ML17138A533
| ML17138A533 | |
| Person / Time | |
|---|---|
| Site: | Susquehanna  |
| Issue date: | 07/09/1973 |
| From: | Becker, Koch PENNSYLVANIA POWER & LIGHT CO. |
| To: | |
| Shared Package | |
| ML17138A531 | List: |
| References | |
| E3-E2-2703, NUDOCS 7903150320 | |
| Download: ML17138A533 (124) | |
Text
m
-I Tj F- N as translatedinto ..~. V .G. I .I. S.H,... ~ ~ ~
~ ra
~ 0 J
~
~gn COi'sSTRUCTIOH Ai'f0 DESIGi't OF THE RELIEF SYSTEN 'i'/ITH Q
PERFORATED-P I PE QUENCHER '0 G
C
.1 iaa 63 lh Q
as translated from ~ ~ ~
I-0 Q
AUFBAU Ui'T'D AUSLEGUiiG NOTICE . ~
RECORDS OF THE THE ATTACHED FILES ARE OFFICIAL Ei THEY HAVE BEEN DIVISION OF DOCUMENT CONTROL.
CHARGED TO YOU FOR RETURNED A
TO LIMITED THE TIME PERIOD RECORDS FACIL'ITY AND ~- 'po MUST BE SRANCH 016. PLEASE DO NOT SEND DOCUMENTS CHARGED DUT THROUGH THE MAIL.
REMOYALOF ANY gM LV REPRODUCTION MUST, PAGE(S) FROM DOCUMENT FOR BE REFERRED TO FILE PERSONNEL
-'"'A.~" r."iQ ~=jj/~/ ~
D(
<<S 4IN DR, DEADLINE RETURN DATE $
43 IT'<@'~~ 4 iild<XIi)HIKE rn
~
i'tl
[ll5 - r; Q
~CD' 00. Qa\
C.'l N
9 J H -3 aa RECORDS FACII.ITYBRANCH taa (PP MS I '2 !
C P
S U .~
AUGUST 1977 V
~,
a
~ Ip TI
'/ ID J
~ ~
S'.
PEIiilq$ yLVANIA PO'PlcR aaL
- 5. LlGH t COf'iPAI'iY 71 BCnr:"ao PXVENEJE. WavER"OWN ga PILLfa'l I Oallla P ~ la 4 jS ~ alla ~ l rL"r SSA. HUE TTS QQI7$ (g;7 ~ c B4,SS
lg l
L ll I
PROPRIETARY INFORMATION This document has been made NON-PROPRIETARY by the deletion 'of that information which was classified as PROPRIETARY by KRAFTWERK UNION AG (KWU).
The PROPRIETARY information deletions are so noted throughout the report where indicated by a) Use of the term KRAFTWERK UNION AG PROPRIETARY INFORMATION.
b) Use of blocked out areas by cross hatch bands in the report text and figures/tables, e.g.
...." with a mass flow density of~W1 Kg/m2s...";
QM~~ mm iii) should be kept below ~~ w~~ atm."
iv) 8/l7/78
l I
I 5
Technical Report Construction and design of the relief system with. perforated-pipe quencher KWU/E 3/E 2 - 2703 Kraftwerk Union 5"1
s l
1
Frankfurt/Main 9 Jul 1973 Place Date Technical Report File number E3/E2-2703 Author Dr. Becker KNJ E3/E2 Dr. Koch Countersignatu"e~/si De artment l
Title:
Pages of text bB Construction and design of the relief Figures '40 system with perforated-pipe quencher Circuit diagrams Key words (max. 12) to identify the report's Diagr./oscillogr.
content: Tables Relief system, chamber steam injection, suppression Reference list Sugary To prevent loads on the pressure suppression system, the blowdown ,
pipes of the relief lines are equipped with perforated-pipe quenchers.
The construction and configuration of the relief system are described..
The requirements under operating and accident conditions are described. The following quantities, which are influenced by the use of quenchers, were investigated and the values to be expected
-forLoad them in the plant are indicated:
on bottom and walls of the suppression chamber during vent clearing and steady-state condensation
- Reaction
- Temperatureforces on the quencher mixing of the pool water
- Steam flow rate as a function of reactor pressure.
This is done in close reliance on previous and ongoing experiments in the condensation test stand in GKM and in the model test stand in the KN3 Nuclear Energy Experimental Facility in Grosswelzheim.
/s/ (Dr. Becker)
/s/ (Frenkel) t
/s/ (Dr. Melchior) /s/ (Dr. Koch) I
/s/ (Dr. Slegers) /s/ (Dr. Simon /s/ (Dr. Domin) II Author's signature Examiner Classifier Class
)
For information Distribution list: COMPANY <<CONFIDENTIAL (cover sheet only): lx KWU/GA 19 Erl lx /PSM 22 Ffm lx /E3/Library 2x /E3/El/LP Additional distribution according to attached )ist Transmission or duplicat.ion of thi" document, exploitat:ion or com-munication of its content not pcrmitt.'ed unless expressly authorized ~
Infringers liable to pay damage . All rights to the award nF'~+>>+~
l Distribution list (internal):
E 3 - Sekretariat E 3/V E 3/V 1 E 3/V 2 E 3/Y 3 E 3/Y 4 E 3/Y 5 E 3/V 4-QKT E 3/V 4 KKB E 3/V 4-KKP E 3/V 4-KKI E 3/Y 4-KKK E 3/E E 3/E 1
~
E 3/E 2 E 3/E 3 3/E 1-LP 2 x E 3/E 2-SA 4 x E 3/R E 3/R 1 E 3/R 2 E 3/R 3 E 3/R 4 E 3/R 5 E
Library 3/Bib) iothek HE/E F
I I
~
~
~
~
~
L
NONLIABILITYCLAUSE This report is based on the current technical knowledge of KRAFTWERK UNION AG. However, KRAFTWERK UNION AG and all persons acting in its behalf make no guarantee. In particular, they are not liable for the correctness, accuracy and completeness of the data contained in this report nor for the observance of third-party rights.
This reservation does not apply insofar as the report is delivered in fulfillment of contractual obligations, nor with respect to licensing authorities or the experts appointed by them.
KRAFTWERK UNION AG reserves all rights to the technical informa-tion contained in this report, particularly the right to apply for patents.
Further dissemination of this report and of the knowledge con-tained therein requires the written approval of KRAFTWERK UNION AG.
Moreover, this report is communicated under the assumption that it will be handled confidentially.
5 4
~
~
4
~
~
~
~
l
TABLE OF CONTENTS I
Pape I
Introduction 5-7 I
2 ~ Statement of problem 5-10 2.1 Function of the pressure relief system 5-1O 2.1.1 Relief funct'ion 5-11 2.1.2 Safety function 5-13 2.2 Operational boundary conditions 5-13 2 ' Permissible pressure loads on the suppression chamber 5-15 2.3.1 Pressure oscillations during vent clearing 5-16 2.3.2 Pressure oscillations during condensation 5-,16 2~4 Test stands 5-16
- 3. Construction of the perforated-pipe quencher and arrangement in the suppression chamber 5-18 3.1 Construction of the quencher 5-18 3~ 2 Arrangement of the relief pipe and quencher 5-19 Bottom load during vent clearing 5-21
- 5. Dynamic pressure load during condensation of steam 5-24 I
I 5.1 Survey of observed condensation phases $
-24 5.2 Condensation with small mass flow density 5-26 5.2.1 Condensation in the pipe 5-26 5.2.2 Condensation with subcritical out flow 5-27 5.2.2.1 Pressure pulsations with subcritical flow 5-27 5.2.2 2
~ Transition range to condensation in the pipe 5-29 5.2.2.2.1 Steady-state phenomena 5-.29 5.2.2.2.2 Non-steady-state phenomena '-31 ~
ig
'l 1
5.3 Condensation with large mass flow density 5-33 5.3.1 Pressure pulsations with supercritical flow 5-33 5.3.2 Steam bubble oscillations near the boiling point of the pool 5-34 5.4 Expected maximum pressure load in the suppression chamber 5"35 5.4.1 Intermittent operation 5-35 5.4. 2 Continuous operation 5-36
'5.4. 3 Circumferential distribution of the maximum bottom load during automatic depressurization 5"37.
6 ~ Reaction forces on the perforated-pipe quencher 5-39 Forces on the individual hole s-ao 6.1.1 Outflow of the water 5-4 0 6.1.2 Expulsion of the air 5-4 0 6.1.3 Expulsion of the steam 5-'4 1 6.2 Calculation of total forces 5-Jl 6.2.1 Vent clearing 5-)1 6.2.2 Steady-state blow-out of the steam 5-4 2 7 ~ Flow-rate capacity with the perforated-pipe quencher at reduced reactor pressure
- 8. Temperature distribution in the suppression chamber during relief processes s-a 6 8.1 Verti'cal temperature distribution 5-4 G 8.2 Temperature distribution in the circumferential direction of the suppression chamber 5-48 Tables Figures References 5-6
I I
!I I
I 1
- l. Entroduction KWU boiling water reactors are equipped with a safety/relief system with which large quantities of steam are conducted via quick-opening valves through the blowdown pipes and condensed in the suppression chamber. These processes have been found to be determinative for the design in regard to the dynamic loads on the pressure suppression system. To reduce the loads, the blow-down pipes have been equipped with quench rs beginning with the Brunsbuttel nuclear power plant (KKB) . This construction provide" the following improvements relative to the open pipe planned initially:
- The quantity of steam flowing down from one valve can be condensed through one blowdown pipe with a quencher connected after it.
" The dynamic pressures that occur in the water space of the suppression chamber are reduced both during clearing of the blowdown pipes and also during steady-state condensation.
- Calm condensation is made possible at high steam flow rates and high water temperatures.
To determine suitable quencher geometries, various designs were studied experimentally in a development program /G/. That test program was carried out primarily in the Mannheim Central Power Plant (GKH) at a scale of l:5 with respect to the flow rate. The, most favorable geometry proved to be the perforated-pipe quencher, 5" 7
l I
which was studied further in an optimization phase and investigated over a broad range of parameters /2/.
The purpose of this report is to describe the essential considera-tions involved in the use of a relief system with perforated-pipe>>
quencher in the reactor, to investigate the operationally relevant quantities and to indicate the values to be expected for them in the plant. We first describe in detail the function of the safety/
relief system using KKB as an example. The construction of the quencher, whose standard dimensions are al: o being used in follow-on projects, is illust ated. Zt should be noted that the outside dimensions of the quencher have been fixed. However, a few quan-tities having hydrodynamic significance and also the hole distribution and the diameter of the supply pipe cannot be specified conclusively until completion of the test program still going on for optimiza-tion of the remaining parameters.
The maximum pressure load on the suppression chamber during clearing of the blowdown pipes and during steady-state condensation is determined by direct transposition of the measuremcnt values fro...
the GEM test stand. Based on the previous test phase with the perforated-pipe quencher, maximum local pressure loads of l~l~~~M~
kg/cm were specified. The parameter combinations listed in.
Section 4 make it possible to respect those specified values.
The following problem areas are also discussed in this report:
Reaction forces on the quencher 5-8
I.
- Temperature mixing of the pool water
- Flow rate through the relief lines in the event of fast pressure relief for elevated backpressure in the suppression chamber.
These questions are examined partially by calculation and partially by experiment.
- 2. Statement of roblem The relief system consists of 7 main valves with connection lines.
These valves are actuated under accident conditions. One or more valves are opened simultaneously.
In this Section we first describe the operation of the system and, using a diagrammatic illustration, show under what conditions
(
xndividual valves are opened. He
~ \
also describe the operational boundary conditions under which the blow9own processes proceed.
Finally, we describe the load limit of the suppression chamber, which must be respected during blowdown processes.
2.1 Function of the ressure relief s stem Each of the main valves is equipped with two separate pilot valves of which one is for the relief function and one for the safety function.
In the safety function, the corresponding control valve is opened electromagnetically by the reactor protection system according to the following criteria.
In the safety function, the safety valves are opened with redun-dancy and diversity at a reactor pressure of MMM~~~~~> I Firstly, the control valve is opened electromagnetically for the safety function. Secondly, both control valves are opened under actuation by main steam after shedding of an additional electro-magnetic load.
5-10
The arrangement of the blowdown lines is illustrated, diagrammatically in Figure 2.1. From the Figure it is evident that the various groups of valves are distributed uniformly over the circumference of the suppression chamber. The three valves that are actuated in the event of automatic depressurization are also distributed over the circumference. Zn this way, adjacent valves are never actuated simultaneously and the pool is heated up uniformly at high thermal load.
I 2.1.1 Relief function Hain. valves are opened by one of their two control valves in the following cases:
<< Turbine tri out Because of the limited capacity of the bypass station SWAM RKK~~~~~~~~~~~Q, 1 valve opens forth.K~~Qjfrom a reactor power of MhM@of full load, and 1 additional valve opens for g~ seconds from a reactor power greater than~~~full load.
These two valves are designated by TSS in Figure 2.1 and are actuated directly by the Geamatic. The reactor pressure is thereby maintained at~~bar.
<< Hi h reactor- ressure Zf the reactor pressure rises impermissibly due to failure of the 5-11
control system or components (e.g., in the event of a turbine tripout caused by failure of the main condenser), then the relief valves open in a staggered manner in three groups in order to control such pressure transierpts. They open at the following set pressures:
First group KRAFTNORK,UNION AG PROPRIETARY Second group Third group INFORMATION The valve groups are identified in Figure 2.1.
- Hold at ressure and tern erature in the event of nonavailability of the main heat sink.
The decay heat after the scram is carried away by periodic opening of several valves in the firstQ~seconds and one valve subse-quently.
- I Emergency shutdown in the event of nonavailability of the main heat sink.
l r
The~reactor is depressurized by repeated opening of one valve manually in accordance with a prescribed'pressure variation in
~(hours.
- Automatic de ressurization in the event of loss of coolant In the event of a loss of coolant, one relief valve is opened automatically in any case in order to depressurize the system 5-12
gradually. This happens only at sufficiently high liquid level and, at the earliest,~~minutes after origination of the acci-dent criterion.
At a reactor pressure below~~bar,h~~~relief valves are opened in, the event of a loss of coolant in order to clear a closed II emergency-cooling loop (suppression chamber pump - reactor-suppression chamber).
Finally, the pressure relief system is used as a redundance for the coolant injection system. If the coolant injection system does not conduct a sufficient amount of water into the reactor when necessary, thenQ~~Qrelief valves are opened in order to bring the low-pressure emergency cooling systems quickly into operation by means of a fast pressure drop.
- Performance test Each valve can be actuated individually by hand during operation.
2.1.2 Safety function In the improbable case that several valves should not open in their relief function during a reactor pressure transient, the control valves open the main valves as described above in .the safety function due to a reactor pressure of hMWW~~~ bar.
2.2 Operational boundary conditions Zn the function described above, the relief system must satisfy 5-13
I I
I
several boundary conditions with respect to the reactor pressure control system and the action on the pressure suppression system.
The nominal flow rate of one relief valve ish KQt/h at a reactor I
pressure ofkMbar.* En accordance with the relevant standards, this flow rate is used as a basis for the calculation of reactor pressure transients tPat can be handled by the safety function.
H As the actual flow rate, the valve manufacturer expects a value
~ gV of~~~/h at a reactor pressure of~~bar. The design of the pressure relief system, and particularly of the quencher, is therefore based on a value of Q~t/h atK~bar.
Furthermore, the pressure relief system is so designed that the flow 'down to sufficiently low reactor pressure is determined by the relief valve, so that, a critical pressure ratio appears over its seat even when an accident pressure appears in the suppression chamber.
I The valve o eninc time must not exceedh~~~~with a dead time of
@~gms. However, shorter opening times down tohM~~W must not,,
produce any impermissible loads.
The nonuniformit of the water heatina in the pool should not I
exceed~+C, apart from regions immediately contiguous to the
- II t
1 Translator's note: for
~
The abbreviation stands a metric ton equal to l,000kg.
l l
I
quencher outlet. This difference must be respected even if the pool is heated by QgC with one or more blowdown lines during a
- blowdown process and simultaneouslyjQ out of ~3RHR pumps are in opeiation.
The'tern erature ran e of the ool water in the quencher's vicinity for,which the blowdown can proceed at full flow rate should be at least betw enQ~C and hQ'C. With a linearly decreasing flow rate, it must still be possible to pass through the temperature range of ~~WC tcQ+'C in accordance with Figure 2.2. This stipu-lation is based on the assumption of a mean pool-temperature of up t~QC for full flow rate and a maximum permissible mean pool-temperature of MQC and the above-mentioned nonuniformity of the water heating.
arms (see Figure 3.1) is Q~3m at normal water level in the suppression chamber. But the system must remain operational in the event of water, level deviations of ~~~ and W~U m. The latter might occur, of course, in the event, of a greatly lowered react'or pressure; s~ Figure 2.3.
2.3 " Permissible pressure loads on the suppression chamber A distinction is l
made between two types of loads which produce pressure loads on the suppression 'chamber during operation of the blowdown lines. Firstly, the expulsion of water from the line causes pressure loa'ds which, as the expelled air expands, cause 5-15
brief pressure 'oscillations at the bottom and walls of the sup-pression chamber. Secondly, pressure oscillations occur during steady-state condensation of steam, Depending on the discharge geometry and water temperature, they ar attributable to the con-densation process. Both loads should be limited to the following limiting values:
2.3.1 Pressure oscillations during vent clearing For a reactor pressure up to~+bar and for the specified valve opening times, the air oscillations at the bottom and wall should not exceed the values
+~~~~Q~~~~Q locally under the. pipe integrally over the bottom.
2.3.2 Pressure oscillations during condensation For the specified range of flow rate and temperature, the pressure amplitudes at the bottom and wall should be maintained belo~
+~~~kg/cm locally under the pipe h~Mkg/cm 2 integrally over the bottom.
2.4 Test stands To obtain empirical information relating to the various require-ments imposed on the'afety/relief system, numerous tests were performed in two test stands. The purpose of this report is to make inferences from the measuiement results to the expectation values in the nuclear power plant.
5-16
Large-scale tests on pressure relief were performed in the con-densation test stand in GKM at a scale of g~g with respect to the flow rate. That test stand is illustrated in Figure 2.4. Results are contained in /2/.
Supplementary measurements were performed in the model test stand (Figure 2.5) in the KNJ Nuclear Energy Experimental Facility in Grosswelzheir. at a scale up to approximately K~~~gBecause of its smaller dimensions, that test stand is much more flexible than the GKY. test stand. It is always preferable when large variations oc o the parameteis are necessary. Another advantage is that the dimensions of the pool relative to the experimental quencher are approximately ten times larger than in GKM and therefore more closely approximate the conditions in the suppression chamber.
Thus, information about the long-range effect can also be obtained there. Experimental results are illustrated in /4/.
I I,
I I
I, I
I I
I I
I
- 3. Construction of the perforated-pipe quencher and arrangement in the su ression chamber Zn this Section we describe the quencher that was designed for plant conditions. Although the construction and dimensions of the quencher are already stipulated, parameters that directly affect the quencher's operation and the loads on the bottom of the sup-pression chamber are to be considered provisionally as starting values for the confirmation tests in GEM. Those quantities will be stipulated finally only after conclusion of the GKH test program.
Table 3.1 contains a compilation of the most important data for the relief system with perforated-pipe quencher. The quantities not yet finally stipulated are provided with an asterisk in that Table.
3.7.. Construction "of the uencher The quencher with the connection for the blowdown pipe is illus-trated in Figure 3.1. The transition member between the blowdown pipe and quencher .arms is a ball. The angles between the arms are so selected that favorable installation conditions are achieved in the suppression chamber (see Figure 3.6). The quencher hole dis-tribution at the present stage of planning is illustrated in Figuie 3.1 and 3.2. The quencher has a total of b~~Woles which clear an outlet cross-sectional area of +~~~M Previous tests /2/ have shown that care mu t be taken to obtain a controlled inflow of water to the steam. For that reason, a uniform 5-18
distribution of holes over the entire area of the arms of the I
I quencher is not suitable, because the water would then have to flow in a direction opposite to the steam's blow-out direction.
Accordingly, the steam is conducted into the pool through hole arrays which are distributed over the+~quencher arms and Ql quenche" ends. The water can arrive at these hole arrays from all sides.
As shown in Figure 3.2, the<Q~mm-diameter holes within a hole array (as in GKY) are arranged in rows with a distance between hole centers oz~~~~M~ mm. WithQMholes per row, we obtain a total of 17 rows per hole array. The distance between hole centers is then'~~~~~ mm from row to row. This distribution has proved to be the most favorable configuration of holes in a large number of tests.
In two of the four arm ends, ~g noles are provided in corre-spondingly constructed hole arrays. These holes generate a thrust in khe suppression chamber's circumferential direction in order to improve the temperature mixing during blowdown of individual quenchers.
3.2 Arrm ement of the relief i e and uencher Connected to each valve is a blowdown pipe which is led downward through a vent pipe into the water space. The vent pipe serves as a protective tube and prevents steam from entering into the air space of the suppression chamber in the event of a leak in the blowdown pipe. The diameter of the blowdown pipe is still 5-19
being optimized for the clearing process; cf: Section 4. The guide of the blowdown pipe is shown in Figure 3.3. The pipe is not connected directly to the protective tube, but rather is held independently at two points. At the top the blowdown pipe is anchored to the valve, which is rigidly mounted to the erection platform in the upper annulus. At the bottom, the quencher con-nected rigidly to the pipe is guided into a mount at the bottom of the suppression chamber (Figure 3.4). It allows axial motions of the quencher due to thermal expansion of the pipe by up tot%%
mm. Transverse and rotational motions of the quencher, however, are limited by guidance in two mutually perpendicular planes.
In addition, the lower mount is so constructed that the flow around the quencher arms is not impaired.
Th e dimensions of the quencher relative to the suppression chamber are shown in Figure 3.5. The narrowest configuration of two bl owdown pipes is illustrated there. The smallest distance between adjacent quenchers is Q~Q m; the smallest distance from the inner wall ishM~lm. The distances are large in comparison to the steam braid (sic) length of ca 'Mcm.
---All 7 quenchers are drawn to scale in Figurc 3.6. The numbers designate the valve group. Ne see that adjacent quenchers are not actuated simultaneously.
5-20
- 4. ttom load durin vent clearin Tests were performed in the GKH test stand to determine the loads during vent clearing. They are described in detail in /2/. In the test series performed most recently (perforated pipe version 7 in /2/), the following parameters were identical to those in the reactor:
- perforated pipe with 4 arms
- hole diameter: h~mm
- arrangement of holes in rows with a row separation of~gmm and a hole separation of kWmm in the row
- mean distance of the quencher from the bottom:LW3 m
<<steady-state steam mass flow density, relative to the total cross-sectional area of the quencher holes
- valve characteristic submergence At the test scale used in GKM, the nominal steam flow density is reached when saturated steam at only ca. Qg kg/cm 2 (absolute) appears before the actual valve. During vent clearing in most of the tests, there occurred transiently a pipe pressure which restricted the'flow rate toward the end of the process. This lt phenomenon is corrected computationally in the transposition of results.
The valve opening time and submergence were varied in the tests.
In addition, tests'ere performed with different quencher outlet 5"21
areas. The previous vent clearing tests in the G101 test stand of relevance for the reactor design resulted in maximum pressure oscillations of g~~~~~kg/cm't the tank bottom. These pres-sure oscillations are influenced not only by the valve opening time and quencher outlet area, but also by the volume of air enclosed m the pipe. In the current test series, the influence of the volume of air enclosed in the pipe will also be studied by varying i s diameter with a true-to-scale reactor quencher.
the ,pipe The purpose of these confirmation tests with the true-to-scale quencher is to determine the associated bottom pressures in GKN, which then can be transposed directly to the plant (except for the calculated correction for the flow rate during, clearing), and which sho'uld not exceed the numerical values indicated above.
Therefore, the limiting values from the G1Q1 tests were taken over for the specification of the bottom load on the suppression chamber during vent clearing. That means that a bottom pressure of VMM~~Qkg/cm is assumed directly beneath the quenchers.
Using suitable combinations of parameters (pipe diameter and quencher outlet area), there are sufficient ways to achieve that
.-:goal.
No statement can be made yet concerning the decrease of pressure in the meridional direction. Accordingly, we conservatively assume a constant pressure at the bottom of the suppression chamber. The same pressure is also assumed at, the side walls 5-22
l I
,up to the outlet height of the vent pipes. From there, it decreases linearly to the water surface.
A distinct decrease in pressure from the maximum value beneath the pipe was measured in the circumferential direction during the KNH blowdown tests. A correspond'ng pressure decrease will also occur during blowdown with quenchers. It will even be steeper, because the center o. oscillation is deeper due to the greater submergence of the quenchers. Therefore, following a loss-of-coolant accident, when the pressure suppression system is already loaded, a smaller load occurs as the total load for the inner bearing. A maximum of
@~relief valves open in such an accident. Zf the pressure load is integrated in accordance with the most 'unfavorable valve configu-ration (Figure 4.1), then we find P mean ~Q%W ' max' i.e.s L%%%%%
kg/cm. ~ The profile in Figure 4.1 was based on a pressure distri-bution described by a 1/r law, as described in /1/.
5-23
I I
I I
- 5. D namic ressure load durin condensation of steam Although the condensation of steam in a water pool represents an approximately -steady-state process, pressure oscillations are produced which load the walls and bottom of the suppression chamber wetted by the water. The discharge quencher of the relief system in the plant has a very broad spectrum of utilization with respect to mass flow rate and pool temperature (see Section 2) .
In that range, the processes involved in the condensation of stea.
can be subdivided into clearly separated ranges.
To begin, we shall briefly list here the condensation phases observed with a perforated-pipe quencher and their most important distinguishing features. That is followed by a detailed descrip-tion of the phenomena found for the individual phases. The pres-sure amplitudes measured in the test stand are also indicated.
Finally, we provide a compilation of the expected maximum pressure load in the suppression chamber due to condensation, including the expected circumferential distribution of the bottom load.
5,1 Surve of observed condensation hases The most important condensation phases of a perforated-pipe quencher are illustrated diagrammatically in Figure 5.1. They can. be observed as a function of the mass flow rate and pool temperature.
If the steam flows out of the quencher at a critical or supercritical pressure ratio, then the speed of sound is reached'n the narrowest 5-24
cross-section. Pressure oscillations produced outside the cpencher can therefore not have any effect on the flow to the quencher. The processes inside and outside the quencher are therefore decoupled. In the following, this phase is designated as condensation with large mass flow density.
On the other'and, if the steam is blown into the water pool with a subcritical pressure ratio, then the processes outside and inside the quencher can have a mutual influence. As a consequence, the" e can even be intermittent operation. This is discussed in gre e detail in Section 5.2.2.2.2.
Finally, if the steam flow rate is so low that the heat of con-densation delivered with the steam is carried away completely by the relief pipe cooled externally by water, then the condensation occurs only in the pipe and no longer outside in the pool. Zn that case, pressure oscillations due to the condensation processes are still observable only to a negligible extent outside the
~
quencher. This condensation phase is of only slight importance in practical application. It is combined with the previously described subcritical outflow as condensation with sm ll mass flow density. That is proper because between these phases there is a transition range in which the two effects are of the same order of magnitude.
5-25
I I
I I
I
5.2 Condensation with small mass flow densit 5.2.1 Condensation in the pipe Examples of results of measurements of condensation in the pipe are shown in Figures 5.2 and 5.3 which wem obtained in the GKt1 test stand (. igure 2.4) . A'etailed description of the different
'test geometries and a compilation of all tests can be found in /2/.
I For the test illustrated in Figure 5.2, the condensation occurred only the in he supply pipe at a flow rate of + t/h for a pool temperature belowMQC. For the test illustrated in Figure 5.3, that was the case at a flow rate of~~ t/h for a pool temperature below Q~C. *) As a rule of thumb we note here that for the test geometry a flow rate of Q t/h is to be understood as kW of nominal flow 'rate.
Since the condensation takes place completely inside a volume bounded by the pipe wall and the remaining water slug, no pressure
- )The amount of heat dissipated along theQ.tm-long pipe section between the pool water level and the perforated collar of the test geometry, and- thus the amount of steam condensed along that pipe section, shall be estimated briefly: Under the assumption that the heat transfer both from the flowing stcam to the pipe wall and also from the pipe to the water pool is very good (in other words, the inner surface of the pipe is heated up to the steam temperature, whereas the outer surface of the pipe assumes the water temperature),
and assuming a coefficient of thermal conductivity of 'Mkcal/m h'V, BLMt/h of steam can be condensed in the first case and~~~ t/h 'in the second case. These values are only a littl'e below the measured flow rates of <Q and ~~ t/h, respectively. The difference can be explained by the fact that a small amount of heat is also dissipated along theg~~m-long section of the pipe lying in the air space.
5-26
I I
I I
pulsations are to be expected in the pool (i.e., outside that volume) due to the condensation. In fact, the test recordings exhibit only negligibly low amplitudes in the vicinity of the resolution limit of the measurement chain (see also Figures 5.2 and 5.3), which are attributable to a hum and the temperature sensitivity of the transducers.
5.2.2 Condensation with subcr'itical outflow 5.2.2.1 Pressure pulsations with subcritical outflow When operating a perforated-pipe quencher with subcritical pressure ratio, regular sinusoidal-like pressure pulsations with frequencies up to nearly MQ Hz were measured at the wall and bottom of the test tank and also in the supply pipe. Examples from the measure-ment traces are shown in Figure 5.10. Pressure amplitudes measured with the model quencher and the frequencies that occurred are illustrated in Figures 5.4 to 5.9. The frequency jumps occurring with the heating of the water during a test are conspicuous. They are associated with abrupt changes of the pressure amplitudes.
Between the jump points, the frequency decreases with increasing water temperature. All signs point to the steam bubbles as the point of origin of the pressure oscillation:
- The heat dissipation becomes poorer with increasing water tem-peiature. The surface area of the steam bubbles and thus their volume increases. But larger bubble dimensions are associated with smaller oscillation frequencies. This explains the frequency decrease with increasing water temperature.
/
5-27
4
- The pressure oscillations propagate both in the pool and also upstream in the supply pipe. Therefore, independently of the frequency, the pressure signals at the individual pressure trans-ducers mu t be shifted in accordance with thei'r distance from the bubbles. Whereas the signal reaches the pressure transducers at the bottom (somewhat more thaniQm away on the average) in% to ms, it requires aboutQQms to reach the pressure measurement point Q, m upstream in the supply pipe at a sound speed of LWm/s and a flow speed of approximatel.:'.QQ m/s. As made clear in Figure 5.l0, the time difference predicted by the calculation was also confirmed by the measurements.
%kMU&%%%%%%%%%%%%%%%%%%%<M%%%%%~~
The reason for the observed frequency jumps remains unexplained.
The oscillation might possibly be caused by regenerative pertur-batiops in, the boundary layer which could also occur, for example, at air jets and also produce frequency jumps there as the parameters are varied continuously /3/.
independently of how the bubble oscillation is generated, the measured pressure pulsations with a uniform frequency and also a substantially uniform amplitude indicate that the processes trans-pire with the same frequency and phase for all steam bubbles (except in the transition regions shown in Figures 5.7 to F 9). The 5-28
synchronization is achieved through the volume of steam flowing into the quencher and relief pipe, resulting in a forced oscilla-tion. The highest pressure amplitudes measured with the model quencher were~~~~:g/cm Transition range to condensation in the pipe 5.2.2.2.1 Steady-state phenomena If the amount of heat necessary for condensation of the supplied steam cannot be dissipated completely through the pipe wall, then the ~ater level in the pipe drops so far due to a corresponding pressure build-up in the steam space that the remaining steam flows out through the released cross-sectional area of the holes and condenses outside the quencher in the pool. In the over-whelming number of tests, the test geometries described in detail in /2/ had discharge areas in the supply pipe (perforated collar and perf'orated neck), which were therefore released first. For I
the tests already described in Section 5.2.1 in connection with condensation in the pipe (Figures 5.2 and 5.3), the first openings were cleared when higher temperatures were reached in the pool,
.=Men the temperature difference across the pipe wall was no longer sufficient for complete condensation of the steam inside the pipe.
Figures 5.4 to 5.6 show measurement results for a steam flow rate of to~at/h. At,.one measurement point located immediately before the inlet into the branch piece of the quencher, pressures of approximately@~~ kg/cm 2 (absolute) and temperatures of at most 5-29
I
%h'KQC were recorded throughout almost the entire test procedure (to pool temperatures up to more than g~C). If it is taken into consideration that both quantities can be afflicted with a measure-'ent error, then these values relate both to saturated steam and also to boiling water. Since clearly superheated steam is supplied, the temperature o the steam flowing into the quencher should rise gradually with increasing temperature of the pool and associated reduction of the heat dissipation through the pipe wall. But no such trend is discernible. This apparent contradiction can be explained by the estimate made in the footnote of the amount of steam already blown out up to the quencher inlet and condensed in the pipe. ~)
Figures 5.7 and 5.8 show measurement results for tests with a supplied steam flow rate between~~ and gMt/h. In that operating range, a considerable portion of the steam is blown out through the model quencher itself. In these tests, the expected increase 0)
If we assume L.Qh.%kg/cm a pressure jump across the collar holes of between (absolute) (this is the mean value between the pressure of the suoplied steam and the pressure at the quencher inlet) and hM kg/~.2 (absolute) (this is the hydrostatic backpressure outside the col)ar), then, as uming a coefficient of post-contrac ion of approximately Q',Q t/h of steam can flo~ out across the collar area of Version 6, whereby a discharge velocity of more than half the speed of sound is attained. In addition, assuming a temperature jump ofi~~~C and a coefficient of thermal conductivity ofh',Q kca'/m h K, an amount of heat corresponding to a condensed steam flow rate of g~gt/h is dissipated through the approximatelyg~m<<long pipe between the pool surface and the quencher inlet. Together with the steam flowing out through the collar holes, this is approxi-matelyl~g of the steam flow rate of +~~~3 t/h suoplied through the valve. Thus, in this operating range, only a small fraction of the steam flows into the, actual quencher. Because of the pre-ceding heat dissipation, it is surely not,superheated.
5-30
of the steam superheating at the quencher inlet with increasing temperature inI the pool is also observed. The fraction of steam condensing in the pipe is clearly smaller than the fraction flowing out. The flow rates associated with this operating point should therefore no longer be counted as part of the transition range.
For low discharge rates, the quencher is not blown free completely but rather the water slug is only moved so far that the hole cross-sectional area necessary for the discharge is released. This equilibrium position of the water level in the pipe or in the quencher is not very stable, however. Zf the water slug is pushed out too far and a too large discharge area is thereby released, then a larger steam flow rate condenses than is supplied through the valve. As a result, the pressure in the relief pipe falls and water flows in through the quencher. The penetrating cold water causes a violent condensation. As shown in Figure 5.11 by means of an extract from the pressure recordings, there is then a collapse of the pressure in the relief pipe.
Both in the test stand and also in the power plant, the relief pipe is provided with a snifter valve which opens at an underpressure of 1%%%kg/cm 2 and allows air to flow in for pressure equalization.
The lowest 'possible pressure is therefore limited to +~+kg/cm (absolute) in both cases. As soon as the air reaches the water level in the quencher or in the relief pipe, the condensation comes 5-31
I
)
I
nearly to a standstill and pressure is built up again in the relief pipe. The pressure initially rises above the steady-state value because the water slug has to be displaced.
This process is clearly illustrated in Figure 5.11. A steady-state pressure of~~g/cm" (absolute) was measured in the supply pipe. A maximum of g~~kg/cm (absolute) was reached when the water slug was expelled, i.e.~~kg/cm more than the steady-state value.
Because of the predominant influence of the pressure and inertia forces, the water motion depends solely on the pressure variation in the quencher and in the relief pipe. The variation of piessure with time depends on the ratio of the steam volume in the relief system to the discharge area of the relief system. Since a time expansion or compression of the pressure variation has the same effect on the inflow and outflow processes, such a parameter varia-tion has a completely neutral effect on the pressure amplitudes that occur. Thus, as long as the pressure minimum is established by the action of the snifter valve, the pressure maximum and thus the entire sequence of motion is also establishe'd for geometrically similar quenchers.
As a consequence of the water motions, pressure oscillations are observed at the bottom and at the wall of the tank when the water enters and also when it is expelled. But those oscillations decay rapidly (Figure S.ll) . According to Figures 5.4 to 5.6, maximum 5-32
I I
l
peak values ofg~~3kg/cm 2 occurred.
5.3 Condensation with lar e mass flow densit In the primary operating range of the quencher, the speed of sound prevails at the outlet. For that reason, the processes inside and outside the quencher are decoupled. Therefore, the processes proceed in a much simpler manner than when there is subcritical outflow.
5.3.1 Pressure pu..sations with supercritical outflow Figure 5.13 shows a typical example of the measurement traces obtained with the bottom pressure transducers in the GD) test stand (Figure 2.4) for operation of the quencher with a super-critical pressure ratio. High-frequency pressure oscillations occur with very small amplitude and without any fixed frequency.
It is obvious that a synchronous oscillation of the steam bubbles does not occur. The reason is probably that the steam bubbles can only have a small mutual effect on each other by means of pressure signal" across the pool. The much more effective mutual influence via the steam flowing inside the quencher is excluded because of the speed of sound occurring in the outlet cross-section.
As a result, the steam condenses without 'appreciable pressure amplitudes occurring. The maximum pressure amplitudes measured with perforated-pipe versions~Lto+l in the GKM test stand are plotted versus pool temperature in Figure 5.13. If we factor out 5-33
I 1
I I
the phenomena occurring just before the boiling temperature is reached in the water pool (described in more detail in the following Section), then the pressure amplitudes are limited to a maximum of Q+Lkg/cm 2 . That value was only reached in one test during a brief interval of time. Otherwise, pressure amplitudes of ~++++
kg/cm 2 were not exceeded.
5.3.2 Steam bubble oscillations near the boiling point of the pool Because of the arrangement of the holes or the steam outlet in rows, between which there are broad lanes for supplying water to =
the steam (see Figure 3.2), the pool can be heated up to the boiling point without recording any rise of the pressure amplitudes (Figure 5.13) . For older versions of the perforated pipe which were pro-vided with a uniform distribution of holes, large steam bubbles were formed beginning at a pool temperature of aboutk%'C /2/.
Since the water in the thin test tank at GKM was not able to escape laterally, the water column "danced" on the steam cushion. The large pressure amplitudes resulting from that motion are illustrated in Figure 5.13.
With the hole configuration described above, this phenomenon can be eliminated completely. Studies concerning the necessary width of the water lanes were carried out in the model test stand /4/ ~
They were used to lay out the hole distribution of the actual quencher.
5"34
5.4 Ex ected maximum ressure load in the su ression chamber Zn this Section we shall describe how pressures measured with model quenchers in the test stand can be used to deduce the maximum pres-sure load in the suppression chamber of the power plant. Because of the fundamentally different physical process, it is appropriate to distinguish between intermittent operation of the quencher with rhythmic inflow and outflow of water, on the one hand, and con-tinuous operation with subcritical or supercritical outflow, on the other hand.
5.4.1 Intermittent operation The intermittent operation o'f the quencher that, is possible for low flow rates was thoroughly described in Section 5.2.2.2.2.
water flows in and out alternately. Furthermore, a justification was presented there for the assertion that a completely corre-sponding behavior of the processes can be expected in the model quencher and in the large-scale version because of the limitation of pressure in the pipe by the snifter valve. We can therefore assume that the pressure oscillations generated at the quencher outlet are the same for thc main quencher as for. the model quencher.
As is demonstrated in the analyses presented in the following .two Sections, for equally large pressure oscillations and at the openings of the quencher the pressure load on the suppression chamber can also be no greater than was measured in the GKM test stand. Therefore, for intermittent operation, a pressure load no 5-35
greater than~~$ kg/cm 2 can be anticipated at the walls and bottom in the suppression chamber as in the test stand.
5.4.2 Continuous operation Any single steam jet that emerges from the perforated-pipe quencher oscillates and thereby accelerates water rhythmically to and fro.
It is only toward the top that the water space is not con ined an" the water can escape. However, because the acceleration o- water is necessary for the decrease of the pressure signal, the unfavorable two-dimensional pressure propagation toward the bottom always occurs in this test tank. A condensation test with a small distance of the quencher from the bottom is compared in Figures 5.14 and 5.15 with two other tests which were carried out with more than twice as large a distance to the bottom. The steam flow rate is approximately the same in all tests. Zn addition, the hole dia-meter in the arms is the same in Figure 5.14. Zn Figure 5.15, on the other hand, the entire outlet area is the same. A decrease of the pressure pulsations with distance of the quencher from the bottom cannot be observed, which should also be expected according to the discussion above. Thus, the pressures measured at the bo tom of the test tank are on the safe side. Ne may therefore assume that the maximum pressure load measured in the test stand due to condensation ~
4++~~~++~~~j~%3is not exceeded at any point.
5-3G
5.4.3 Circumferential distribution of the maximum bottom load during automatic depressurization The perforated-pipe quencher is so designed that the outflow of steam from the quencher at reactor rated pressure occurs with a supercritical pressure ratio for which, according to the dis-cussion above, the smallest pressure loads occur. Operation of the quencher with subcritical pressure ratio and also intermitten" operation for very low flow rates occur during the automatic deprcs-surization described in Section 2. During that operating condition, only 4~valves respond in KKB. They are distributed uniformly over the circumference (Figure 2.1) and thus are widely separated from one another. Therefore, the maximum pressure load occurs only in the close vicinity of the quencher. At a large distance from the quencher, the load decreases sharply, so that the total. bottom load is considerably below the peak values near the quencher. The total reduction factor shall be determined in the following on the basis of a conservative estimate.
For simplification we shall think of a circular disk covered uni-formly with synchronou ly oscillating steam bubbles and having the radius of the quencher and a thickness a little greater than the thickness of the arms (Figure 5.16) . For an observer located centrally under the disk, only the disk's diameter and not its thickness is of significance. Thus, the pressure decrease in accordance with the law for an oscillating spherical bubble (p ~ r = const.) /5/ occurs for. that observer in such a way as if 5-37
a large spherical bubble with the diameter of the disk were oscillating. That bubble is also illustrated in Figure 5.16.
However, in the conservative estimate being made here, the pres-sure decrease in the vertical center section shall be completely neglected.
For an observer at the bottom shifted laterally in the circum-ferential direction, the pressure decrease with the circular disk is faster than with the previously described enveloping sphere because of the smaller thickness.. Accordingly, the decrease of pressure with the sphere's radius shall be assumed conservatively in the circumferential direction. The developed circumferential distribution of the maximum bottom load calculated in this way is shown in Figure 4.1 in a normalized depiction for the two valves with the smallest separation that are actuated simultaneously during automatic depressurization (Figure 2.1) . According to that, the total pressure load on the bottom is less thanL+%of the peak value. For the maximum pressure amplitudes oQ+~~kg/cm 2 indi-cated in Section 5.2.2, the total load on the ~bottom is actually less than+~+kg/cm 2 5-38
- 6. Reaction forces on the erforated- i e uencher The forces exerted on the system consisting of valve, blowdown pipe and perforated-pipe quencher during blowdown of the relief valves can be divided into internal and external reaction forces.
The internal forces include pressure x area, frictional forces, the recoil on the valve, the dynamic pressure in the quencher and deflection forces inside the valve and quencher.
Insofar as no resultant components occur, these forces are allowed for by a suitable wall thickness of the material.
Among the external forces we must distinguish between the primary forces produced directly by the recoil action of the outflowing water, air and steam, and secondary forces due to the inflowing water. The steady-state and non-steady-state processes must be investigated for both, the internal and external forces. This Chapter is limited to the external forces.
Vent clearinc rocess When the relief valve is opened, a pressure builds up in the space above the water 'surface in the pipe. This pressure accelerates, and expels the water column. It is followed by compressed air and then by the steam.
The order in which the individual media are expelled is determined by the geometrical configuration and is made certain by experimental observations, A certain degree of mixing at the interface between 5-39
ig I
l I
steam and air is probable and depends essentially on the diameter/
length ratio of the blowdown pipe.
6.1 Forces on the individual hole The forces that now act on each outflow cross-section with area fL'are the recoil hv and p fL, where p is the unrelieved pressure in" the narrowest cross-section. (Zt is assumed that the, latter is still inside the outflow cross-section.)
Thus we find for the force K L acting on ~n individual outflow cross-section:
L PL'LL For subcritical outflow, we now set hp ~ 0 for simplicity and for critical outflow we calculate bp from the equations of gas dynamics.
6.1.1 Outflow of the water If we assume for the final velocity of the water a realistic value of%am/s in the individual outflow cross-section and set hp =+~%
then we get a KL of g+~Q//q 6.1.2 Expulsion of the ai" The air emerges initially at the final velocity of the water and expands immediately so that the air flowing after it expands in a cavity. If we assume critical expansion and a clearing pressure PF 4% kg/cm %%%++>a mass flow rate mL4%%~+Xg/s and a velocity VLI%%% m/s, then we obtain a force of KL +%%4kp.
5-4 0
6.1.3 Ex ulsion of the steam 2
Based on the outflow area of the quenchez%%%%3 m and the maximum flow rate per individual hole m ~+~+ kg/s, we obtain I
an individual force of ~~kp for the outflow of steam.
6.2 Calculation of total forces Since the hole configuration is symmetric, no resultant forces should occur, apart. from those caused by fabrication inaccuracies and by nonuniform outflow of steam.
Nevertheless, to obtain information for strength calculations we made assumptions such as the masking of the upper or lower half of the hole array or of an entire hole array. Due to these unrealistic assumptions that were made in determining the reaction
'll forces and due to round-offs, a safety. margin is allowed for in the calculation.
To calculate the resultant forces with partially masked hole arrays, the components of each individual force in the horizontal and vertical directions are added up.
Assuming the individual force of g~+kp calculated for the expulsion of water, and making the assumptions described below, we obtain reaction forces pl p4 (see Figure 6.1) of the following order of magnitude:
5" 41
l l
'I I
KRAFTWERK UNION AG PROPRIETARY INFORMATION 6.2.2 Stead -state blow-out of the steam After expulsion of the water column and air cushion, a pressure fluctuating about a mean value is set up in the relief pipe. Under the assumptions made above for the ma..king of the hole arrays, we obtain the following reaction forces for the design pressure in continuous operation at++kg/cm 2 The application points of these forces are shown in Figure 6.1 ~
A summary of all calculated forces and of those that were indicated earlier in a specification for the quencher design demonstrates 5-4 2
'l l
l l
ig I
that the specification values can be considered conservative I especially because of the unrealistic assumptions made in the calculation.
Based on the geometrical construction of the quencher, we can now use the calculated or specified forces to calculate all possible moments that act on the quencher as a whole or on any individual arm. These moments were all taken into consideration in preparing the Fab ication Specification. For the sake of clarity, we shall dispense with any description of those calculations here. he shall only indicate the computation method and the assump.ions made in determining the reaction forces.
5-43
I I
I I
I I
- 7. Flow-rate capacity with the perforated-pipe quencher at reduced reactor ressure At high reactor pressure there prevails above the seat of the safety/relief valve a critical pressure ratio. The flow rate through the valve is then'determined primarily by the reactor pressure and is approximately proportional to it. 7f the reactor pressure drops to low values, then it finally falls below the critical pressure gradient above the valve seat and the flow rate decreases more than proportionally with the reactor pressure.
Since. the flow-rate capacity at reduced reactor pressure plays an important role for the case of automatic pressure relief described in Section 2, it was investigated in detail for operation of the relief system with the perforated-pipe quencher.
The flow rate per relief valve for low reactor pressures is plotted in Figure 7.1. The upper curve corresponds to the flow rate through the valve as indicated by the valve manufacturer. En the plotted range of reactor pressures, these values are also appli-cable in practice for the flow rate through the relief system with a subsequently inserted perforated-pipe quencher. The reduced flow rates illustrated by the lower curve are set for a pressure 2
of Le~kg/cm I (absolute) in the air space of the suppression chamber. The flow rate through the relief sy" tern,at reduced reactor pressure is not influenced substantially by the perforated-pipe quencher (for reactor pressures~~Qkg/cm 2 (absolute)), but is influenced more. greatly by a rise of the pressure in the suppression chamber.
5-44
I I
I
The variation of the liquid level in the core shroud during automatic depressurization following ah~~cm 2 leak and failure of the coolant injection system is plotted in Figure 7.2. The illustration'makes it clear that there is practically no dif-ference in the time variation of the depressurization for the two flow-rate variations under consideration. The reason for this is that the difference in thc flow rate is still small when the low-pressure coolant injection system responds (i.e., at a reacto" pressure o bMbar) . As the accident proceeds, heat is then dissipated to an increasing degree through the injected cold wate" and an additional pressure drop is produced.
I 5-45
- 8. Temperature distribution in the suppression chamber during relief rocesses A most uniform possible temperature distribution is necessary in the water storage tank of the suppression chamber because then
- the heat capacity of the water is better exploited and
- thermal stresses at the walls and bottom due to nonuniform heating are avoided.
For a detailed study of the temperature distribution during relic processes it seems appropriate to consider separately the mixing in. the vertical and circumferential directions.
8.1 Vertical tern erature distribution A uniform vertical temperature distribution is easy to achieve if the heat is supplied very deep in the water storage tank; The hot, specifically lighter water rises, mixes with the colder water above it, and we obtain a uniform heating of the pool in the vertical direction. For that reason, the perforated-pipe quencher is installed very deep in the water space of the suppression chamber, as is illustrated in Figure 3.3.
Results of tests on vertical mixing were obtained in thc GK'4 con-densation test stand /2/ and are also illustrated in Figure 2.4.
Temperatures mea"ured at the tank bottom and at the tank wall for the large flow-rate range of QM% are plotted in Figures 8.1 to 8.5.
I Values read off simultaneously are connected by similar types of lines. We see that a uniform temperature with a maximum scatter 5-46
I range oi~~~~'C was found up to the boiling point both beneath the quencher at the bottom and also at the tank wall.
The narrowly restricted lateral geometry of the tank (Figure 2.4) has an extremely detrimental effect on the temperature mixing.
On the one hand, the warm water must rise. On the other hand, however, water must also flow downward toward the quencher. These motions counteract each other.
Zn contrast, in the suppression chamber of the power plant all the water flows toward the quencher from the side. Thus, there are clear inflow and outflow conditions. For that reason, we can anti-cipate a better vertical temperature mixing there.
Et should be emphasized that the water beneath the quencher near the bottom is also heated, although the impulse of the steam flowing out from the perforated arms is directed entirely upward /2/. The fact that a sufficient amount of heat is nevertheless supplied to n
the'water beneath the quencher enables us to infer an intense large-scale turbulence in the pool, which is produced by the flow and condensation processes at the quencher.
A strong turbulent mixed flow in the pool during condensation of steam through the perforated-pipe quencher was also able to be observed optically in the model test stand (Figure 2.5), Those tests are reported upon thoroughly in /4/. Those tests provide primary information concerning the large-scale temperature distri-bution. Accordingly, they shall be discussed in the next Section.
5-47
I I
I I
N 8.2 Temperature distribution in the circumferential direction of the suppression chamber The distribution of relief pipes in the suppression chamber as illustrated in Figure 3.6 makes it clear that the valves associated with the individual valve groupsg which respond jointly according to the cases described in Section 2, are distributed uniformly along the circumference. The geometrical arrangement therefore guarantees a uniform heat supply in the circumferential direction.
However, we mu::t also consider the case that a single valve responds for a longer period of time. In that case, the steam is supplied only at one point of the suppress'on chamber ring. The heat transport described in the preceding Section due to the characteristic motion of large turbulence elements [turbulence bubbles) in the pool is more effective in the horizontal direction than downward toward the bottom, since the steam is blown into the water primarily horizontally, as can be recognized from the hole distribution in Figure 3.2.
Information concerning the large-scale temperature distribution in the suppression chamber was obtained with the model test stand (Figure 2.5) . The water surface area there (relative tq the pipe cross-sectional area) is about ten times greater than in the .thin GKM tank. Therefore, the pool itself is also much larger compared to the quencher. The length of the test stand in relation to the quencher corresponds to approximately half the developed circum-ference of the suppression chamber ring. Figures 8.6 to 8.9 show
examples of temperature distributions in the pool. They were recorded with two different versions of the perforated pipe and with distinctly different mass flow rates. The uniform temperature distribution at the bottom and at the wall demonstrates the action of the turbulent mixing observed optically: The temperature differences are at most~~C. It may be expected that this effect will act in the same way in the suppression chamber of the power plant.
The mixing in the circumferential direction can be further improved by a slightly unsymmetrical distribution of holes on the quencher, whereby a one-sided impulse is exerted on the water mass, which gradually sets the water into a slow rotary motion. In this way, the heat is carried away from the quencher and colder water is conducted to it. Such a slight unsymmetry is provided for, for example, by a hole distribution at the ends of ~~~arms facing in the same circumferential direction.
An important active element for the mixing of the water in the suppression chamber is the closed-circuit cooling system. After exceeding an average pool temperature o~~C, water is taken from
-.-4he deepest point of the pool, cooled and distributed near thc water surface along the circumference of the suppression chamber In that way, a recirculation of the water in the suppression chamber is accomplished every half hour by the four RllR legs shown in Figure 2.1 ~ These measures alone already produce a uniform temperature distribution, since, for 1
example, an emergency shutdown 5-49
I I
I I
I I
I
of the reactor extends overhMhours, which is more thar;i~Ltimes longer.
Because of the many processes described above to achieve an equalization over the entire water volume of the suppression chamber, it can be assumed that the maximum deviation of the temperature in the pool of the suppression chamber does not exceed 3C, except in the immediate vicinity of the steam outlet.
550
I I
KRAFTWERK UNION AG PROPRIETARY INFORMATION 5-51
KRAFTWERK uNION AG PROPRIETARY INFOK1ATION Figure++iiiii 2 j 5-52
KRAFTWERK UNION AG PROPRIETARY ZNFORtIATION Figure. 2.2 and 2.3 5-53
B itd 2.4 G K lvl Kondensationsversuchssta nd Gkdl - Condensation test stand
Bild 2.5 Modell - Versuchsstand Gwh Fig@re 2. 5 GK!! r.,ode3. test stand
I I
I I
i I
I
SECTION A-B SCHNITT 4-B
/ SECTIO.': C-D VIEN A / SCHNITT C-D 4NS~-HT 4
/
g~.bores 0g B t ld 3.1. LOCHROHRDUSF F3 NWhMMW Q~MTI.be (Configur-tion of the hole arrays will not specified until after the confirmation tests in GKN)
roQs
~02ei!en 0
k 0
WQ Lochfcldcvs schnitt "i~ Extract from hole array Q~ ,I I
I B i l d 3.2 Lochbeleg~un der Di'jsenarme
B~ld: 3.3.
Fuhrung des Abblaserohres mit LochrohrdQse (KKB)
Guidance of the blovdovn pipe with perforated-pipe quencher (KKB) 5-58
~ ~
'; <I I\
\ r l] I,I I
II I
li Bitd 3.4
. Bodenvercinl<erung der 'use Bottom anchor of the'quencher
'I
'l
I I
I I
I I Fi ure 3.5 I I D imens ion s o f per forated I I pi pe quencher F3 and t. /
arrangement in the KKB suppression chamber I
I I I I I I
I I I
/
I I I I I,
I t
/
/ I l
/ // \
//
/ /
/
/ /'ild 3.5.
/
Abmessungen dert ochrohrdQse F3 r
/ I/
und Anordnung in der Kondensations-karnrner K K B
~ ~
~ ~
90
~ ~ 105
~
350 00 325'P',
265'70'eight Q - Gruppenbezeirhnung ~~
Group desxgnaMon
~FS ure 3 6.
Arrangement oS the quenWers'n the suppression chambers.KKB I
5-,61
KRAFTWERK UNIOIJ AG PROPRIETARY INFOKCATION 5-62
KRAFTWERK UNION AG PROPRIETARY INFOKIATION 5.1 through 5.11 5-63 through 5-73
I I
I I
I
Time axis Zeitachse
~0 w.w.~g~~~
Scale
- -Bottom pressure transducers L<~~ QHkgrcn ~ P
~
IWf v~~~~~ ~ ~ ~ A Fi ure 5.12 Pressure pulsations at the bottom during outflow with supercritical pressure ~atio (stochastic process)
,GKM test no, 224 (perforated pipe version 7)
.m @t/h zF =1</Pi cm'ole = i~i~Zg kg/m s
I KRAFTWERK UNION AG PROPRIETARY INFOKCATION Figure.. 5.l3 thro a.'x 5.l6 5-75 through 78
KRAFTWERK UNION AG PROPRIETARY INFORMATION 5-79
KRAFTWERK UNION AG PROPRIETARY INFOK4ATION 5-80
KRAFTWERK UNION AG PROPRIETARY INFOKCATION Figure .. 7.2.
5-81
KRAFTWERK UNION AG PROPRIETARY INFORMATION Figure.. 8.1 through 8.9 5-82 through 5-89
I I
1 I
I l
REFERENCES
/1/ f'renkel, Becker, Nowotny, Schnabel, Koch:
Brunsbuttel nuclear power plant Pressure distribution in the suppression chamber during clearing of the relief pipes, taking into consideration KWW tests in November 1972 AEG-E3-2486, January 1973
/2/ Hoffmann, Knapp, Meyer, Waldhofer, Werle, Mebhior:
Condensation and vent clearing tests in GKM with perforated pipes KWU-E3-2594, May 1973
/3/ Wagner:
Oscillation phenomena in axially symmetric free jets .of high supersonic speed impinging on a wall Diss. TH Aachen, 1970
/4/ Werner, Melchior:
Tests on mixed condensation with model quenchers KWU-E3-2593, May 1973
/5/ Weisshaupl, Koch:
Formation and oscillation of a spherical gas bubble under water AEG-E3-2241, May 1972
/6/ Slegers, Molitor, Hoffmann Outlet geometries for the pressure relief pipes in the boiling water reactor; first development results AEG-E3-2465, December 1972