ML19340A513

Eccentric Rod Burnout at 1,000 Psia W/Net Steam Generation

ML19340A513
Person / Time
Site:	Dresden
Issue date:	04/06/1959
From:	Levy S, Polomik E, Swan C GENERAL ELECTRIC CO.
To:
References
GEAP-3148, NUDOCS 8008070700
Download: ML19340A513 (61)

Text

__.

Cincering 38Dort GEAP-32kg

& - /d 9'

• )yMZe c Ecli.dk Comm Jgan.s.

4. Le debb*

w

.2. 2/ 4 /

dh )

b l de / 4,,ena,,,/ pi Ol I

s ECCENTRIC RCD BURNOUT AT 1000 PSIA WITH NET STEAM GENERATION By S. Levy E. E. Polomik l

C. L. Swan A. W. McKinney April 6,1959 y'[

p

~'

~,

l

.f 1

~

~'

' k(

GENERAL ELECTRIC ATOMIC POWER EQUIPMENT DEPARTMENT T.1OfS:iN 10 kEGCLA!0tiY 0 it.

r SAN JOSE, CAllFORNIA 80080 70 7GPO

[cgylA0lY DDCXET fliE CDPY

(/,sgolw A : <ew

• w nde...c /-

f/W1f/ r GEAP - 31h8 ECCENTRIC ROD BURNOUT AT 1000 PSIA WITH NET STEAM GENERATION By S. Levy E. E. Polemik C. L. Swan A. W. McKinney April 6, 1959 u

}

e ECCENTRIC RCD BURNOUT AT 1000 PSIA WITH NET STEAM GENERATION By S. Levy, E. E. Polomik, C. L. Swan, and A. W. McKinney*

SUMMARY

i Burnout and pressure drop tests were performed with an eccentric rod geometry to simulate possible maldistribution of flow in multirod fuel assemblies of boiling water reactors. Data are presented for a uniformly heated rod, 0 5h0 inch in diameter and 8-1/2 ft. long, located within a circular pipe of 0.875 inch inside diameter. Test variables include one concentric and three displaced rod geometries, exit steam qualities from 9 to 66 percent by weight, flow rates from 0.25 to 1 34 x 106 lb/hr.ft.2 and system pressures of 1000 psia. Test results are as follova:

1.

Burnout heat fluxes with net steam generation are the same for the concentric annulus of 0.1675 inch and the eccentric flow spacing of 0.096 inch. Burnout values decrease by only 8 and 15 to 36 percent when the minimum flow annulus is respectively reduced from 0.1675 to 0.061 and 0.033 inch. The small decrease in burnout heat flux with eccentricity points to considerable transverse mixing.

2.

Two-phase pressure drops at various eccentricities are in good agree-ment, thus confirming the existence of transverse steam quality mixins.

• The authors wish to acknowledge the assistance of J. A. Kervinen and F. S.

Raczynski in loop operation.

u

3 Singic-phasa pras ure drop ttsts exhibit flev channeling, tho friction factor decreasing as the eccentricity increases. The single-phase results can be used to calculate ' actual' eccentric hydraulie dia=eters and to adjust the measured burnout values for the variation in length to diameter ratio with eccentricity. Corresponding reductions In ""rn-out heat flux over and daove those measured experimentally would amount to 3, 6, and lo percent for the three eccentricities tested.

k.

An analytical model is postulated to deter =ine the degree of transverse mixing. The model based upon subdividing the ficw area into two parallel channels indicates that mixing extends well over half of the flow zone.

It also points out that the measured total two phase pressure drop with heat generation agrees with predictions from accepted steam slip and two-phase pressure drop relations.

INTRODUCTION Steam distribution is seldom uniform at every cross section of a multi-rod fuel assenbly (See Figure 1). Uneven ~ power generation and different flow characteristics often produce local variations in steam quality.

In particular, the corner rod so=etimes exhibits the highest heat production combined with the lowest hydraulic diameter. These conditions could lead to higher steam quhlities or ' channeling' at the corner and the burnout heat flux there could not be specified without first establishing the degree of ' channeling' produced by the proximity of the corner rod to the fuel assembly channel.

Conditions at the corner rod can be approximately simulated by using an eccentric rod within a circular pipe. Various eccentricity settings can be utilized to reproduce corner rod to channel spacings and its variations with 2

f 9

O O O O O

<O VVVVVV O O O O "uet UVVU n,

A, n On n

O t

FU E L CH AN N E'.

FIGURE !

MULT! ROD FUEL ASSEMBLY

manufacturing toleranccs. Burnout testa have been pcrformed with such a simplified geometry and are described in this reports.

Burnout zesults with net steam generation in eccentric annuli are not available in the literature.

However, pressure drop and heat transfer results in an eccentric vertical annulus at pressures close to atmospheric are given in reference (1). Also, burnout measurements at 1000 psia for concentric annuli are reported in reference (2) for horizontal flow and in reference (3) for vertical flow. These and other available b;rnout data are reviewed in the next section.

REVIEW OF AVAILABLE BURNOUT DATA A ce=plete su= mary of burnout studies for water with net steam generation is given in reference (4). The data covered therein include test results in rectangular channels and round tubes at pressures between 500 and 3000 psia. Gen-erally speaking, burnout heat flux was found to depend upon the following variablea:

1.

Burnout decreases with increasing enthalpy at the burnout point.

2.

Burnout is essentially independent of inlet subcooling and mass 6

velocities below 1.6 x 10 lb/hr.ft.2 Burnout values increase slightly with flow rates above 1.6 x 106 2

lb/hr.ft,

3 Burnout decreases witn increased length to hydraulic diameter ratio (L/D). DoublingL/Dfrom50to100reducestheburnoutfluxbyabout six percent.

h.

Burnout heat transfer varies slightly with system pressures between 600 and 2000 psia, and increases slowly for preesures below 2000 psia.

Test data in the range of interest of this investigation have been repro-duced in Figures 2, 3, and 4.

Figure 2 shows all the available test results in round tubes at 1000 psi. F16ure 3 illustrates the corresponding data for rectangular

_4 1

SOURCE PRESSURE L/D G(LB/HR FT )

TUBF SIZE SYMBOL 2

s MIT 975-1065 PSIA 50 0.86 - 1.9 X lo O.i805ID X 9 INCH O

MIT 1000 52 0.021-0.073 0.1805 IO X 9.41NCH 0

ANL 1000 64.5 1.55 - 4.96 0.1801D X il 5/8 tNCH BETTIS 1000 67 1.23 - 4.71 0.187 ID X 12.5 INCH v

ANL 1000 76 0.42-1.64 0 306 ID X 231/4 INCH 8

2 e

• e e

a ese e

e ae e,.

=A,

e A

A,

Ag o's;*

v e*

, e'..

A A

.A e O

A AAA A

A A A

8 10 5

e se AA A A d'"

8 A

^^

8 7

a e

en a

6 5

4 O

3 e

O O

2 O

@ O O

O O

O gg l

l l

l l

l l

l l

l 1

0l l

l 4

5 6

7 8 9 10 20 30 40 50 60 70 80 90 100 STEAM OUALITY, WEIGHT PERCENT FIGURE 2 AVAILABLE BURNOUT DATA AT 1000 PSIA IN ROUND TUBES WITH NET STEAM GENERATION.

2

' SOURCE PRESSURE L/D G(LB/HR FT )

CHANNEL 4YMBOL s

BETTIS 800 PSIA 33 0.18-0.61 X lo 0.101 X l X 6 INCH v

BETTIS I200 33 0 18 - 0.704.

O.101 X IX 6 INCH a

BETTIS 000-830 153 0.2 - 3.6 0.097 X l X 27 INCH

+

BETTIS 120 0-1215 153 0.18 - 3.0 0.097 X l X 27 INCH A

BETTIS 800 242 0.19 - 0.8 7 0.059 X l X 27 INCH O

BETTIS 1200 242 0.18 -- 0.8 8 0.059X l X 27 INCH 2

y v v

s a

y A

+

a a v 0

4 y a

,A,

,+

+

A A

A+

A n

Y 4

A A

+

A

+

6 A+

^

10

+

s A

C O

A

+

AA +

A 4

8 A

4 A

0 A

1 A

O 4

e

+

0 m

6 S

a S

5

+

+

d4 o

D 0

3 e

O Cf 2

o O

5 I

I I

I I

I I

I I

I I

I 10 6

7 8

9 10 20 30 40 50 60 70 80 90100 STEAM OUALITY, WEIGHT PERCENT FIGURE 3 AVAILABLE BURNOUT DATA 'AT 1000 PSI A IN RECTANGULAR CHANNELS WITH NET STEAM GENERATION,

I l

1 SOURCE HEATER DIAMETER L/D FLOW RATE LB/HR FT 2 e APED O.375 INCH 217 0.2-1.d X los s

• APED O.540 INCH V.3 0.2-1.0 X l o 2

e e

• e s

e A

lo e

n e e t'*

4

!E 7

p e 4A a

^ ^

j

^4 4

A a a 4

a

1. a r '.

-i

^

5 j

..,e

..s m

,o^^Aa Q

4 a

a i

a A

e e

=

e e*

3

'e 5

l l

l l

l l

2 X 10 4

6 8

10 20 30 40 60 80 100

' TEAM QUALITY, W E'. J HT PERCENT S

FIGURE 4 AVAILABLE BURNOUT DATA AT 1000 PSI A IN ANNULI WITH NET STEAM GENERATION.

s channals and Figura 4 shows previously obtained data with a vertical annular geometry (3)

In order to amplify the abscissa scale of all three figures, burn-out heat flux was plotted versus steam quality instead. of enthalpy at the burnout point.

The most remarkable features of Figures 2 to 4 is the spread la experi-mental results.

In Figure 2, the system geometry and pressure are practically constant and most mass velocities are below the range where they affect burnout.

The deviation in test results may, therefore, be caused more by the experimenter, his loop, and his measuring techniques than by the controlling variables. Relative improvement in the spread is noted in Figures 3 and 4.

In each instance, informa-tion from a single experimental group is presented and in each case part of the scatter can be traced to variation in the length to diameter ratio. Yet, the deviations still exceed normal. They may be due to the neglect of an important correlating variable or to tre need of further improvement in experimental methods.

In the case of Figure 4, the latter reason mey prevail. As is pointed out later, considerable progress was made in eliminating the scatter of Figure 4 by innovations in test section design and burnout detection metheds. The data presented in Figure 4 were.obtained with the same size heated red and facility.as in the present work. The flav annulus was, however, maintained by means of slightly less accurate internal ~ ceramic spacers and the existence of burnout conditions was not verified for each point as in the present work. These modifications and other improvements are discussed at greater length in the next section.

DESCRIPTION OF EQUIPMEIR The test facility as depicted in Figures 5 and 6 is comprised of two stainless steel loops into which are inserted flanged pipe sections conto:uing .e '

& y- ;:

d

..:[Q p 4

3, q

N 7h [

b

&.z.

s

e. '

laki Tre dk

/

' " ~ '

~~

M 1

g t

L

. ---)

.e l

',*

• l '..

Ts

>g-t;

~;

~

.%g, ~s % v n

c ax

~'j "

} **/

Q4 a

FIGURE 5 HEAT TRANSFER FACILITY l

(g PRESSURE F t rl-F AN ONTROL CONoENsER VS TEri 7

/

~

f y-4-

/ / / / / / //,.f W

3" sea. 80 Pere

(-

I

-t-TWO SPEEo FAN - 5 H.P.

STEAM To con-2" SCH. 80 PIPE

\\

PRESSURC CONDENSATE RETURN PREsaunt TEMPERATURE CONTRot 1550 GAuoE GAvat i

SYSTEM LO*

RELIEr

\\

\\

r

-STEAM DRUM 4

.'i 1550 Sue-La COOLER RELIEF f

/ //// ///[-

(_-

~

STEAM W FAN

/

l

! TWO-SPEED r

)

DRUM 3

MOTOR 10 HP r

l

'70 GPid P UM P

---

A MAY DE ADDED FOR L

2" SCH. 80 PIPE FUTURE STEAM ENTRAlHMENT NATURAL OD 4

STUDIES VITH 9

~

CIRCULATlott N

ROD SECTION.

LOOP FORCED CIRCULATION LOOP ~

TEST ULTRA-SONIC ULTRA-SONIC SCCTlON l

FLOWMCTER FLOWMETER 15 - ROo y

y V

'v' I O G. P. t 1.

k" SCH. 80 PIPE I" SCH.

t 4

80 P l P E -

D 2

Ar

_,,A Yu J

.I" SCH. 80.=lPE

(

CHARGiflG PUMP l

s x

I i

y s

TO atowoOWN TAM:(

ILL1MG P Ull P FRc: DtMINERAtizEo WATCR STORAGE FL OW D I AGRN.1 B_E.s.T._TBN!SFER A. Fl,QlD. FLOv/ TEST FAC f L I TY FIntIt'r L j

f

electrically heated test sections. Steam generated in either loop passes to an air-cooled condenser where it is converted to saturated liquid before being returned to the main circuit. Subcooling is achieved by removing saturated liquid from the loop, passing it through another air cooler, and returning it to 'the loop upstream of the test section. A suitable transformer with an induction regulator for voltage control provides electrical power to the test elements. A control panel placed adjacent to the loops contains the necessary indicating and recording instruments and controls.

Test Loop T::e test facility consists primarily of two interconnected stainless steel loops, one intended for testing with natural water circulation, the other for testing with forced circulation. The entire assembly is designed for pressures of 1500 psia and te=peratures up to saturation. The experimental cystem includes the following:

1.

The primary loop where water is circulated in a closed system from a nine foot vertical steam. drum, down a four inch townce=er, on to an electrically heated tect section, and back to the steam drum.

2.

A packless canned rotor centrifugal pump installed in the forced circulation loop to provide a 92 foot driving head at 150 gpm.

3 A fin fan air cooled condenser which will remove up to 4,270,000 Btu /hr.

to condense 6;600 lbc/hr. of steam at 1000 psi. A ds=per in the steam condenser aje inlet is used to regulate the rate of steam condensation, i.e.,

loop pressure.

4.

A subcooler system which bypasees a portion of the downcomer water flov through a finned fan cooler located adjacent o the Heat Transfer I

Facility. The a=ount of subcooling is automatically or manually regulated l

by a Bailey pnsumatic tcmperatura control system.

5 A bypass da-iaa*alizer used to maintain water purity at the megohm level. Process water temperature is reduced to 100 F in a cooler before it passes through a mixed bed denineralizer resin. An economizer reheats the demineralized water before entry to the sub-cooler pump.

Power Supply Alternating-current resistance heating is utilized to simulate nuclear heat production. A 12 KV power supply is fed into a General Electric induction regulator which provides plus or minus 53 percent variation of the supply voltage.

The unit, rated at 628 KVA, contains three single-phase regulators which are mounted in an oil filled, air cooled tank. Automatic regulation of load voltage is incorporated in the equipment to compensate for variations in line voltage.

The load terminals in the induction regulator are connected to the high vo;tage terminals of a 1780 KVA transformer.. Four vindings are available in each phase of the secondary transformer to obtain output voltages from 30 to 360 voltc in three steps. The windings of each phase can be parallel connected, eries-para.'lel or series connected to vary the phase voltage. The twenty-four s

transformer tags are connected to terminal posts on a linkboard. From the link-board, multiple conductors carry the current to the test element terminals.

Test Section The test section is shown in Figure 7

• dater enters at the bottom of the test secticn and flows upwards in the annulus between an electrically heated tube of 0 5ke inch outside diameter and a ctainless steel pipe of 0.875 inch inside diameter. The heater tube, made of 304 stainless steel, is 8 ft. 6 in.

long and has a 0.0h9 inch wall thickness.

It is positioned within the outer pipe 4

1 1;_j 9'-O" REF.

FFECTIVE HEATING ROD LENGTH I20' 8'-6a 8'- 9=

1.050+

1.046 A

A s

D

~

n 7'-9h,,

.{

[

=Jj_. 73 m

3 ___

7'-6" T

~~

d HEATER ia ROD 6'-9 7

-tZ l--

i20*

n.

ta 3

S'-94 SECTION A-A SPACER j

A1

~~A n

y' 1]

,, PRESSURE TAP AND n

/

THERMOCOUPLE LOCATION 4'- 1 O"

/

i

/

/

STAINLESS STEEL PLUG 4 '-6 "

/

n SAPPHIRE 3'-9 f

,. <)ts %<

. _r B

B k

a g

f] =9 4

2'-6 A

A

{q".'].

9 p

- x> -

18"

6" CH 3"

L

~ - -

SECTION B-B y u o e

,__,, _y u o o u y

o v

lI f

b"

~

INLET i

w 2"

(

1 N

~

FIGURE 7 TEST SECTION

't by means of groups of three spacers shown in Figure 7 The three spacers are located one inch and 120 degrees apart. Additicnal single spacers are provided to prevent electrical shorting and to maintain the close spacing required on the narrow annulus or eccentric side of the test section.

Each spacer, as 'shovn' in Figure 7, consists of a sapphire rod 0.220 inch in diameter backed by a stainless steel plug.

Positioning of the rod is obtained by initially making contact between tha rod and outer pipe on the eccentric side. Reference readings are then taken 4

for all spacers from the extremity of the stainless steel plug to the diametrically outer radius of the external pipe. The two spacers on the non-eccentric side are next withdrawn an amount about 10 mils in excess of that required for the rod setting. The eccentric spacers are inserted the needed distance and tightened into position. The non-cceentric spacers are finally tightened. The position of the rod could be and was verified by measurements through the pressure tap holes.

These measurements indicate that the heater rod location was known within + 6 mils.

ltessure drop and water temperatures are measured at various levels along the test section. Curomel-alumel thermocouples in a 3/16 inch diameter stainless steel well are installed in the steam-water flew at the 0,1-1/2,3,4-1/2,6, 7-1/2 and 9 ft. levels. Pressure taps are located at the same levels. Each pressure tap is connected to e seal pot and the seal pots to a common pressure manifold.

Thermocouples are installed inside of the heater rod. Thermocouples made of chromel-alumel wire are splot velded inside the rod at various positions.-

Four thermocouples are located at the exit end of the rod while two more are installed 1 and 2 ft. from that end. The thermocouple leads are brought out of the test'section through the hollow bottom electrical conductor. -.-

1 Instrumentation and Control The loops are instrumented to indicate and record power supplied to i

the test section, system temperatures, pressure, water level, and flow. The instruments and readings are as follows:

i 1.

Two Brown instruments sixteen point recorder to determine the tc=per-atures in the test section, steam drum, downcomer, condensate, and subcooler.

l 2.

Barton liquid level transmitters to measure the liquid level in the

)

l steam drum.

3 An Ashcraft 2000 lbs. Sauge and two Heise gauges from 0 - 1100 and 600 - 1600 psi to determine the loop pressure.

4.

A resistance thermometer located in each main loop to sense the tem-j perature of the condensate and subcooler mixture.

5 A Bristol flow meter to measure the water flow leaving the subcooler.

6.

Orifices and Foster flow tube to obtain the amount of condensate returned to each loop from the main steam condenser.

7 Orifices and Foster flow tubes installed in the downco=er and cross over to the pump respectively to measure the loop recirculation rate.

8.

A recording vatt meter, a vatt hour meter, and an indicating volt 1

meter and nemeter to obtain the power input to the test section.

_EXPERDEITfAL ERRORS Burnout measurements are effectively represented by plotting heat flux versus atcam quality as in Figures 2 to 4.

The accuracy of the test results i

can thus be evnluited by obtaining the errors associated with these two parametera.

l ;

H:st Flux Uniform heat generation was used in all of the present tests and the heat flux at the burnout point can be computed by dividing the power input to the test section by its heat transfer area. The main errors in such computed values result from the inaccuracy of the power readings and the assumption that the average rod heat flux is identf. cal to the local one at the top of the heater rod.

The power input was obtained from a power meter. The meter is estimated to have a deviation of no more than 0, 1, 2, 6, 8, and 15 kw at scale readings of 0,100, 200, 300,140, and 500 kw respectively.

The differences between average and local exit heat flux can be traced to two causes. First, the temperature of the rnd varies axially (the variation is largest at high subcooling). Second, the rod thickness can change in the axial direction. Resistance measurements indicate that the maximum deviation from the average resistance is below 3% and the axial change in resistivity with temperature is below 0 5 percent.

For heat fluxes above 10 Btu /hr.ft.2 (~l400 kW) the maxir.m error is 5 7% if an additional O.2 percent error is allowed for tolerances in rod dimensions, i.e., heat transfer area. The corresponding uncertainty interval for a 95 percent confidence level in burnout heat flux can be obtained by the method of Kline and McClintock(5)

It is about 3 percent at the highest flux value of the present tests.

Steam Quality The exit steam weight fraction or quality is obtained from a simple enthalpy balance Q

(hg-h) i U

x=

(1) hfE

-lo-w

where steam quality by weight x

=

heat generated in rod, Btu /hr.

Q

=

flow rate, lbs/hr.

V

=

g enthalpyofsaturatedwater, Btu /lb.

h

=

t enthalpyofwateratinletconditions, Btu /lb.

h

=

hfg = heat of vaporization, Btu /lb.

The flow rate v is calculated from v = 0.1261+ (3600) C Il + " (T13 - 75) hT

_&A_1 M

Do V

Troom T1 V100 V

\\/

/CALB Tg Here C

=

D orifice discharge coefficient, dimensionless a

=

linearcoefficientofthermalexpansion,1/0F orifice diameter (measured at 75 F), inches Do

=

3 VTy = specific volume of fluid at flow meter temperature T13,ft/lb S

flow chart scale reading, dimensionless on a ' square' marked chart

=

So full scale flow chart reading, dimensionless on a ' square' marked chart

=

T height of water measured by ranometer across pressure transmitter at H

=

CALB calibration temperature (for any scale deflection S), inches 3

V specificvolumeoffluidattemperatureofcalibration,ft/lb

=

TCALB Az height between pressure taps across orifice, inches

=

3 VT specific volume of water at room temperature, ft /13

=

room The method of Kline-and McClintock(5) has been applied to Eqs. (1) and (2) to establish the un.artainty in steam quality. The error analysis for the test data of reference (3) is descr.ibed in details in reference (6).

It is based upon the following estimates of error in each of the variables in Eqs. (1) and (2).

Power Q:

the maximum deviation was given in the preceding section.

Pressure variables hf and hfg: Heise gauges with a one psi scale interval.

were used to measure the test section pressure. The maximum error is estimated to be below + 5 psia.

Temperature variables T13' YTg' h :

thermocouples at the flow meters and at i

the test section inlet have a calibration scale certified to be currect within + 4 F.

' Allowing for a 1 F error in cold junction correction, the 0

temperatures and corresponding properties are knownwithin 5 F.

C Flow meter scale variables:

the flow was recorded on a Bristol square root chart. This chart could be read to 0.1, 0.07, 0.05, t.ad 0.03 at chart values of 1.0, 3 0, 7 0, and 10.0 respectively.

Other variables:

their =aximum variation is summarized below:

D 0.0105/0 767 o

az O.0203/.293 C

3/0.625 D

V 0.00003/.1607 Tm V

0.00003/.1607 TCALB Typical uncertainty values in the exit steam quality values of reference (3) are reproduced below. These nu=bers yield a 95 percent confidence level (6),

Pressure Fass Flow Subcooling Power Exit Quality Quality psia lbs/sce Stu/lb kw Wt%

Error 1000 0.839 114 5 180 13 1 0.88 I

1000 0 796 35.h 122.4 16.5 1.60 1000 0.295 85 0 80 26.5 2 32 1000 0 353 69 9 100 30 7 2.03 i 2

The prLc.:dtng tabulation reveals that the uncertainty interval approaches at most 10 percent. The corresponding uncertainty interval in exit enthalpy is about 2 3 to 2 5 percent.

BURHOUT DEIECTION Rod element burnout was detected by a Safety Monitor System which causes 4

the main pcVer circuit breaker to trip when burnout conditions are impending.

The I

system relies upon an electronic circuit which monitors the balance between the voltages measured across two adjacent one foot sections at the exit end of the rod.

This voltage remains in balance during normal operation, at which time, a pair of thyratron tubes firing on alternate half-cycles hold a relay open in the breaker trip circuit.

Burnout produces a sudden increase in rod resistance. This leads to an unbalanced input voltage to the Safety Monitor circuit. The unbalance prevents one thyratron tube from firing which deenergizes the relay in the breaker circuit, causing trip. The sensitivity of the circuit to a veltage unbalance can be adjusted so that, in effect, the temperature at which trip occurs can be selected.

The circuit time constant is less thsn 8 millineconds.

Very natisfactory and reliable operation of the trip system has been obtained to date.

However, burnout detection does not stop with an effective trip circuit.

It was reslized early in the tests that the effects of rod eccen-tricity would not be established without eliminating all potential causes of

' false' burnout.

Several steps were taken to insure that the tests were giving true, repeatable burnout points for the specific conditions being tested. These steps are:

1.

Temperature Traces at the Burnout Point The Safety Monitor System can trip falsely whenever an electrical short occurs.

Sudden increases in water subcooling can sometimes have the same effect. The net result is that the recorded burnout heat flux is too low. To verify that the Safety Monitor trip was caused by a temperature rise at the exit end of the uniform wall rod a 30 gauge thermocouple wire was attached internally to the rod at that point. This thermocouple temperature was recorded on a fast response Sanborn instrument and temperature traces obtained at burnout condi-tions. A typical trace is shown in Figure 8.

The thermocouple reading remains practically constant as the input power to the heater rod is raised to approach burnout '. Just before burnout, a small change in power produces a very large increase in temperature. As shown in Figure 8, the temperature rise stops only when the power is turned off.

Temperature traces similar to Figure 8 can be used to detect electrical shorts. An electrical short leads to a power trip without a large rise in rod temperature. The rod resistance is rapidly unbalanced and the accompany-ing temperature rise lags before being propagated from the point of shortage to the recording thermocouple. Ccnversely, the Safety Monitor System could be used to discover defective operation of the burnout thermocouple. Repeated trips at the same loop conditions and enemalous temperature recordings are l

the warning signals. The Monitor and Sanborn recording thus complement each other very well. The use of both systems together inst'ad of singly as in e

previous experiments makes it possible to checa their performance and detect electrical shorts and defective thermocouples without loss of a test section.

All burnout points presented herein utilize thi s dual system of burnout detection. A monitor trip and a trace of the tjgj; shown in Figure 6 were simultaneously obtained for every burnout run.

800 POWER OFF 780 760 RUN NO. 55 u.

3 HEAT FLUX 429 x 10 BTU /HR FT2 740 W

FLOW O.747 LB/HR FTZ b 720 INLET SUBCOOLING 60.6 BTU /LB

<xe EXIT QU ALITY 31.8 w

$ 700 w

H

@ 680 m

J g 660 mw

$ 640 620 600 -

I I

I I

I 580 O

I 2

3 4

5 6

7 8

9 10 11 12 13 14 15 16 17 TIME, SECONDS FIGURE 8 - TYPIC AL TEM PERATURE TRACE AT BURNOUT.

i Continuous temperature recordings at the exit end of the heater rod offer several othe r advantages. They make it possible to determine the time and temperature range of a burnout point and to invest 16 ate the effects of heat flux and quality upon the mode of burnout. Before examining a large number of temperature traces, it is worthwhile to note that burnout condi-tions can be approached by varying any one of three variables (power, sub-cooling, or flow) while maintaining the other two constants.

Innediate burnout response is obtained by increasing the power; the response is some-what delayed for changes in subcooling or flow.

In all three cases, however, 0

i it was found that burnout was a sharply defined point. A 2 F change in sub-cooling at fixed flow and power settings would produce a sudden temperature rise as shown in Figure 9 (a). The temperature decreases if the subcooling is increased (about 1 F) and rises again to produce burnout when subcooling is lowered.

Four tracan are shown in Figure 9 The traces 9 (a) and 9 (b) represent fluctuating.igh ' film boiling' temperatures before burnout. These are representative of low heat flux conditions and decrease in subcooling before burnout. Figures 9 (c) and 9 (d) show the charp temperature rises obtained by increasing the power input at high heat flux.

2.

Qta Repeatability The first burnout point taken for a given rod setting was repeated at the end of each operating day. This served to establish not only that burnout l

results were repeatable, but also to verify that the test section geometry was relatively preserved. Typical data repeatability is illustrated in the following tabulation for a minimum flow annulus of 0.096 inch: :

i e.

P y OWER TRIP y POWER TRIP T1ME

~ONE SECOND

~

~ ONE SECOND

~

TIME

'l l l : : :

215* F 430*F

'_.m

~_

613

• F t 641*F 3

2 TRACE 9(a) RUN NO.59 ; HEAT FLUX 436 X 10 BTU /HR FT ;

TRACE 9 (c) RUN NO. 63; HEAT FLUX 663,000 EXIT QUALITY 30.4 PERCENT BTU / HR FT ; EXIT QUALITY 2

18.6 PERCENT POWER TRIP POWER 'I RsP

/

- ONE SECOND TIME

~

~

- +- TIME

I

: : : l l l l l

.l

: ' :-4 W l : ' '

~

t N

I p

~

- 86*F

- 215* F 637'F

~

)

}

L _.

J

~

-w i

TRACE 9(d)

RUN NO.42 ; HEAT FLUX 997,000 2

2 TRACE 9(b) RUN NO. 56 ; HEAT FLUX 422,000 BTU /HR FT ;

BTU /HR FT ; EXIT QUALITY EXIT QUALITY 48.i PERCENT 21 PERCENT i

FIGURE 9 INTERNAL ROD TEMPERATURE TRACES AT BURNOUT

Flow Subcooling Pressure Heat Flux, Exit Quality Run 1b/hrft2 Etu/lb psia Btu /hrft' Weight %

6 49 o.772 x 10 53,3 955 ggg x 103 32.1 6

53 0 765 x 10 60.0 998 439 x 103 31.8 6

55 0 747 x 10 60.6 1002 429 x 103 31.8 6

59 o.776 x 10 64.0 loo 6 436 x 103 30.4 3

Data Reduction Data reduction from day to day was utilized to determine test data deviations and the next runs settings. Additional and repeat runs were initiated whenever required.

4.

Hydraulic Stability Pressure drop across the test section waa recorded as the burnout point was approached. Ench power increase at constant subcooling and flow is accompanied by an increase in pressure drop. Examination of the traces determines the degree of hydraulic stability before burnout. Two typical traces are shown in Figure lo. Figure lo (c) corresponds to stable runs while Figure lo (b) shows the smallest degree of hydraulic stability achieved in the tests. Note that the pressure taps were all located on the eccentric rod side.

TEST RESULTS -- PRESSURE DROP Single-Phase Flow Pressure drop was measured for cold water flow without heat addition.

Test results obtained for the concentric rod geometry and three eccentric settings nre given in Table 1.

A typical set of pressure drop runs is plotted in Figure 11 for an o.096 inch minimum flow annulus.

It is noted that the frictional and o LOO g

i i

i i

x 90 z

uz

, 80 1 70 0

h SAFETY o

. g ;,j MONITOR 2 50 12 %., 3,.,,,..,..,,;,,...

TRIP DUE TO 2q_OY[YN{U!!,,Ep'f';'.pv"

~T BURNOUT o.

O 30 O

E 20 o

en y 10 m

I I

I I

I a

O O

2 4

6 8

10 12 TlM E, M INUTES IO(a)- RUN 67-ECCENTRIC ROD WITH O.033 INCH MINIMUM FLOW ANNULUS O LOO w

I I

I I

I I

I I

N SAFETY I

MONITOR z 90

..;FDh, ff_} "'~$lgt:

TRIP DUE TO BURNOUT

- 80 n',,..

. r.na

..;.~

......glN.).j!?,' ; d' A.i '..

r4 7,9;;,

; < pij.

p.

c..._

L 70

"-@ r-M'A' ;,4 7

S

- W M f' "*j,)f.i y y

.. i f 60 0

50 3

o E 40 n.

O 30 O

y 20 a

$ 10 w

I I

I I

I I

I I

I 0

O 2

4 6

8 10 12 14 16 18 20 TIME, MINUTES IO(b)- RUN 65-ECCENTRIC ROD WITH 0.061 INCH MINIMUM FLOW ANNULUS FIGURE 10- PRESSURE DROP TRACES ILLUSTRATING SYSTEM HYDRAULIC STABILITY

l 1

I I

I I

I I

I I

70 a

/

un 60

/

Eo

] c4 g*

- RUN NO.6 28 o R UN NO. 7 3 50

+ RUN NO.8 o

x RUN NO. 9 A RUN NO.10

$o I

3 0

~

+

~

t;0 s-u-u.

x

{ O 30

+

OF l

l s

/

f:/,/^

10 yh.

I O

O I

2 3

4 5

6 7

8 9

DISTANCE A1.ONG TEST SECTION, FEET FIGURE 11-SINGLE PH ASE PRESSURE DROP DATA FOR ECCENTRIC ROD WITH O.096 INCH MINIMUM FLOW ANNULUS l

TABLE 1 -- SINGLE-PHASE PRESSURE DROP DATA Minimum Inlet Run Flow Annulus Flov Temperature Pressure Reynolds Number Velocity Pressure Drop (inch H O) 2 2

inch lb/hrft F

psia non-dimensional ft/see o-3' o-6' O 9 '_

O 6

k 1

0.1675 1 39 x 10 81 65 1.87 x lo 6.20 27 54 8i.3 6

4 2

0.1675 1.n x 10 83 65 1 54 x 10 4.96 18.o 36 54.4 3

0.1675 8.33 x 106 93 65 1 30 x lo 3 73 10.0 20 5 31 7 4

4 0.1675 6 95 x 105 94 65 1.09 x 10

-3 11 70 15 0 22 5 4

5 0.1675 5 56 x 105 94 65 0.88 x-10 2.49 4.2 9.4 14.8 6

4 6

0.096 1 38 x 10 121 65 2 97 x 10 6.18 25 47 5 71 6

4 7

0.096 1.18 x 10 121 65 2.47 x 10 5 31 19 37 56 6

4 8

0.o96 1.07 x 10 121 65 2.23 x 10 4.82 16.5 31 5 47 9

0.096 0.86xlof 4

121 65 1.80 x 10 3 87 11.0 22 32 5 lo 0.096 0 73 x lo 121 65 1 52 x 10 3 29 90 16 5 25 6

4 23 n.

0.061 1 39 x 10 84 65 1 94 x lo 6.23 26.1 48 5 74.5 6

g i

12 0.061 1 39 x 10 92 65 2.14 x lo 6.21 27 5 48.0 74.5 6

g

~.061 1.13 x 10 93 65 1 79 x lo 5 05 16.8 32.0 49 5 13 o

14 0.061 8 37 x 105 loo 65 1.41 x 104 3 74 n.o 20 31.0 15 0.061 5 56 x 105 4

99 65 0 94 x 10 2 50 lo 15 1 139xlof 16 0.033 86 65 1 98 x 10 6.20 23 5 46.5 71 l'7 0.033 1.n x 10 90 65 1.66 x lo 5 16 15 5 30 9 47 5 18 0.033 8 36 x 105 94 65 132xlok 3 83 8.5 17 7 27 5 5

19 0.033 7 31 x 10 96 65 1.18 x lo 3 27 6.8 14 3 22.4 4

20 0.033 6.96 x 105 98 65 1.15 x 10 3 12 6.9 12.4 19 5 21 0.033 5 57 x 105 99 65 0 97 x lo 2.49 35 8.0 13 1 l

spacer lesnes can be approximated by means of strai ht lines. This is under-S standable nince the spacers are about uniformly distributed betvcen pressure taps and fullydeveloped flow is established beyond the 1-1/2 foot position.

If the spacer lessec are-ectimated by means of accepted contraction-1 expansion formulas,* the experimental results can be applied to calculate friction factors.

It is possible to calculate a friction factor for each of the tap reading i

as illustrated below for Run No. 7 l

l position 0 1/2' O - 3' 0 1/2' O - 6' O 1/2' O - 9'

, friction factor 0.0283 0.0271 0.0276 0.0275 0.0276 0.0274 The accuracy of the computed friction factor is expected to improve with distance along the test section. This is due to the larger pressure drop reading and the smaller effect of a + 1 inch error in the transmitters. For this reason only the readings from 0 to 9 feet were used in Figure 12 to determine the effect of eccentricity upon friction factor.

However, it is worthwhile to note the small deviation in friction factor-vith ponition for Run No. 7 It means that each of the O to 9 foot value plotted in Figure 12 corresponds to about 6 points. Exam-ination of Figure 12 reveals that the friction factor decreases as the eccentricity increases.

In other words, channeling cccurs with single-phase flow.

This result agrees with the test data of reference (1). A 30 percent decrease in friction factor was reported there for a 1/8 inch annulus of 1.08 inch inside diameter when the flow annulus was reduced by 30% on one side. The present' tests predict a i

smaller effect of eccentricity. Tentative friction factor lines have been drawn for each geometry in Figure 12.

In drawing mean friction factor lines special c=phasis was placed.upon the runs at high velocity (i.e., high pressure drop).

+ Spacer losses account for 22 to 7 percent of'the mcasured pressure drop 1ceses. i

I i

I I

I I

I e CONCENTRIC ROD A O.096 INCH MINIMUM FLOW ANilULUS

• O.06l INCH MINIMUM FLOW ANNULUS

+ 0.033 INCH MINIMUM FLOW ANNULUS 0.05 c:

O O.04 COMMERCIAL PIPE EQUATION u

,,, o----a g

o 0.03

,]b-(,D e

F

+

• ' + ' ' ~+ ~ *[_

~~~~___, N i

O.02

' ' ~

+,,,,' ~

0.01 6

8 104 2

4 6

8 105 2

REYNOLDS NUMBER 1

FIGURE 12-SINGLE PHASE FRICTION FACTOR FOR ECCENTRIC ROD

Further, th concentric rod line was selected to agree with the friction factor equation for commercial pipe, namely d

0.160/n

=

gg where N

Reyn las number, non-dimensional.

=

BE To verify that the channeling effects were preserved at high temperature, heated runs with high subcooling were utilized to obtain points in the Reynolds number range of 10. Properties were evaluated at the bulk conditions between tap 0 and tap 1-1/2', and tap 1-1/2' and 3'.

The friction factor as plotted in Figure 0.14 12 was corrected by the ratio )

to account for the variation of properties.

The ratic of viscosity was estimated by computin6 vall temperatures from the accepted Colburn equation for pipe flow.

It is important to note that the high temperature runs are in agreement with the cold pressure drop runs even though the results are based upon readings 1-1/2' apart and the properties vary along the test section.

Two-phase Flow Two-phase pressure drop runs were performed for all the geometries studied I

in single-phase flow. The data are su=marized in Table 2 and a typical set of measurements is reproduced in Figure 13 The effects of eccentricity upon two-phase pressure drop annot be as clearly defined as for single phase flow. The major reason is that the flow system is vertical and calculations of two-phase frictional losses require pastulating the steam slip in terms of quality. While the steam void data given in reference (7) at 1000 psia could be utilized, it is not apparent how these data could be applied to eccentric rod geometries without making an assumption about flow channeling. A preliminary estimate of eccentricity effects, however, can be obtained from F1 ure 14.

Here two sets of two runs as 6

30 -

_m O

Cm TOTAL PRESSURE DROP (FRICTIONAL + ACCELERATION + HYDROSTATIC + SPACER) m MEASURED FROM INLET OF TEST SECTION, IN H O 2

m m

m m

m u

u m

a m

m O

m a

m m

O N

a m

m O

O O

O O

O O

O O

O O

O O

O O

O O

d O

gg f

I I

I I

I I

I I

I I

I I

I I

O y

_l_'

x.

m O&&

,m

\\\\.

\\\\

• +o$

m

\\\\

(n Z w os 04 mmmm C

O s

CCCC y

m ZZZZ N

p ZZZZ O

F 4

,O O O O ee O$

M.

O m

2)

H M

Oo gm

.e o

4 5

2 m

2; s

g s

+

o

+

x e

O O

e 9

+

O

17 0 FLOW HEAT FLUX SUDC00 LING PRESSURE LD/HR FT2 nTU/HR FT2 OTU/LR PSIA

+ ECCENTRIC

/

15 0 - ROD nisiuuu

,. o.:

/

FLOW ANNULUS r/

8 0.096 INCH 0.766 n 10 348s103 71.0 996 1

G g"

O ECCENTRIC O

goo,,y,uum

/

O FLOW ANNULUS g

0.03 3 INCH 0.803:10 36S a l03

/j 6

78.0 999 y

12 0 9

a. 0.2

/e P

/

4yO 10 0

/

g O Z

.. o, y E

• Y o =

v/

/ [e

$y 80

+

/

8 g~

/[

= = 0,1 60 k-as0

/

y

/

dM If #

yy s 'fh AO

~

M O

/

f 4

,t' /

se0

+

20 m

O q

/

b U

J O

by2 FLOW HE AY FLUM SU BC00LIN G PRESSURE

- Y 3 18'MR 772 jiU/HR FTI 12 0 J y /1 B P9tA T*

p w 2

+ c0NcENTRio

= *0 4 g.

O R00 0.4 F 106 339,io3 56 6 1000

~

g O ECCENTRIC

,p 3

0.

@g 10 0

- R00 MINPMUM g

FLOW ANNULUS a

8. g.

006l INCH 0.45 108 330:105 eg,y icol g

W p

=0 3 j

g 80

.J4 y2

/ f' m

,, n,,

g,o,3 60 J

a= 0.I,, '

s. 0.2 j

o 40 -

Y, L

x=0 20 -

I,

= = 0. 8 "Qh sao O

O I

2 3

4 5

6 7

8 9

iSTANCE ALONG TEST SECTION, FT.

FIGURE i4 - EFFECT OF ECCENTRICITY UPON TWO - PH ASE PRESSURE DROP

= -. _

l t

TABLE 2 - 'IWo-PHASE PRESSURE DROP DATA Minimum Meauared Pressure IIeat Flux, Exit Quality Pressure Drop (inch H O)

Run Flow. Annulus Flow Subcooling' psia Btu /hr ft' Weight %

o-3' o-6' o-9' 2

2 inch lb/hr ft Btu /lb 22 0.1675

' O.473x lof 56.6 1000 331 x 10 40 5

-2.6 1.6 17 5 3

23 0.1675 1.22.x lo 282 5 986 899 x 103 91 15 1 34 7 84 6

24 0.1675 1.23 x 10 89 7 1000 513 x 103 16.1 17 5 59 5 165 6

25 0.1675 0 742 x 10 190.4 1003 493 x 103 18.2 19 10.8 40 5 6

3 26

.o.096 0 76 x 10 71.0 996 348 x 10 21.2 63 18.8 50-27 0.o96 o.963 x lof 70.8 986 450 x 103 22 3 14.

46 5 125 3

0 28 0.696 0.$ x lo 62.6 966 345 x 10 22.6 6.0 19 50 5 6

29 0.o96 1.23 x 10 78.6 978 671 x 103 11 5 11 5 37 5 107 6

30 0.061 0.838 x 10 95 8 999 465 x 103 24.9 63 23 5 73 6

3 31 0.061 0.419 x 10 52 7 985 330 x 10 44.3

-31

-1.0

~14 6

32 0.061 0 91e x 10 208.0 1001 602 x 103 13.8 7.o 33 52 6

3 33 0.061 1 31 x 10 250.0 903 792 x 10 4.8 17.o 59 5 loo 6

34 0.033 o.803 x 10 78.0 999 365 x 103 20.8 31 12.8 41 5 6

3 35 0.033 0260 x 10 80.1 987 227 x 10 50.2

-7 0

-21.0

-32.0 6

36 0.033 1.23 x 10 174.0 975 608 x 103 8.6 13 2 31.0 74 5 6

37 0.033 0742 x 10 34.1 992 273 x 103 21.1 30 u.3 35 5

)

-nearly similar as possible are plotted. By examining the pressure losses in terms of steam quality it is noted that in both instances the pressure drop is slightly lower as the eccentricity is raised. The difference in pressure drop i.

-is, however, very sma*.1 and one can initially conclude that channeling for two--

phase flow did not increase over that for single-phase flow. A more detailed discussion of this point'is given in a later section. An analytical model is i

postulated there which relies upoa the data of reference (7)..The model which i

subdivides the flow region into twa zones is used to evaluate the accuracy of i

the steam slip data of reference (7) and the pressure drop relations of Martinelli-

}

Nelson i

TEST RESULTS -- BURNOUT l

Effect of Eccentricity I

j To investigate the effects of eccentricity burnout tests were performed i

at 1000 paia with a concentric rod and with three displaced rod scometries. The i

minimum flow annulus was successively reduced to C 096, 0.061, and 0.033 inch compared to a concentric flow spacing of 0.1675 inch. All test results are tab-4 ulated in Table 3 Burnout values for a concentric rod are shown in Figure 15 Coordinates t-of burnout heat flux versus steam quality, used therein, were found adequate for correlating the test data.

It is noticed that the spread in' experimental results is plus or minus 10 percent or just about within the maximum error range of the system. The test results also fall above the recommended design line of reference (3). It is worth adding that the concentric runs were performed last in this series of tests and further examination of eccentric data vill reveal why the

. number of runs was not larger.

4 34.

1500 1000 900 e

\\

800 700 e

y 600 o

h D

500 Ce e i

t e

400 e

N E

cm HECOMMENDED DESIGN LINE 300 x

REFERENCE (3) 3 e\\

u.

og tiw I 200 I

I I

I I

I I

I 100 6

7 8 9 10 20 30 40 50 60 70 80 90100 STEAM OUALITY, WElGHT PERCENT FIGURE 15 BURNOUT DATA FOR CONCENTRIC ANNULUS AT 1000 PSI A.

7-TABLE 3 -- BURNobT DATA Minimum 4

Run Flow Annulus Flov Subcooling Pressure Heat Flux Exit Quality inch lb/hr ft2 Btu /lb psia Btu /hr ft2 Weight I

38 0.1675 0.261 x 106 6

70 7 1003 263 x 103 61 5 39

.o.1675 0 337 x 10 60.1 1002 296 x 103 6

53.6 40 0.1675 0 559 x 10 63.8 loo 6 375 x 103 1

6 38.1 i-41 0.1675 0 909 x 10 60 9 1004 454 x 103 27.6 6

42 0.1675 0 997 x 10

-165 0 1007 647 x 103 21.0 l

43 0.1675 1.14 x 106 266.6 6

997 892 x 103 1h.8

-44 0.1675 1.19 x lo 54.9 1002 513 x 103 22 3 45 9.1675' o.955 x lo 82 3 1001 513 x 103 6

'59 2

46 0.1675 0,742 x 106 172.0 1012 514 x 103 23 1 47 0.1675 1.17 x 10 72 9 999 528 x 103 6

6 20.8 l

48 o.1675 1.16 x 10 71 9 1002 524 x 103 21 3

+9 0.096 0 772 x 106 58.8 955 448 x 103 32.1 1

50 0.096 1.07 x lo 126.0 1000 606 x 103 6

21.1 il o.096 1.19 x lo 238.0 993 840 x 1M 13 6 52 0.096 1.19 xlof 252.0 967 889 x 103 14 5 53 0.096 0 764 x lo 60.0 998 439 x 103 t

6 31.8 34 0.096 0 351 x 10 47 3 950 315 x 103 56.2 6

is 0.096 0 747 x 10 60.6 1002 429 x 103 31.8

~

36 0.096 0.422 x 106 83 8 1003 359 x 103 48.1 6

j 17' O.096 o.999 x 10 74.1 1023 505 x 103 24.8 4

38 0.096 1.17 x 10 184.0 1012 752 x 103 16.2 6

i9 0.096 0 767 x 106 63 8 1006 436 x 103 30.4 i

o 0.061 0.815 x 10 96.2 1001 479 x 103 6

27 3 il o.061 0.4k2 x 106 51,9 990 332 x 103 4.9 xlof 263 0 975 938 x 102 11 5 5

i2 0.061 1.29

3 0.061 0 922 x lo 216.c 1027 663 x lo?

18.6 xlof 264.]

825 967 x 103 A

o.061 1 32 11.1

5 0.061 1 34 x lo 294.0 755 1,040 x 103 6

9-3

6 0.061

.o.819 x 10 101.0 1000 471 x 103 25 6 7

0.033 o.811 x 106 101.0 1002 443 x 103 6

i 8

0.033 0 553 x 10 51 3 lolo 354 x 103 23 5 38.1 l

- o o.033 1.23 x 10 284.0 1001 616 x 101 85 0.248xlof 9

0.033 3

186.0 1014 275 x 10 65 5 1

6 1

1 o.033 0.809 x lo-208.0 1001 h46 x 103 24.2 2

0.033 o.98 xloy 296.0 780 771 x 10 to,o 3

i 3

0.033 0 98 x 10 299 0 79 5 771 x 103 6

10.1 4

0.033 1.15 x 10 25 3 1005 427 x 10; 22.6 6

5 0.033 1.1h x 10 107 0 1022 539 x 103 17 7 l

6 0.033 0.835xlof 194.0 985 580 x 103 19.8 7

0.033 0.14 x.10 151.0 1001 625 x 103 6

16.-1 i

9 0.033 0.81 x 10 30.1 1004 397 x 105 30 7

[

y

o.033 0 535 x 106 39 2 1000 348 x 103 40,5 i

0

Test results for a 0.096 int: ainimum flow annulus are plotted in Figure 16. The mean straight line drawn through the concentric data is reproduced here for comparison purposes. It is seen that excellent agreement is obtained between this eccentric geometry and the concentric rod case. For all practical purposes, no reduction in burnout heat flux was obtained by dis-placing the rod about 50 percent of the flow channel distance! Note again that the experimental data spread is small and below 10 percent. The repeatability of test results is also illustrated by the close grouping of four points at a steam quality of 32 percent.

Burnout-fluxes for a minimum flow annulus of 0.061 inch are shown in Figure 17 The mean line for the concentric and 0.096 inch geometries is also plotted on the same figure. Data for the 0.061 inch minimum flow annulus fall slightly below the previous results. A mean line tarough the runs falls about 8 percent below the line proposed in Figures 15 and 16. Test runs are once again correlated within plus and minus 10 percent.

The results for a minimum flow annulus of 0.033 inch are plotted in Figure 18. A greater departure from previous runs is noted and the deviation decreases as the steam quality increases. A mean curve through the experimental data can be drawn to obtain a correlation within plus or minus 15 percent. Closer examination of the figure, however, reveals that for this geometry, a flow effect can be detected. The mass flow rate for each run is shown in brackets in Figure 18.

One notes that the burnout heat flux decreases as the flow rate increaseo. This trend can be traced to the larger boiling lengths associated with increacing velocities. For the system of coordinates utilized in Figure 18 two runa at the same heat flux and exit quality have different boiling length if their flows are different. The higher the flow, the larger is the boiling length, l

1500 CORRELATING LINE FOR CONCENTRIC ROD DA','A 1000 900 e

800

$ 700 5m 600 a

f500 e

'k

.e e 400 5

's RECOMMENDED DESIGN LINE h.

E m

e REFERENCE (3) s N*

I 200 100 I

I I

I I

I I

7 8

9 10 20 30 40 50 60 70 80 90 IOC STEAM QUALITY, WEIGHT PERCENT FIGURE 16 BUF.NOUT DATA AT 1000 PSI A FOR ECCENTRIC ROD WiTH 0.096 INCH MINIMUM FLOW ANNULUS.

1500 s

CORRELATING LINE FOR CONCENTRIC ROD DATA 1000

• \\

AND FOR ECCENTRIC ROD WITH 0.096 INCH 900 s #e MINIMUM FLOW ANNULUS.

800

\\ s a 700 s

5 N

CORRELATING LINE FOR O.061 INCH 600 N

-f g

MINIMUM FLOW ANNULUS.

4

\\'x 500

\\s 's u.

400 N N g

g k

s RECOMMENDED DESIGN LINE x

REFERENCE (3) 3

\\

l N

200 100 6

7 8

9 10 20 30 40 50 60 70 80 90 IOC STEAM QUALITY, WEIGHT PERCENT FIGURE 17 BURNOUT DATA AT 1000 PSI A FOR ECCENTRIC ROD WITH 0.06l INCH MINIMUM FLOW ANNULUS.

1500

% \\

CORRELATING LINE FOR CONCENTRIC ROD AND FOR 1000 s

ECC NTRIC ROD WITH 0.096 INCH MINIMUM FLOW N

ANNULUS.

900 N

800 0.98 X IOb \\ g N

6 700 (0.98 XIO )

N g CORRELATING LINE FOR ECCENTRIC ROD S

( 1.14 X IO )

N WITH 0.061 INCH MINIMUM FLOW ANNULUS.

h 00 (l.23 XIO )

(0.04 X10 )0 g 8

8 %

I N

500 M {[81XIO8 N

N (l.15 X 10 )* joAIX10 )%

8 8

400 e

e 6s z

(0.81 X 10 )8)\\

s (0.55 X IO eb b

(0.54 X IO ) N s

RECOMMENDED DESIGN

\\

~

(O 25 X IOh EQUATION-REFERENCE (3) s g

u.

N s k

W N

200 i

i I

I i

1 I

l I

I I

I I

l

,gg 6

7 8

9 10 20 30 40 50 60 70 80 90 100 STEAM QUALITY, WElGHT PERCENT FIGURE 18 BURNOUT DATA FOR ECCENTRIC ROD WITH O.033 INCH MINIMUM FLOW ANNULUS.

i 1.e. tne greater the possibility for channeling.

The overall reduction in Figure 18 in burnout heat flux is still relatively small when it is realiced that the flow annulus on one side of the rod is only one fifth what it used to be.

One must, therefore, conclude that consider-able mixing takes place in a two-phase mixtur'*.

This mixing accounts for the close

. agreement obtained in Figure 14 for two-phase pressure drop runs at various eccen-tricities. There is, however, no doubt than a limited amount of channeling occurs with eccentricity. Burnout occurred n the eccentric side. Thermocouple readings and examination of the test section verified it.

Figure 19 is a picture of one of the test sections and illustrates the localiced burnout area on the eccentric side. Further, as discussed in the next section, rod eccentricity produces a greater reduction in burnout heat flux than the one shown in Figures 17 and 18.

Application of Data to Multi-Rod Geometry In a multi-rod geometry displacement of one of the rods relative to another rod or the fuel assembly channel does not affect the system hydraulic diameter. This is not true in the case of a single rod displaced within a pipe.

The ' actual' or effective hydraulic diameter must increase to account for the lover friction factors obtained in Figure 12.

A larger hydraulic diameter in turn means a lower length to diameter ratio. According to the correlations of reference (4) this should lead to a higher burnout heat flux if perfect mixing The ' actual' reduction in burn'ut heat flux with eccentricity can be occurs.

j estimated by using the relations of reference (4) and the data of Figure 12.

It is shown in reference (4) that burnout varies as f3~ '

1 The hydraulic 1

diameters to be suosLituted in the exponent are obtained from Figure 12 by 1.16 assuming that the pressure drop varies as D at a constant value of G.

The i __ _

vf /

LA

~

,t t_ __ _

FIGURE 19 LOCALIZED BURNOUT ZONE ON ECCENTRIC SIDE OF HEATED ROD.

following tabulation gives the total estimated effect upon burnout.

Concentric Minimum Flow Minimum Flow Minimum Flow Rod Annulus o.096" Annulus o.06f Annulus o.033'

1. Actual hydraulic diameter

.o279

.o304

.0336

.o376 from Figure 13, ft.

2. Length to diameter ratio 305 28o 253 226 32-0.0012h o.6935 o.7146 0 7382 0 7625

k. Increase in burnout heut; 1.0 1.03 1.065 1.10 flux from L/D effect

5. Relative burnout heat 1.o 1.o o.92 0.64 - 0.85 flux from Figures 17, 18, and 19

6. Net relative burnout heat 1.0 0 97 o.86 o.58 - 0 77 flux (from 5 and 4 above) 7 2 (Minimum flow annulus) y,o o,57 0 37 o.20 Actual hydraulic diameter The above tabulation shows that channeling effect existed at all eccen-tricities. Also, application of the data to a displaced rod in a multi-rod assembly can now be obtained by utilizing lines 6 and -7 of the above table.

Comparison with Available Data The data in Figures 16 to 18 confirm several of the trends reported in reference (4). Burnout with net steam generation is independent of velocity for cass flow rates between o.2 to 1.2 x 106 lb/hr.ft.2 Inlet subcooling effects i

are also negligible.

In the case of the concentric rod the flow and inlet sub-cooling were purposely varied in runs 42, 44, 45, 46, 47, and 48. No systematic i

trend was noted in terms of flow, inlet subcooling, and boiling length. Further, the effects of system pressure are not too large.

In Figure 18, runs at 750 and 800 psi pressure agree with tests at 1000 psia.

A comparison of the present test data with available test results at 1000 43 -

)

psia at high L/D ratios is given in Figure 20.

Agreement is noted with some of the UCLA data at lov steem quality and some of the Bettis data at high steam quality.

Further comparison between the three sets of information is not warranted until a better understanding of burnout is obtained. Se"eral explanations can be advanced for the experimental deviations in Figure 21.

1.

It is possible that experimental techniques need improvement. Main sources of error in the authors' opinion are false burnouts, variation in test section geometry during operation, and experimental inaccuracies.

2.

Another possibility is that one of the controlling variables was not included in the correlations.

It is hard to believe that burnout heat flux can be a single function of exit quality.

It must also depend upon the flow characteristics at the exit end of the rod. Different burnout values for short and long heated rods tend to verify this. Two-phase pressure drop flow pattern, and steam slip may be expected to vary with heated length, heat flux and flow. Correlations to date may be inadequate because they.cannot account for these variables.

3 Another area of uncertainty is the role played by the test loop.

Devia-tions arising from different investigations indicate that the loop system must affect the results.

In reference (9) the effects of loop geometry were underscored. A very stable circuit with very high head pump could lead to higher burnout heat fluxes than a low head loop.

ANALYSIS -- ECCENTRIC RCD Method of Solution An analysis was undertaken to determine the degree of mixing in an eccentric rod geometry. The basic model used in the analysis subdivides the flow 3,000,000 2,000,000 B TTIS DATA "g

(*f ECTANGULAR CHANNELS UCLA R

L/D*153 DATA IN TUBES 5 1,000,000

. s,, N.

eL/O=109 ta.

r I

[

800,000

\\

\\

\\

s 600,000

\\

g

{

PRESENT DATA

\\

l g\\

[

y IN ANNULUS L/Ds305 g \\ \\l l

.J 400,000 N N'

\\

300,000 N

~

BETTIS DATA ISI RECTANGULAR CHA8tNELS L/D= 242 L

I I

I I

I I

I I

I I

800,000 I

2 3

4 6

8 10 20 30 40 60 80 800 STEAM QUALITY,WElGHT PER CENT FIGURE 20- COMPARISON OF PRESENT RESULTS WITH AVAILAB'LE BURNOUT DATA.

OUTER PIPE O.875 INCH I.D.

HEATER ROD O.540 INCH 0.D.

\\444\\

l

/ -

/

/

FLOW

/

N'h\\

REGION 2 (N

' NT N N b:

\\

n FLOW D

REGION i

\\NNNN FIGURE 21 ECCENTRIC ROD GEOMETRY-SUBDIVISION OF FLOW AREA

area into two parallel channels as shown in Figure 21.

The pressure drop in the flow direction is assumed the same for the two channels and the flow rate and steam quality in each channel is accordingly adjusted. The distribution of steam quality within the fuel assembly is then obtained step by step in tae flow direction. The equations are derived in reference (10) and are reproduced in Appendix A.

Thei can be solved once the channel geometry and heat transfer characteristics a e specified.

The calculations are based upon the following assumptions:

1.

Two-phase pressure drop was obtained from the correlation of Martinelli and Nelson.

~

2.

Experimental void data of Larson at 1000 psia were utilized in the comuu-tations. Corrections have been made to allow for voids in the subcooled region in terms of fuel rod heat flux.

j 3

The friction factor was assumed to be the same for the two flow channels l

and to remain constant in the flow direction.

4

)

h.

The change in mass flow rate in each channel was neglected compared to the change in quality over the small incremental length used in the i

calculations.

}

l 5

The pressure loss due to spacers was apportioned over a finite interval l

vhich can be specified.

Results Several cases were computed on an IBM-650. A typical set of results l

1s shown in Figure 22 for a minimum flow annulus of 0.090 inch *. By varying the OFor simplification purposes, spacers were left out from this preliminary analysis.

Later calculations indicated that their presence accentuated the flow maldistribution. -

i 0.9 0.8 REGION 2

'f REGION I 2p O.7 STEAM OUALITY O.6 1

GlON l z

p =lT/9 9

p = 2Tr / 9

/

y p =n/3

/

E O.s

1. = 1T/2

$g HEAT FLUX =200,000 BTU / HR FT2 y

PRESSURE = 1000 PSI A

-0 FLOW = 0.5 LB/SEC U

.4 FRICTION FACTOR = 0.037 h

MINIMUM FLOW ANNULUS = 0.090 INCH o

O

$O.3 w

H 0.2 COMPLETE MIXING CURVE O.1

/

/

/

p = TT / 9

1. = 2TT/9 0 -

/

g g,', #

STEAM QUALITY

/r =TT/ 3 IN REGION 2 g..l/2

,e i

1 1

I I

I I

O I

2 3

4 5

6 7

B OISTANC E FROM BOTTOM OF TEST SECTION, FEET FIGURE 22-ECCENTRIC ROD AN ALYSIS - STEAM DISTRIBUTION IN 1 'RMS OF ASSUMED MIXING.

size of the two parallel flow channels the degree of mixing can be estimated by comparison with the experimental data. The analytical model in Figure 22 indicates that mixing extends well over half of the flow area if the steam quality in the eccentric zone is not to produce burnout. Also, it appears that transverse mixing extends over the entire cross section and that only the transverse flow time lag accounts for the reduction in burnout. This is all the more apparent when the rodel is utilized at higher eccentricities. For instance, for a minimum flow annulus of 0.061 inch the maldistribution of flow is about 3061arger and the results point to a greater and greater mixing zone.

Pressure drop curves corresponding to Figure 22 are shown in Figure 23 Examination of the predicted two-phase pressure drops shows that they are not highly dependent upon the assumed flow geometry. The analytical model can, there-fore, be effectively used to evaluate the relations of Martinelli-Nelson and Larson by comparison with the experimental pressure drop results. Such a compari-son is shown in Figure 24 for the concentric rod data. The comparison between tests and prediction is very satisfactory, especially when one realizes that the analysis is based upon smooth pipe friction factor (about 0.016). When the measured single phase friction factor (about 0.024) is utilized, the deviations in Figure 24 are practically eliminated. Test and calculated values for the latter instance are given in Table 4 and the discrepancy between measurements and predictions falls below 10 percent.* The analytical model thus confirms the validity of the Martinelli-Nelson and Larson relations.

• The analysis does not allov " r variation of the friction factor with subcooling and is based upon a friction factor calculated for saturated conditions.

49 -

700 p = U/9 p = 2TT /9 p = TT /3 m

$ 600 p=U/2 s

3 d 500 HE AT FLUX = 200,000 BTU / HR FT 2 b

PRESSURE: 1000 PSIA

[

FLOW = 0.5 LB/SEC

$ 400 INLET SUBCOOLING = 3O BTU /LB o

FRICTION FACTOR = 0.037 MINIMUM FLOW ANNULUS = 0.090 INCH s

/

+

9 i

D H

s

= 200 f

8 5

E S 100 E

I I

I I

I I

I I

I O

O 1

2 3

4 5

6 7

8 9

DISTANCE FROM BOTTOM OF TEST JECTION, FEET FIGURE 23-ECCENTRIC ROD AN ALYSIS -TWO-PHASE PRESSURE DROP PREDICTION IN TERMS OF ASSUMED MIXI!JG.

l 200 5

f 5

i l

[. iso

/

xg RUN NO.23 f

l

$ 160 160

+

2 X=0 3

/

/

8 I40

[

i40 280 e

f O

RUN N O. 25

[

(20

/

12 0 240

/

,/

O

[

y,g RUN N O. 2 4

[2 100

/

10 0 200

/

/

8 O

/

f o

so

,/

j Xs0 PREDICTION

[

80 16 0

/

?

/

/

7 o

[ 60 j

60 o,

o

/

PREDICTION I20

/

/

/

8

/

/

/

d 40

/

[

40 PREDICTION ao l

c 0

/

/

/

/

E 20 20

/

I 40 -

/

R o/

A l

I I

I I

I I

I I

I I

I i

I I

I I

I l

I I

I I

I I

I I

o o

o O

I 2

3 4

5 6

7 8-9 O

I 2

3 4

5 6

7 8

9 0

1 2

3 4

5 6

7 8

9 DISTANCE M EASUR ED FROM INLET OF TEST SECTION FT.

FIGURE 24 - COMPARISON OF MEASURED AND PREDICTED TWO-PHASE PRESSURE DROP

TABLE 4 COMPARISON OF MEASURED AND PREDICTED WO-PHASE PRESSURE DROP EASED UPON MEAS 1EED SINGLE-PHASE FRICTION FACTOR Run No.23 Run No.2h Run No.25 i

Pressure Drop Measured Calculated Measured Calculated Measured Calculated (inch H O) 2 1

0 - 1 5 FT 26.2 23 7 26 3 24.1 19 9 17 5 0-3 FT 51.1 47 7 53 5 k9 39 1 34 9 O

.4 5 FT 77 70 9 85 5 76 3 59 51.2 0-6 FT 106.7 95 2 131 5 130 82.8 72.8 0 - 7 5 FT 141 5 128 5 188 202 111 104.8 0 - 8.5 FT 173 255 129 7 0-9 FT 192 273 148 5

-f

i 1

i REFERENCES 1.

Stein, R. P. et al, " Pressure Drop and Heat Transfer to Non-Boiling and Boiling Water in Turbulent Flav in an Internally Heated Annulus", Chemical Engineering Progress Symposium Series No. 11, Vol. 50, 1954.

2.

Foley, D.

J., Batch, J. M., and McEwen, L.

H., " Boiling Burnout at High Pressure in Horizontal Annuli", AEC Reactor Heat Transfer Conference Paper No.12, November 1956. (Classified) 3 Galson, A.

E., and Polomik, E.

E., " Burnout Data Applicable to Boiling Water Reactors", ANS Pittsburg Meeting, June 1957 4.

DeBortoli, R. A., et al, " Forced Convection Heat Transfer Burnout Studies for Water in Rectangular Channels and Round Tubes at Pressures Above 500 PSIA",

WAPD-188, October 1958.

5

Kline, S'. J. and McClintock, F. A., " Describing Uncertainties in Single Sample Experiments", Mechanical Engineering, January 1957 6.

Goss, C.

L., "A Statistical Method of Error Analysis with Application to Burnout Data", GEAP-2057, January 1957 7

Larson, H.

C., " Void Fractions of Two-Phase Steam-Water Mixture", M.S. Thesis, University of Minnesota, 1957, 8.

Martinelli, R. C. and Nelson, D. B. " Prediction of Pressure Drop During Forced Circulation Boiling of Water", Trans. ASME, Vol. 70, 1948.

9 Lovdermilk, W.

H., " Boiling Burnout with Stable Flow in Tubes", presented at Working Seasion of National Heat Transfer Conference, Chicago, 1958.

10.

Levy, S. and McKinney, A.

W., " Calculation of Steam Distribution in a Multi-Rod Assembly on an IBM-650 (Code PARCH)", GEAP-3125, March 1959 APPENDIX A ECCENTR7C RCD ANALYSIS -- DERIVATION OF EQUATIONS 1

Consider two parallel flow systems with heat addition in the flow direc-tion. The flow distribution between the parallel systems is determined by assuring that the pressure is the same for the systems at any position in the flow direction.

The controlling equations can be derived and are shown below:

1.

Geometric Relations The following relations are valid:

A1+A2 (A1)

A

=

Sy+ S (A2)

S

=

2 4A D1=

l (A3) 31

^2 2=

(Ak)

D S2

~

(A5)

D = -

where 2

A Area of flov, ft

=

S Wetted parinater, ft u-D'=

Hydraulic diameter, ft Subscripts 1 and 2 are used to identify the parallel. flow systems, while no subscript is utiliced for the combination of the two parallel systems.

1 2.

Heat Balances 1

The quality distribution in terme of positica and flow rate can be l

specified frca heat balances.

' 4

l A G h dx dz*

HyQ1

=

1 y fg 1

or dX1 Hi Q1 dz Al G h

1 f6 (A6) and dx2 H2 42 (A7) dz A 02 h 2

fg d3, HQ (A8) dz AGh fg 1

H Hj + H2 (A9)

=

i HQ = Hy Q1+H2 42 (A10)

AG = A1 G1+ A2 G2 (All) with Q = Heat flux, Btu /hr. ft.2 2

G = Mass flow rate, lb/hr. ft hf = Heat of vaporization, Btu /lb 6

X = Steam quality by ve16ht H = Heat transfer perimeter, ft.

= Diotance in flow dirce tion, f t.

Integration of Equations (A6), (AI), (A8) defines the value of quality in terms of position Z.

=

8 Hi Q1

+

dz (A12) x1 AlG1hgg h,8 so

- Ms f

H2 Q2 dz, (A13)

+

x2 hfg A G 2 2 hfg

\\

l

• Note that for simplification purposes G is taken constant over the interval dz.

Calculations to date have indicated that this is a valid assumption.

l l i

v m s

x=

_ b dz (A14) gg AuMg 0

I!cre 8 represents the quality at Z = 0.

It is assumed that the hpg two parallel systems have the same initial quality at Z = 0.

Note that ohs can be positive or negative to correspond to subcodea or net steam conditions.

3 Pressure Equations The pressure drop is made up of frictional, hydrostatic, acceleration losses and spacer losses.*

~

fl Vf 2

1 1

1 dpl bD U+U

-R 28 D1 y

+

_f gl yf yg

'2 2

2 G1 X

d V

1 Yf(1**l)

G1 g

+

+

g g1 6

R61 (1-Rgy) 2g (A15) where f

= friction factor 3

Vf =specificvolumeofwater,ft/13 p

= Martinelli multiplier, two-phase to single-phase pressure drop R

= volume fraction occupied by Sas g

3 V

=specificvolumeofsteam,ft/13 f

K

= spacer loss coefficient

'4 hen x is necative, the acceleration term third on the right of Eq. (A15) is written 2

V I

G1 d

f g

, 1-Bgp

• Note that for simplification purposes G is taken constant over the interval dz.

Calculations to date have indicated that this is a valid assumption.

- 56 a

o v.

The friction factor f is part of the input data and must be specified for each computer calculation.

It can be obtained from 0.160/

(A16) f =

f>

vhere JIg absolute viscosity of water, lbs/hr. ft.

=

Rewriting Eq. (A15) yields 2

dp1 Vf G1 (g1 g1

-(1-x1)2 dR (1-x )

V

• 1 V

• 1 dR g1 y

6 6

d g

)

2D (1-Rg) dx

- (1-Rg)

V R V

2 1

y f g g Ig dx K ) + __1 dx 1

1 1

y R.

+

6 dz 2d:

Vf 1 v7 -

V,

( Al'/)

G s

In the above expression, the size of the interval dz was left open whenever a spacer occurs and must be specified.

A similar equation can be derive.1 for dp2, and substitutions made for d

dx1, du2 Putting dp1 = dp2 2

r i(1-x1)2 Vr.01 f1 p1 K1 HyQ1 Y

O (1-x1)

V f Gy gl g

___+_

+

5 2

f6 6

(k1"NG )2 d*1 k1~"6 )+ v-G 2D

01e l A 1

l h f

1 1

2 2x1 Vg x1 dR 1

1i Vr G2 flN1 K2 H2 42 Y

g3 f U2 h

'" T/

'R 2

(A18) dx1 I F ~ %j g

(2D1 2dLJ A2 hfg 8

~

gy (1-;:2)'

01:gg "g (1-xa) g 2x2 Y

X V

dR 1

I g

2 S2 n

(1-RGg) 0;'2

( 3?~N8) vf R R

^2 f

2 Q

G2 The function p anel Rg are available in terms of x.'

h.

Functions % and Rg The function % is obtained from the Martinolli-Nelson equatten.

ir l l

particular at 1,000 psia p = 1.0 x<0

% = 1.0 + k9 x 0 < x < 0.064 (A19) p = 2.2 + 30 3 x x > 0.064 The value of Rg is taken from the best fit of University of Minnesota data. At 1,000 psia x (0.25 x 1.058)

Rg

=

025 x + 1.056 - (1-x)

(A20)

Below x = 0.1 the following eepressions were used:

Rg = 5 26 x + j$

0 < x < x' (A21) 107 Rg = 2.00 x + ES

,- 32

<x<0 (A22) 107 (10()(2.00)

}

Rg = 0 x <

2 00 107 4

l vith x' obtained from x' (0.25 x' + 1.058)

= 5 26 x' + 22 (0.25 x' + 1.056) - (1-x')

10I l

5 spacer Lossec The spacer losses were obtained by means of accepted single-phase con-traction-expansion formulas. The single-phase losses were multiplied by the local value of the parameter

% to obtain two-phase losses.

4 6.

Method of Solution i

'The pressure drop and mass balance equations were solved on an IBM using a fourth order Runge-Kutta integration scheme i

i

_ 58 1

-.,