

W,a '

a tg L (.-

4;< _

+.

, - ' 'lu ' '- ,5Q

• I

., i

=

k'5,. i s q .; ;.s + 4 6L'

[.' r,., Mp =  : J' pya;;.:l, lg:^ .-

..s. .-- . . . . l}.Q! -

4 ,, 4 - [-

y&gk]C 1$$0d' ?'([ > . ' . '

4

nstramer tr.; n Oescriptien A Turner Cesigns Model 10 fluorometer was used to c'easure the dye concen-tratier by fluorescence. The flucremeter has multiple ranges to expand the range of concentration which can be measured. Two range settings are avail-

{

atie, X1 and X100 having a 100 to 1 effect on sensitivity, sensitivity can be changed in each range from X1 to X31.6 in four steps, having maximum 30-fold effect en sens:tivity. At 2 ppb the fluorometer weuld read in the upper One third cf the maximum sensit:.vity scale on the X10 range ensur'.ng good resolu-tien and a wide range of readable concentrations in the range of linear re-spense. The fluorometer may be cperated in either the through-flow moce or the grab sample mode by a miner' change. The threugh flow mode was used exclus;vely for this test work. Figure 1 shows the fluorometer and attendant readeut equipment set up in the through flow mode.

?

+ ,,  %

5

- f+ s el '

. - :.* n:_..

f '*

'e d

A q . . . . . . ':

(

} , e

y. . .

j}gQfg

. _A ; m . . .~. .. . . ... .

_f,.

, A k

FIGURE 1 FLUOROMETER IN THROUGH FLOW MODE WITH DIGITAL VOLTMETER,

. RESISTANCE THERMOMETER AND STRIP CHART RECORDER

.e

.3-. 1 lit -

h%f

, y .

S-i h A digital - vcitmeter - (DVM) was used as the primary readout device to achieve r

= greater resolution than the standard panel meter supplied with the fluorome-

- ter, full' scale on the panel meter was equivalent to two volts on the DvM and 4 .

the DVM had a resolution of-0.001 volt. The transmission characteristics of the primary filter in the fluoremeter changes slightly with temperature, b therefore, a platinum resistance temperature sensor was mounted on the filter.

s , A similar temperature ' sensor. was mounted in . a 1/8 inch diameter ' rod for-h, measurement of through flow temperature. A third sensor was used to measure ep M . the calibration solution . temperature during calibration. Resolution cf the O' temperature readout was 0.1 T.

Dye was injected at the centerline of the cendenser C discharge via a- f our

foot- long one inch schedule 160 pipe installed in an instrumentation m .

- penetration TE 60182. The primary dye injection flowrate was low, about cne ec/see, so that a secondary dilution flow was required to carry the dye into c

the pump flow rapidly. A 2 gpm secendary flow was provided from the condenser s

inlet waterbox.

.k;' ,

Primary dye injectien was acccmplished . by a three roller fit.xible tubing q- eenstant displacement pump. The pump was driven by a stepper motor controlled by a signal generator modulating a power supply. The dye reiease rate was centro 11ed via the signal generator frequency to achieve a mixed concentration greater than 1 ppb at all flowrates. Dye injection flowrate was measured by.

the volumetric method. The injection pump and a.100 mi pipette <with reduced area measuring stations' were . supplied frem a 20 liter mariotte. vessel, to maintain a constant inlet pressure on the injection pump. When the large-a vessel was shut off via a valve, dye was supplied to the pump from the 100 ml-

- pipette. A> timer with 1/100 second resolution was started and stopped as the e meniscus of the dye passed the : measuring , locations. As the measuring lo-

• cations. were in the small diameter tube, the meniscus moved rapidly, which reduced the uncertainty of the time measurement. The minimum time to inject

= 100 ml varied from 110 to 190 seconds for one pump operation and a minimum of two injection flow measurements were recorded for each test.

i i

. 4

w -

w.

6 1

4 L+ Sampling ef the mixed flow was accomplished just upstream of the discharge  ;

L -. . transition structure at the 8 inch chlorination line downstream of the .!

E i

isolatien valve. The B inch line was temporarily plugged and a one half inch

. sample'line was conducted to the fluorometer to obtain a representative sample. 4

- of the' mixed flow. This location was about B20 feet downstream of the in-jection . location and six bends were included between the injection and >

i sampling locations. The pipe length, including the effects of the bends, was determined from published data to be adequate to attain essentially complete  :

a

{ mixing of the-injected dye with the flow, i ,

l'  ;

Test Procedure -

j 4

[ tufficient preparatien time was available to prepare two initial calibration  ;

samples from the test dye solution fer use in the in situ calibration of the fluorometer. Two 1000 ppb solutions, prepared independently as a check,' were  !

c

! used for calibration solutien preparation with site water. The measured  !

j g concentrations of the calibration solutiont vere within 0,2 percent. To allow real time flow calculations, the fluorometer was calibrated daily just prior. ,

to test -initiation. The_ calibration system consisted of an insulated container filled with an accurately measured ' volume of site water, . a small

-centrifuga1' pump to mix the calibration solutions with the water and circulate the solutions . through the flucrometer, and the calibration solutions with r appropriate : measurement glassware. Precise volumes of solutions of - known concentrations were injected into the water volume to achieve a series of accurately known concentrations. The pump mixed the solutions thoroughly, o

which was indicated by,a constant fluoremeter reading. The temperature of the solutions increased slowly with time due to the energy input of the pump,'but F . the temperatures were recorded and appropriate corrections were made. Figure i 2n shows a sample calibration with the least squares best fit line used to ,

t calculate concentrations from fluoremeter readings. Also shown on the plot b , are the-residuals, that'is, the variations in percent of full scale of the Q

-individual data points from the best fit line. The standat1 error estimate of i

- the curve fit was 0.14 percent of full scale, j 1

1 t 1 i.

V *

.m ~

v 7

v- 7, y l l

.. )

i The output of the fluorometer was recorded by strip chart to document the time I i

. historyc ef ' each ~ test - and to assist in the evaluation of the mixing. :The l r

t- temperatures of the-water.and the fluorometer filter were recorded at various i times. .As dye injection commenced the time was logged. The dye injection '

flowrate was monitored during .the test by measuring the = time required to  ;

inject 100 ml of dye.. The dye temperature was recorded periodically.  !

) :.

r Prier. to recording data, sufficient time war allowed to achieve steady con- >

P>, ditions.

e The elapsed time required to achieve a steady measured concentration. j 6

L, from initial dye detection was less than one minute and at least three minutes .

p' were allowed to assure steady condi*tions. The minimum total injection time >

t'

-j was ten minutes. Twenty voltme'tet readings taken st 3 second intervals were ,y i

L averaged by calculator to evaluate both the average reading:and the variation. j five averages were recorded for each test. Temperature of the water and the

?

- fluorometer filter were also recorded with netations on the time of the test .

n- and other pertinent-conditions.

f: . o a

. .LO

~

1 2

< 3,

./

O' 1. 5 ._ .

- 'm

0. 5% ~ '

=

t!!

l w

,

• W.

W 1. 0 - - - e *

, g -

o

- t

f. 5 C - e d 2 .

B w

_ g L,, O. 5 O 5%

U t .

9+ .

~

g

> ' QM 't f f f !t f f 9  ! e f t  ?' !e f f I 40 45 1. 0 1. 5 LO q cAusnAr2DN YALtX ppb

. FIGURE 2 CALIBRATION OF FLUOROMETER

x. ,

4

( g A -

g'

y , +; v.

Tom -o

.a . --

8-TL 0 TEST RESULTS ll c-

[ "; The measured head data andLealculated flovrates are presented in Table 1 -and

{' the totel- head versus flowrate curves are presented in rigure13 and 4 with.

'least squares'second order curves fit to the data. The Figures _ include the -

' manufacturer's pump curve for reference.

M Table 1

_ ; .'g .

, 4. Pump Performance s

j; "Pressyre Pu$p.keuse, TotalHgad Flowrate g,

Pump Test psi

• Head Teet' Teet gpm

• L' A^ 1 41.7 +5. 99.9 112,500

_ 2 32.42 +4 80.4 149,300 P 01 28.02 3 +4- 70.7 -166,200

," 4 16.72 +3 .46.6 184,900 F, 5 38.22 -4 93.2 129,000.

i 1 >

B

1 19,72 +1 55.0 169,700 IJ ,

2- 29.22 -1 78.1 143,600'

'. 3 36.22 0 92.6 124,100 40'42' b 4 . O 101.8 100,600 l- 5- 25.42 -3 71.1 '150,100-L;

'C 11 ~19,65 -1 56.0 173',400 2 30.25 Oc .79.4- 148.300

-: 3 34.65- +1. 88.3~ -137,000 4 -39.95 +1 99.9 115,000-5 43.45 +2 106.6 >97,500 h-U '

'All 1- -25.95' -1 70.9- 469,900f 2- -- 30.45- +1- 78.9 437,800

_ 3 34.85 +1 88.6 404,000 Pr

• 4 40.95 +2 101.2 337,700 h;, 5

-43.15 ' *2 106.0 304'800

~

1~5 Pump pressure was measured with a calibra?*d pressure gauge connected

~ ;

to the pump' discharge vacuum . breaker downstream of the pump shutof f

?.: valve.' A salinity of 31 parts per thousand, determined ~to be

<- representative from the Normandeau Summary Document'of December 1977,

. 4

^ '

> 0 I.

l e 'L T.  :-l.1 e

4 y t ,

%'o; s:

.p . L . .; 6 a, +

a. -

s n ,

9 .,

+ '

't 3

resulted.in a specific weight of'63.'158 lb/ft , which was used to convert psi to; feet of water. .

i

, 2. The pumphouse water level reading.was obtained from the control' room.

j

3.- The total head was' calculated as the pump pressure expressed in feet

^

ll,

.plus 10 feet minus'the puephouse head plus the velocity h'ead-at the,

~

, pressure measurement.locatien.-

The test results n for 1 Pumps A and C f all . on a single curve and the - least

' squares second, order' curve' fit to the data for A'and C is shown in rigure.3.

. -l The lperformant.e. of Pump B was somewhat lower than Pumps A and'C. At a given

flow the pump head for Pumps A and C is, about 8 feet higher than the pump -

curve. The! deviation increases to about 12 feet at lower heads. In terms of'

'flowrate'at 'a given head, Pumps A and C produce about 10,000.gpm more than'the i pump curve atithe lower heads and the difference increases to about 14,000 gym 7 s

, at Ethe' higher heads. Pump B follows the ' pump ' curve ~ at heads less than . 80' l

": feet, but!is' above the pump curve at higher heads. '

,t The > three pump = tests -are presented in Figure '4 with the total measured flow t plotted against the measured head. -Tne maximum' flow measured was 469,900. gym I

.atta: head-of 70.9 feet. The' manufacturer's pump curve flowrate was multiplied" 'I

.by1three.and plotted-for;comparisen. At high flowrates, the. measured head is about 6 feet greater - than the corresponding .manuf acturer's curve. At lower flowrates, the variationLincreases'to about 9 feet.-

) >  ;

O)Y I

h n >

I[5 i

i 4

-$ : s I r

{'.'k- -

+

,7 7

p

. ij' +

E .

^

k, < 10 4-5,f '.~

i: 120.

i

\*

e _

[ t k . 100.

J.y ,

4

. s PUMP M 80.

~

CURVE g T

1 - .-

e g . .

y 60. __

d +

~

o - o PUMP A \.

^

PUMP S

40. __- +

PUMP C

20. __

u

[. '

0. ' l 8 I i I i I  ! I i

80. 100. 120. 140. 160. 180. 200.

FLOWRATE 1000 gpm i

FIGURE 3 SEABROOK STATION CIRCULATING WATER PUMP TEST DATA -

INDIVIIDUAL PUMP OPERATION o .

r I

~

m-

a w .

Q '. .

t*a t >

33

, '8 c.

e,  :

-' :, ( t f

-

: t .. 4

.. p L:s

.j r ,

, 120.- o, t . ,

l.? f ( -  ;

a- _

T a

.a

' ^

, ~ } 99. _.

l y

Hm r i

i. . . ,

i.' w ;

E> PUMP

80. -

r.

CURVE l 1

s, I

4 'I O '

4

.. ~'

b -<e -

L 60. '

-w_.

a -

THREE PUMPS a

(;U Q n3 i.. -

M 40. _

e-s ,) : .

c. 20. __

n b

-4 ;. .

p. 4 E

0.- ~' ' ' l' ' ~ i ' I 't i ' i l I '- ' i k

? 300. - _ 350, 400.- 450.- 500.

.F e FLOWRATE-1000 gpm 4

?

s

• 5.

[g.f~ FIGUREI 4 SEABROOK STATION CIRCULATING WATER PUMP TEST DATA 1: .

- THREE PUMP OPERATION .,

. he g ,

.s

. 1

_, 's a

j'

~

l r -

.. 12-

[ -i p i

_ FI.OW MEASUREMENT UNCERTAINTY Estimates of precision indices were made from the standard deviation of the t

, measurements / while bias uncertainties are estimated from comparative- tests and experience. The bias and precision components are propagated separately {

l from the measurementsLto the final result. The elementary error source uncer-l tainties for'each' component are combined by the root sum square U.35 ) method.-

Precision. uncertainty is estimated as the precision index multiplied by :the l, Student t' factor to achieve a 954 confidence level. The overall uncertainty-

{

of the result is reported as the sum of the bias and precision uncertainties.

1 e ,. . t

'b -

. 3; The dilution flow measurement method is based on the conservation of a tracer  ;

continuously injected into the unknown flow. The fully mixed concentration is  :

measured, and the ratio of the injected concentration to the mixed cencen- '

tration equals the. ratio of the total flow to injected flow. Therefore, the I uncertainty of the dye dilution flow measurement depends linearly- on the >

uncertainties due'to the injection flowrate measurement, the relative concen-tration . measurements , and the degree of mixing of the injected dye. at the- l sampling location. In this case, due to the length of the penstocks, the-mixing was relatively complete and the uncertainty,due to mixing was small.

Examination of the fluorometer output versus time indicated ' the- short term

- concentration fluctuations were about 0.5%, which would be characteristic cf

-'relatively-complete mixing. The precision index due to temporal fluctuations

' is estimated at 0.5%. A bias uncertainty of 0.2% was estimated.for incomplete spacial mixing. -In addition, uncertainties in the concentration measurement due to data acquisition and reduction are also considered.

Injection'Flowrate Uncertainty Injection mass flowrate was determined by the volumetric method, that is determination of the time required to inject a given volume times the density of the-dye.

t Sp*

4 9

3

hkS h

33 p>

s 13' -

1 L

E M g =V 1 p[/t g (1) n Bias uncertainties in the injection flowrate include temperature effects on the volume. determination, temperature effects on the density determination,

~

3 and- time measurement. Precision uncertainties also occur in- the time and r

L: . temperature measurements. From the average of - the measurements, the time (T precision.index was estimated at 0.20%. The maximum range of the dye tempera-ture measurement, .1.6'r, was used to determine a density precision index*of F 0.0154. A 15'r variation from standard temperature for volume measurement k> - results in a volume bias ' uncertainty of 0.03%.

A 2'r temperature bias results -

lE-in a 0.02s density bias uncertainty. .The time bias uncertainty was estimated

[. at 0.10%

e TABLE 2 - INJECTICN T1.0WRATE UNCERTAINTIES (%)

Elementary-Error Source Bias Precision t

Volume Manufacturer Spec.

o .

0.10 NA c Temperature (15'r variation) 0.03 NA Density 0.02 0.015

}; ~ Time- 0.10 0.20

Root Sum Square (RSS)'= 0.146 RSS = O.201

[.

l' O Concentration Measurement ib , '

The . concentration measurement uncertainty is estimated by evaluation of the m

fluorometer calibration uncertainty, the data acquisition uncertainty, and the data reduction uncertainty,

r. < 4 :

.J,

'4-ay.

9

y ,

& yo _.

9'

.- 14 3

Calibration ':ncertainty The' elementary. error sources for the calibration uncertainty include prepara-

. tion'. of the calibration samples, temperature ef fects - on the fluorescent activity, - and . electronic- noise and envirenmental effects on the instrumenta-tion. . All concentrations were norma'.ized by the injection concentration, so

, that no uncertainty due to the injection concentration' occurred.

g -c calibration samples were constructed with five serial dilutions, an overall -

r, , dilution of 1 to 2,000,000, to produce a 10 ppb calibration solution. For the ti i- .

g instrument calibration procedure, an' initial volume of 10 liters was circulat--

E ed through the fluorometer and 'v olumes of the calibration solution were added to attain - known concentrations. The calibration volume ne'azurements were subject to the same bias uncertainties as the injection flowrate volume uncer '

tainty , i.e. 0.1% f rom - manuf acturer's specifications and 0.03% from tempera-

, ture effects. Table 3 lists the bias uncertainty in the calibration solution

+ preparation and for the calibration precedure. The precisien index of the volume -measurements is dependent on the number of dilutions constructed. The volume precision uncertainty for a single measurement is estimated : at 0.05

- percent and the precision index -of a number of measurements is estimated as - 4 t the RSS : of the number of measurements made. The precision and bias uncer-

- tainties of the ' calibration solutien are estimated in the first part of Table 3 and are included-in the overall calibration uncertainty.

4 l

h e

-'4J

.e euf '-

h J: s  ;

~ ~

if[??TF,J' ~ '

p 4 E i- n

• F , ,' t

,n e ._- 15 4,. >

N 'i h TABLE 3 CALIBRATION UNCERTAINTY (4) r t ,

h e l Elementary Error source -Bias Precision }

calibration Solution ~ .;

.' Initial volumes-(5) measurements 0.10 0.11  ;

. temperature (15'r) 0.03 NA  !

t

[ Total volumes (5) measurements temperature (15'r) 0.10 0.03-0.11 NA L.-  :

i RSS = 0.148 RSS = 0.110 '!

F ',

Calibration procedure Initial volume (10) measurements 0.10 0.16  !

r s

temperature (15'T) .- 0.03 NA L Sample volumes (8) measure'mehts 0.10 0.14  !

temperature (15'r) 0.03 NA .,

riuorescence Temperature 0.04 -0.14 p . Environmental Effects (1.5'T) 0.15 -NA e

. Electronic Noise. NA 0.0B

• p n- RSS = 0.260 RSS'= 0.277-

.3 i

_ .  ; .,- '< Electronic noise :was evaluated by determining the precision index' of the F .cutput voltage from fully mixed calibration solutions at a ecnstant tempera- +

E ture.- The precision index of - the average reading determined for several' [

[n ' . series of measurements was 0.08%.

• i riuorescence-Jintensity is dependent on temperature and, the temperature i-E correction for fluorescence has the form i

~

c=c Re 0.0144 (Tr -Tc ) (2)

I s _

Therefore; the sensitivity of the fluorescence to a temperature change is- i

, 1.444 times the change in temperature in degrees T.- The temperature bias c . uncertainty due to e:aironmental ef fects on the thermometer is estimated ,at [<

. 0.03 T or 0.04%. The temperature precision index is taken to equal the Ly.

-i' resolution of' the ; thermometer, to.1*r, resulting in a temperature correction i r

~

~

1 precision index of 0.14% for fluorescence.

m4 s .

- la

-'6 l_

).

-~

3 .

y

(. o-*

• 16 4

9 Environmental ef f ects on the fluorometer are basically caused by temperature effects en the transmission coefficient of a critical filter. The trans-mission coefficient is estimated to change by 0.10% per degree. The filter terperature was monitored and during calibration the range was 1.$'T.

The calibration precision index is estimated at 0.26%. The precision uncer-tainty of a typical calibration curve (calculated from the standard errer estimate of the linear regression curve) was found to be 0.lEt of full scale, which is less than but in general agreement with the above precision estimate.

Data Acquisition and Reducticn Elementary error sources for the determination of the concentration include the effect of temperature on fluorescence, environmental effects en the fluorometer filter temperature, and eleet:0nic noise. The electronic noise v precision index was estimated previously at 0.05%. The bias uncertainty fer the maximum dif f erence frem the calibration temperature of 4.1'F in filter temperature, was estimated 0.41%. A precisien uncertainty due to the filter temperature variation was estimated at 0.06% due to an average filter tempera-ture range of 0.6"F. Sample temperature is measured during the test run and is normally relatively constant, a typical temperature variation is less than 0.l'r, The precision - index of the temperature measurement is taken as the resolution of the thermemeter, 0.1'r, resulting in a temperature correcticn precision index of 0.14%. The temperature bias is estimated at 0.03*T, resulting in a bias uncertainty of 0.04g. The temperature correction coeffi-cient, in Equation (2) is an average value. Laboratory checks have shown that the value is no more than 5 percent in error, so that for the small tempera-ture corrections used, about 2'F, the bias uncertainty in concentration due to the coefficient is estimated at 0.14%. Table 4 summarices the data acquisi-

- tion and reduction uncertainties.

4 4

.ns i

e, $

i 0

m

. ,, % 'O s

&:l [  :. . - ^17 e *L ~.

TABLE 4 DATA ACQ'UISITION AND FIDUCTION UNCERTAINTIES (%)

L Elementary trror Source Bias Precision t .,

g ' Environmental Ef fects (4.1'r) ~ 0.41 -0.06

,y Effeet of. Temperature on O 04 0.14=

p" . Fluorescence

[' ; "

n [ LElectronic Noise NA 0.08 is .m r .:r F ', , Temperature Correction Tactor

',' 0'.14 NA

'RSS = 0.412 RSS = 0.172

!s ,

4 2  ; Total Uncertainty y

p

/.h '

Table ,5 ! summarizes the RSS bias and precision uncertainties for the four' -F 4

componentr of the flow . measurements extent of mixing, injection -- flowrate,

' concentration - measurement,~ and data acquisition and reduction. The bias and.

i (precision uncertainties may be combined by assigning a-Student t-factor to the, precision: indices to attain the 95% genfidence: level. 'The number of measure-mentalfor.:each precision index were greater than 20, so'a Student t-of 2 was o assigned.. .

4 g /r ,

m :'., ' '

TABLE 5 ' SUP. MARY OF UNCERTAINTIES - DILUTION METHOD .. (%)

N

' ~

Source?  : Bias Precision

Mixing- 0.20 0.50-6; Injection Flowrate 0.146 0.201

' Calibration. 0.260 0.277

~

. Data Acquisition and <

.e Reduction 0.412 0.172

,h a- .RSS = 0.547 RSS = 0.630 d i' .' .

M The overall flow measurement uncertainty is 1.81, at the 95% confidence level,

a-

'i.e.,'0.547 + 2t0.630). i l

1 D Qt l 1

G 4.

}  : q:

X , ,

g@'."

.l APPIND

X A L

' Temperature . Probe Calibration Temperature. Themometrics,. Inc. 5-10 Thermistor Standard s/n 189. NBS 0

Standard >

Traceable calibration table.

a ,.

Sensitivity Approximately 5.0 ohms /0.01'c at 0.00*C n" .

1.0 ohms /0.01'C at 35.00'C l?! ,

(' ' OHM-Meter Data precision Model 258 s/n 2778 ractory calibrated 6/6/E5 Resciution 1 CHM k . Temperature NESLAB, Inc. Model F.7,I,8

< JBath -Rated stability of O'.C.1'c o < .

y -. Injection Station Readout Newport Q:104R s/n Be4684 g.. Digital indicater - controller 4

Resolution 0.1'r

[ Probe Omaga Ingineering Inc. Model PR-11-2-100-1/8-6 1/2E

.7 Dye ~ Temp' Platinum Resistance Thermemeter

- -Sampling Station Readout Cmaga Precisien RTD Readout

[ ,.

. Model 199-P2-A-X s/n 034582 Water Temp. ~ Cmaga platinum ETD element model T3105

, ' Probe Mounted in stainless steel well Auxiliary Cmega Model PR-11-2-100-1/8-6 1/2-E '

Probe- Platinum resistance Thermemeter r

l Dye Temp. 67.9 '77.5 63.4 Calibrated by PMB'. ,1 g . Water Temp 67.6 77.1 63,0- 9-13-85 it _ Auxiliary Probe 67.4 '76.9 52.B.

'4970: 4040

- Reference . CHMS- 3561 s.

i 'T 66.'96- 76.46 62.40 Lc .,

jr

^ -

hN Y.:

% f mw >

4,.. .ll - s yb,.-.i. 'f' ll -; < q:  !;

-.s.

OsQQ W %.t. >> . .- -

% o.

+ '

gg , ' -

.m ...m .

[pp/ gn , .N+

0',

r '( ',--.' i 1

7: #

~

APPENDIX B'

, . ' TIMER CALIBRATION

b

-Timer- Newport Digital' Counter - AE 80139

< c. ,

A A Laboratory:

4- ,, .< Hewlett Packard 5300A Counter-

, ' $.. , ' > standard  ; serial #1120A03702 . AE 60032

_-.g.

a I q

t IiL t w,o * > Laboratory . Standard calib' rated periodically sith' WT NBS- radioistation. '

?

g'NE) ' .-..' Calibration'

.p e ,

4,

.n L,'

w<

Newport 1 ;84.929 67.673 75.282 . 72.116'

@%. 'A.

~

.w HP ~ ~ .84,931 67.673 75.282: <

72.117

, i ._

t

_ . ,. . --? 4 5" -

,g  % ,.

% g 3p1' i 1 I

^

'_ i( . i

'"',M

,' 'y .

i, .

o 4

h 3

4 4-y 5

1

! - one -

) '.

.\

\ \

'? s.-; -

i.}*

k

• f 4-

,..yVJ.: :. k .

V; .~ . s _ '_ f s ' *'i'*

^ !:} }

'-  ;[.

os :.  :. K ,

.s .x&'

A c

' :9 v l * ':. (3

$f r

• g s 5 t

,;.Gr

., 7 i ,

.L, 3 s j

s.. ~ = ,

sM d A m

o,C . ..

t Unit d St;tes Nucl:ar R:gul0tery Commiscion J;nuary 31, 1990 l

Attention Document Control Desk Page 4 t

f

• I l l

e.

r.

ENCLOSURE 2 TO NYN-90029 Seabrook Station NPDES Permit Modification l

1 i

i a

f

.O j, I , UNITE 3 STATES ENVIRONZENTAL PRSTECTIEN ASENCY j .. RE010'N I

,, j - J.F. KENNEDY FEDERAL BUILDING,80STON, MASSACHUSETTS 02203 2211 January 18, 1990

, Ted C. Feigenbaum

-Senior Vice President-and Chief Operating Officer b New Hampshire Yankee Division Public Service Company of New Hampshire P. O. Box 300' i Seabrook, New Hampshire 03874-

• h Ret Seabrook Station NPDES Permit Modifications Permit No. NH 0020338 Dear Mr.'Fe'igenbaum Your letter of November 6, 1989, has been received and_ reviewed with the New Hampshire Department of Environmental Services concerning your request for modifying the referenced permit. ThisL permit will expire on July 5, 1990.

- Di sch a _r_g e _ _0 0_1_ :

EPA and the State _are in concurrence with you'r letter that there will be no discernible impact upon the environment by the ,

anticipated circulating cooling water flow rate of 720 MGD. It is recognized that this flow rate' exceeds the_ maximum permitted flow rate of 594 MGD although well below the maximum flow rate of

.1,187 MGD which had.been approved for two-unit operation. The rationale for this determination is presented in Attachment I.

D_i s_c h a r g e_s_ _ _0 2 2_,_ __0 2 3 ,__ _ a nd _ _0 2_4_ :

It'is agreed that the natural buffering capacity of-the diluent streams (Discharges 001 and 002) for the three internal streams

.(Discharges 022, 023, and 024) will insure that the receiving waters will be~ environmentally protected even though the internal streams may exceed the pH requirements of the receiving waters.

The pH limitations for these three internal streams can be deleted without jeopardy. Further, Discharges 001 and 002 (combinations of various internal streams) have pH limitations that recognize and' limit the pH prior to discharge into the receiving waters.

The experimental data justifying this conclusion is given in

. Attachment I.

Determination It has been determined that since these two modifications are of minor technical nature they will not jeopardize the environment

o- ,-

when Seabrook Station begins full operation. These modifications, therefore, will be incorporated into the next modification or reissuance of the station permit. EPA regulation 40 CFR 122.62 (a)(16) allows EPA to modify the existing permit in order to accommodate this type of new information when it is presented to the Regional Adminstrator.

As a cautionary note, the existing permit limits for Discharges 001, 022, 023, and 024 are still in effect until the permit has been formally modified or reissoed to include these requested changes.

Should you have any questions, please contact T. E. Landry of this office at 617-565-3508.

Sincerely, dQuW/6m#g Edward K. McSweeney, Chief Wastewater Management Branch cc EPA Compliance, Attn: Steve Silva EPA Permits Processing, Attn: Veronica Harrington NH DES, Attn George Berlandi NH DES, Attn: Charles Thoits i

4

g .

eu e

, )

l t

ATTACHMENT I l

l.. Di sc_ha rg eiOOl .

Paragraph I.A.l.m of the referenced permit defines the source  !

documentation establishing the circulating cooling water I effluent limitations. These documents established under the worst case basis that a flow rate of 1,187 MGD (at minimum flow conditions during low tide) and a discharge of 39 'F l above ambient (at maximum temperature discharge conditions) would not significantly impact the of fshore receiving waters.

When the operational conditions for one-unit operation in'the' current permit were developed, the 1,187 MGD was divided in half since only one unit was to be completed or 594 MGD. '

Actual pump tests showed flow rates that exceeded this value.

In retrospect, the value of 594 MGD as the maximum flow rate was found to be in error for three distinctly different reasons:

-i

a. The original 2-unit flow rate was based on low flow  !

at low tido conditions and it did not include the higher flow rate that would be induced by the increased Net Position Suction Head of high tides.

b. With all other parameters the same, the flow rate for one unit is more than one-half of a two-unit operation because flowing water friction within the tunnel is reduced by more that than one-half; thereby, producing a greater flow rate for the same pumping ,

equipment.

c. The motor / pump sets were found, during station tests, to have a greater pumping capacity than originally designed.-  :

Therefore, the combination of these three effects will produce a maximum flow rate of about 720 MGD. This value is above the permitted value of 594 MGD but well below the originally approved value of 1,187 MGD.

I The heat load imposed upon the receiving waters is exactly one-half of the two-unit heat load previously evaluated and approved. Since the heat load remains unchanged, the maximum temperature of the discharge during the high flow conditions will be reduced by a factor of 594/720 or 82.5% of the-low flow discharge temperature.

2. Discharges _0,2A ,0 L a,nd 024. I During station test operations and during review of operational n

.s

2 procedures, three internal streams (Discharges 022, 023 and 024) were found to exceed the pH limitations imposed by the NPDES Permit limitations and the pH Water Quality values for the receiving marine water envircnment.

In each case, these streams are combined with several other internal fresh water and salt water streams before being released through Discharge 001 to the Atlantic Ocean and through' Discharge 002 to the Browns River. Both of these streams have-Water Quality pu limitations that have to be met before being released.

The central question is: "Can the pH requirements for the

' internal streams be deleted and still insure that the human health and'the aquatic community will be protected?" The objective is to take advantage of the large buffering capacity of marine waters in the system to naturally neutralize the three streams prior to discharge from Discharges 001 and 002.

The buffering capacity of the several internal streams and of salt water itself is very great. This buffering capacity was verified by 4 simple laboratory experiments. The pH of two 100 m1 samples of fresh water was changed from 6.7 pH to 11.0 and 2.0 respectively. The pH of two 100 m1 samples of salt water were changed f rom 7.7 to 11.0 and 2.0 respectively. - Salt water was then added to each of the four samples through a titration burette recording the changing pH values. The end-results are as follows:

Sample T i__t r_a n _t_

a. 100 ml fresh 0 pH = 11. 0 65 ml marine resulting pH = 8.9

b. 100 ml fresh 0 pH = 2.0 800 ml marine resulting pH = 6.7

c. 100 mi marine 9 pH = 11.0 6,000 ml marine resulting pH = 8.5

d. 100 ml marine 9 pH = 2.0 800 ml marine resulting pH = 6.7 Therefore, it is practicable to remove the pH limits from the

.three fresh water internal streams (Discharges 022, 023, and 024) based upon these data since Discharges 001 and 002 contain salt water components and have pH limitations before release into the receiving waters. The pH of the internal streams will be neutralized within the plant system before the discharge mixes with the receiving waters. The very large salt water dilutional ratios for the discharges into the ocean (Discharge 001) and into the Browns River (Discharge 002) further insures rapid and complete neutralization for protection of human healir and of the aquatic community.

-