- Email: [email protected]
Passive heat removal experiments for an advanced HTR-module reactor pressure vessel and cavity design
ELSEVIER
Nuclear Engineering and Design 146 (1994) 409-416
Nuclear Engineedng Design
Passive heat removal experiments for an advanced HTR-module reactor pressure vessel and cavity design L. Wolf, A. Kneer, R. Schulz, A. Giannikos, W. H~ifner Battelle Ingenieurtechnik GmbH, Diisseldorfer Strasse 9, 65760 Eschbom, Germany
Dedicated to Prof. Dr.-Ing. H. Unger, Lehrstuhl fiir Nukleare und Neue Energiesysteme, Ruhr-Universit~it, Bochum, Germany, at the occasionof his 60th birthday.
Abstract In the search for further reducing the residual risks of possible major reactor pressure vessel failure during and in the aftermath of severe accidents of modular HTRs, an alternative RPV has been designed and a sample vessel already fabricated by the firm Siempelkamp, Krefeld, FRG. This alternative RPV design is made of high quality, ductile sphero east iron with axial and circumferential wire or more recently cireumferentially flat band prestressing. This specific Siempelkamp design has been tested and qualified already on behalf of a series of experiments in the sample test vessel. Also, this design was used for the control gas vessel in the THTR under operational service conditions. In order to demonstrate reliable decay heat removal under most severe conditions, a 1 : 1 scale, 20° sector of the vessel/cavity, termed INWA-facility (inactive decay heat removal) was fabricated and tested at Siempelkamp. The cavity was cooled by natural circulation of water flowing in tubes embedded into the cast iron structure of the cavity. A total of 5 experiments were performed with this setup examining a variety of changes in constructive details, surface and cooling conditions. Each experiment was performed both for operational conditions and depressurization transient, typical for a 200 M"~thHTR-module. Experimental test times ranged up to 1000 hours. Pre- and post-test predictions with the TOPAZ-code accompanied the INWA test series. The paper describes the INWA-faeility and reports some of the experimental results as well as the predictive capability of TOPAZ by comparing the data with computational results. The INWA-results qualify the pre-stressed cast iron vessel together with the natural circulation cooled cavity even for the worst of severe accident conditions. Even in case of total failure of all cooling capabilities in or at the cavity structure the vessel surface temperature remains below critical values.
1. Introduction Since 1973, the design and planning of German pebble bed high temperature, gas-cooled reactors was accompanied by the alternative reactor pressure vessel design using the concept of prestressed cast-iron (PCI) vessel [1]. In 1975, developmental work started and in the following
years, it was demonstrated, that PCI-vessels can be operated up to 350°C. The required materials for cast-iron, liner and wires were systematically examined by material testing programs which have been documented. Both, the control gas storage tank for the T H T R as well as the large-scale H T R test vessel at Siempelkamp, Krefeld, Germany, impressively
0029-5493/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0029-5493(93)E0246-G
410
L. Wolf et al. ~Nuclear Engineering and Design 146 (1994) 409-416
demonstrated the predictability, the feasibility as well as the licensibility by the authorities according to standard codes and regulations. Public acceptance of future nuclear power seemingly necessitates increased application of passive design features without resorting to active components. In addition, emphasis is put to further reduce remote, residual risk contributors, such as for instance pressure vessel failure. The prestressed cast-iron pressure vessel combined with a passive heat sink capability integrated into the reactor silo structure shows promising passive, and forgiving features satisfying a number of presently discussed requirements of nuclear systems of the future. In late 1987, the firm Siempelkamp proposed [2] such a combination of reliable pressure vessel with the advantages of a passive decay heat removal system fully integrated into the reactor silo cell structure. This proposal was based on the design features, overall measures and characteristics developed by SIEMENS/KWU (INTERATOM) for the 200 MWth HTR-modular reactor design, with the idea to substitute the PCI-vessel for the common steel vessel and to replace the silo structure by the composite cast-iron/concrete structure developed by Siempelkamp [2,3], as described in the following.
2. PCl-vessel/passive cast-iron-concrete composite silo structure system Figs. 1 and 2 show axial and transverse cuts through the overall system proposed. All major elements of this passive combination of PCI-vessel with advanced composite structure for the reactor cell structure are numbered and fully itemized in the footnotes of Figs. 1 and 2. The proposed system fits into the overall design of the 200 ~ t h SIEMENS/KWU HTR modular pebble bed reactor design with an operating system pressure of 6 MPa. During normal operating conditions, the energy loss across the pressure vessel of 350 kW is transferred by combined convective and radiative processes to the inside surface of the cavity, where
prestressed cast;iron pressure vessel (~)Reactor cell Axial prestressing tendon~(~ Top s l a b ~ ) Heat exchanger Hoop prestresstng cable ~.~ Ltner wtth support structure Coollng tubes ( ~ Section for test setup
Fig. 1. Vertical cross section of HTR-module with PCI-vessel/cavity design.
it is conducted to the cooling tubes embedded into the cast iron profile. Finally, the majority of the energy is transported away by the water coolant, the rest to the outside surface of the reactor silo structure. In case of an accidental depressurization transient of the system, the energy to be transferred to the cell cooling structure increases to 890 kW. The coolant will rise to the natural draft coolant tower by virtue of buoyant natural circulation. To
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
maintain the heat sink capability under all circumstances, two redundant coolant piping systems are used. During the developmental design of the aforementioned system, it became obvious that a number of input parameters for design and safety analyses were not known with sufficient accuracy. Due to its specific construction, based on a variety of lined-up individual components in series, the vessel component is much more complex than the presently used steel vessels. Therefore, concern was raised, whether the total entity of prestressed components can mitigate the decay heat released over a spectrum of operational and severe accidents as a number of complex combinations of conductive and radiative heat transport mechanisms prevail from the inside vessel through the various components and the special cavity towards the outside concrete surface of the vessel confinement compartment. In order to solve the open questions concerning decay heat transport
411
,/o
I 1 Electric heater and core vessel 2 Liner 3 Pressure vessel 4 Axial prestmcalng tendon 5 Hoop prestressing cable 6 Cover plate
7 Cast Iron block of the reactor call 8 Cooling system 9 Reinforced concrete 10 Sprinkler system 11 Insulation
Fig. 3. INWA test facility (dimensions in ram).
as well as to demonstrate the feasibility of the combined design a 1:1 scale, 20° sector of the vessel/cavity was fabricated and erected, termed INWA-facility, at Siempelkamp. Fig. 2, item 11, indicates the sector element chosen for the experimental demonstration.
3. INWA test facility
~)Core area (~)Ltner (])Liner support structure (~)Slock of cast iron body ..~Axtal prestruss~n9 tendons (~Honp Rrestresstng cable (~T)Reactorce;1 (~)R~nforced concrete "(~Cast iron proftles Q~Cooltng tubes ]Q~ Section for test setup
Fig. 2. Horizontal cross section of HTR-module with PCI-vessel/cavity design.
Fig. 3 gives a perspective view of the INWAFacility [3], which was constructed with all details of the realistic PCI-vessel and cast-iron/concrete composite structure with circumferential tendon reinforcement. The height of the sector was chosen to be about 2 m - only a fraction of the total axial system extension. Therefore, natural circulation
412
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
condition in the embedded tubes had to be simulated accordingly by appropriate forced convection condition. Also, because of the sector shape the hoop prestressing cable could not exert the proper force on the PCI-vessel structure as under normal operating conditions. However, the experimental structure was solidly damped at the floor. In order to provide an ultimate heat sink in case of failure of both redundant, natural circulation piping systems, a film cooling device, see 10 in Fig. 3, was installed to rinse down the outside reactor cell composite surface. The complete INWA-facility was embedded within 500 mm thick thermal insulation. Also, as shown in Fig. 3, the supporting base structure was carefully designed for its isolating properties. Super reflecting metal foil was applied for the floor, ceiling and sidefaces of the reactor cavity to provide proper thermal boundary conditions. All of these precautionary measures were necessary to keep heat losses from the INWA-facility as low as possible. Based on the experience elsewhere, the crucial role of the heat loss control was well
known to affect the successful performance of these types of experiments and their reliable interpretations. The driving thermal power is provided by the electric heater element, see item 1 in Fig. 3. This heating element consisted of 4 independently controlled heating circuits to assure a surface temperature as uniform as possible by properly accounting for the end heat losses. Thermographic control photos indicated maximum temperature deviations of only 4°C located towards the heater plate sides.
4. Measurement systems
In order to reach all objectives of the INWAtests, the complex three-dimensional structure was instrumented with a total of 123 embedded and surface thermocouples, the majority of which were placed along the radial centerline of the sector. Furthermore, to obtain temperatures and heat fluxes across gaps and contact interfaces between components, emphasis in instrumenta-
cooling system (heat sink) radiation radiation and ~ convection
radiation and . heat flux boundary
~
convection conduction
conduction
natural convection or film cooling
Fig. 4. Heat transport phenomena in the INWA-module, MP1-3: Positions of measurement and calculation comparison. Nodalisation of 2-D INWA FE-model.
L. Wolf et al. /Nuclear Engineering and Design 146 (1994) 409-416
tion was put to position a sufficient number of thermocouples at the relevant surfaces and at and into the wire packages. Furthermore, to properly evaluate the heat loss, thermocouples were placed towards the boundary sides of the facility. Special instrumentation was positioned inside the cavity to determine the temperature distribution, velocity fields and radiative transfer. For this purpose, 3 globe thermometers, 4 turbine anemometers and special thermocouples were attached at movable rakes. Four calorimeters were installed into the embedded cooling tubes. After the first preliminary experiment, the importance of correct heat flux control became apparent and a special heat flux sensor was additionally placed at the heater plate surface facing the liner component. This allowed redundant and diverse control of the heat flux boundary condition. All measurement signals were sampled, converted and documented by a PC-based system and software.
5. O~ective and test matrix
Fig. 4 shows a horizontal cut through the INWA-facility and at the same time indicates the
413
nodalization used for the thermal analysis with the FEM-Code TOPAZ [4]. Also, this figure reveals the variety and complexity of heat transport processes occurring in the different components of the PCI-vessel and composite reactor cell structure. The objectives of the INWA-experiments were as follows: (1) Demonstration of reliable and predictable passive heat sink capability under operational and accidental reactor conditions without invalidating thermal limits of individual components and the integrity of PCI-vessel structure. (2) Determination of appropriate values for thermal resistances, thermal conductivities (axial and circumferential packages of cable wires), and heat transfer coefficients at various interfaces. (3) Evaluation of the importances of convective and radiative heat transport mechanisms in gaps, most importantly across the cavity. (4) Determination of the effects of changes in constructive details (addition of a protective cover plate, shielding the circumferential cable package, surface treatment by adding a black coating onto the surfaces facing the cavity). (5) Evaluation of the impact of total failure of
Table 1 Summary of major characteristics of the 5 INWA-experiments Exp. No.
Description
Cooling
Film cooling
0
Heat flux increase (10%) 200 h to 370 h
on
off
1
Start of the DPE at 504 h Experiment with cover plate
on
off
2
Start of the DPE at 408 h
on
off
3
Start of the DPE at 577 h Black coating onto the surfaces facing the cavity
on
off
4
Start of the 1.DPE at 283 h Start of the 2.DPE at 788 h Black coating onto the surfaces facing the cavity
off off
on off
5
Start of the 1.DPE at 100 h Start of the 2.DPE at 350 b Steal vessel instead PCl-vessel
on off
off off
DPE = Depressnrization Event
414
L. Wolf et al. / Nuclear Engineering and Design 146 (1994) 409-416
INWA EXPERIMENT 5"
Tnmsient
i
Sleady-State Ol~'atlen -
L 0.
t00.
200.
300.
i 400.
000.
Fig. 5. Heat fluximposedon liner surface for INWA-tests.
the natural circulation heat sink capability a n d / o r cell outside surface film cooling on the thermal responses of the PCI-structures and the composite reactor cell structure. A total of 5 experiments have been performed and evaluated thus far, testing the PCI-vessel structure under different conditions cited above. Some characteristic features of these experiments are summarized in Table 1. A reference test with a section of a commonly used steel vessel is presently underway. Upon its completion, a direct comparison and demonstration of PCI versus steel vessel behavior is possible. All experiments were performed using the transient heat flux profile depicted in Fig. 5 for controlling the heater plate in order to reach steady-state operational or to simulate accident conditions according to the predicted behavior during a depressurization accident of the modular HTR-reactor designed by S I E M E N S / K W U . The curve shown in Fig. 5 accounts for estimated heat losses (20%) of the INWA-facility.
21)0,
4~.
600.
~,
triO0
R A D I A L C O O R D I N A T E F R O M H E A T P L A T E [mm]
600.
rIME[hl
9.
Fig. 6. Comparisons of cylinder block temperature profiles (4~= IIY, normal operation).
tx~ t-
O
I~0.
500,
1500,
R A D I A L C O O R D I N A T E F R O M C Y L I N D E R R I B [mrn]
Fig. 7. Comparisons of radial temperature profles across cavity(normal operation).
5" L.~ ~e
6. Experimental results
Figs. 6 through 9 present measured radial temperature profiles through the PCI-structure and cavity for steady-state normal operation and the depressurization transient, respectively. These
0.
........ ~0. • .....
4ee.
6e0.
~o.
t~0.
R A D I A L C O O R D I N A T E F R O M H E A T P L A T E [ram] Fig.
8.
Comparisons of cylinder
block
(¢) = 10°, depressurizationtransient).
temperature
profiles
415
L. Wolf et al. ~Nuclear Engineering and Design 146 (1994) 409-416
figures show the results of all experiments without protective cover plate shielding the hoop prestressing cables. The comparisons between individual curves in the respective figures clearly demonstrate the small variance of the experimental results, thereby demonstrating consistency and plausibility of the data despite of thousands of hours of operations of the INWA-facility. Under normal, steady-state operation, maximum temperatures at the liner surface range around 275°C (Fig. 6). The radial temperature profiles across the cavity are essentially fiat (Fig. 7), with steep gradients at the PCI-vessel and inside cavity surfaces, indicating a natural circulating flow pattern along the structural surfaces with a large, essentially stagnant core in the cavity. Changes made for the different experiments have only benign effects on the steady-state temperature profiles. Figs. 8 and 9 show the radial temperature profiles for experiments 2, 3 and 4 at specific instants in time. Again, the variance between the curves is very small. Maximum liner temperatures reach a value of about 470°C. For experiment No. 4 the radial temperature gradient decreases towards the cylindrical fin of the outside PCI-vessel surface because this experiment was performed without the natural circulation heat sink capability in the composite reactor cell structure. The
Table 2 Heat transportby radiation and convectionbetween the PCI and the inside cavitywall Exp. Time Radiation Convection Radiation Convection No. (h) (W/m 2) (W/m 2) (%) (%) 0 1 2 3 4
190 400 650 330 530 513 750 250 450 950
764 378 1170 701 1744 708 1786 672 1722 1712
61,3 185 334 61,4 105 55 91,6 53 54 49
92,6 67,2 78,0 92,0 94,3 92,8 95,1 92,7 97,0 97,2
7,4 32,8 22,0 8,0 5,7 7,2 4,9 7,3 3,0 2,8
simulated failure of this system leads to a temperature increase of the inside cavity surface and as a result to a reduction in heat transport from the vessel surface across the gap. This can be clearly seen from Fig. 9 which shows that the gap temperature level increases from around 1500C for e x p e r i m e n t s 2 and 3 and up to 220"C for experiment 4. A concise error analysis for the temperature sensors resulted in absolute errors in the interval [ + 1,7°C, +3°C] which is too small to be presented in the figures. Table 2 summarizes the evaluation of the special measurement techniques applied in the cavity for the determination of the respective fractions of convective and radiative heat transport across the cavity. The respective values indicate the overwhelming importance of the radiative transport (92-97%) versus the convective transport (3-8%), with the exception of INWA experiment No. I which was performed with a shielding cover plate protecting the azimuthal wire package.
Lrd
7. Accompanying numerical analyses b"
0.
5~.
1000.
151X).
R A D I A L C O O R D I N A T E F R O M C Y L I N D E R R I B [mm]
Fig. 9. Comparisons of radial temperature profiles across cavity (depressurization transient).
All INWA-experiments were accompanied by design, parameter sensitivity-, as well as bestestimate pre- and post-test computational analyses with the Finite Element-code T O P A Z [4]. The T O P A Z - c o d e was extended by including 2-D radiation properties using the FACET-methodol-
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
416
be predicted with confidence as depicted in Fig. 11. -'-
MP-3 E:EP
....
~*~-2 CALC
" -"
I4~-3CAL~
8. Conclusions .
...:'2=2'-='2:'.: 2 : . . . . .
g. d
~m,
~0.
TIME [h] Fig. 10. Comparison between data and post-test prediction with TOPAZ.
ogy [5] to derive the multi-dimensional view factors. The nodalization for the two-dimensional TOPAZ computations is shown in Fig. 4 which at the same time indicates the positions MP1 through MP3 for the comparisons between the data and the open post-test predictions shown in Figs. 10 and 11. Figs. 10 and 11 show for INWA-experiments No. 2 and 4, respectively, that after a substantial learning period [3], the optimization of the code input parameters result in excellent agreements between measured data and computed results. Even worst-case scenarios, with failure of all heat sinks, as tested during experiment No. 4, can now
MP-I EXP.
'
l
--- .~:2c,~cI
200,
400.
600.
800.
1000.
[hi Fig. 11. Comparison between data and post-test prediction with TOPAZ.
The large-scale INWA-experiments have convincingly achieved all of their objectives cited above. Most importantly they demonstrated in large-scale the passive heat removal capability of the combination of PCI-vessel and composite reactor cell structure as suggested by Siempelkamp for the specified conditions examined. Detailed insights obtained from the experimental data allow one to properly specify input parameters for code prediction now, which were previously not readily available. Measurement techniques and evaluations were fully tested at high temperature levels and hundreds of hours test time for each experiment. The experience gained from these experiments may well supplement design needs of reactor cavity cooling systems developed elsewhere. The facility is available for more and possibly different experiments in the framework of international cooperation.
9. References [1] Battelle-Institut e.V., Feasibility Study of a Prestressed Cast Iron Reactor Pressure Vessel (in German), Vols. I and II (1973, 1974). [2] B. Beine, Integrated design of prestressed cast-iron pressure vessel and passive heat removal system for the reactor cell of a 200 MWth reactor, Prec. 10th SMiRT Post Conf. Seminar on Small and Medium Sized Nuclear Reactors, Anaheim, CA, USA, Aug. 1989. [3] B. Beine, Large scale test setup for the passive heat removal system and the prestressed cast-iron pressure vessel of a 200 MWth modular high temperature reactor, 3rd Intl. Seminar on Small and Medium Sized Nuclear Reactors, New Delhi, India, Aug. 16-28, 1991. [4] A.B. Shapiro, TOPAZ: a finite heat conduction code for analyzing 2-D solids, University of California, Lawrence Livermore National Laboratory, Rept. UCID-20045 (1984). [5] A.B. Shapiro, FACET: a radiation view factor computer code for axisymmetric, 2D planar, and 3D geometries with shadowing, University of California, Lawrence Livermore National Laboratory, Rept. UCID-19887 (1983).
Nuclear Engineering and Design 146 (1994) 409-416
Nuclear Engineedng Design
Passive heat removal experiments for an advanced HTR-module reactor pressure vessel and cavity design L. Wolf, A. Kneer, R. Schulz, A. Giannikos, W. H~ifner Battelle Ingenieurtechnik GmbH, Diisseldorfer Strasse 9, 65760 Eschbom, Germany
Dedicated to Prof. Dr.-Ing. H. Unger, Lehrstuhl fiir Nukleare und Neue Energiesysteme, Ruhr-Universit~it, Bochum, Germany, at the occasionof his 60th birthday.
Abstract In the search for further reducing the residual risks of possible major reactor pressure vessel failure during and in the aftermath of severe accidents of modular HTRs, an alternative RPV has been designed and a sample vessel already fabricated by the firm Siempelkamp, Krefeld, FRG. This alternative RPV design is made of high quality, ductile sphero east iron with axial and circumferential wire or more recently cireumferentially flat band prestressing. This specific Siempelkamp design has been tested and qualified already on behalf of a series of experiments in the sample test vessel. Also, this design was used for the control gas vessel in the THTR under operational service conditions. In order to demonstrate reliable decay heat removal under most severe conditions, a 1 : 1 scale, 20° sector of the vessel/cavity, termed INWA-facility (inactive decay heat removal) was fabricated and tested at Siempelkamp. The cavity was cooled by natural circulation of water flowing in tubes embedded into the cast iron structure of the cavity. A total of 5 experiments were performed with this setup examining a variety of changes in constructive details, surface and cooling conditions. Each experiment was performed both for operational conditions and depressurization transient, typical for a 200 M"~thHTR-module. Experimental test times ranged up to 1000 hours. Pre- and post-test predictions with the TOPAZ-code accompanied the INWA test series. The paper describes the INWA-faeility and reports some of the experimental results as well as the predictive capability of TOPAZ by comparing the data with computational results. The INWA-results qualify the pre-stressed cast iron vessel together with the natural circulation cooled cavity even for the worst of severe accident conditions. Even in case of total failure of all cooling capabilities in or at the cavity structure the vessel surface temperature remains below critical values.
1. Introduction Since 1973, the design and planning of German pebble bed high temperature, gas-cooled reactors was accompanied by the alternative reactor pressure vessel design using the concept of prestressed cast-iron (PCI) vessel [1]. In 1975, developmental work started and in the following
years, it was demonstrated, that PCI-vessels can be operated up to 350°C. The required materials for cast-iron, liner and wires were systematically examined by material testing programs which have been documented. Both, the control gas storage tank for the T H T R as well as the large-scale H T R test vessel at Siempelkamp, Krefeld, Germany, impressively
0029-5493/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0029-5493(93)E0246-G
410
L. Wolf et al. ~Nuclear Engineering and Design 146 (1994) 409-416
demonstrated the predictability, the feasibility as well as the licensibility by the authorities according to standard codes and regulations. Public acceptance of future nuclear power seemingly necessitates increased application of passive design features without resorting to active components. In addition, emphasis is put to further reduce remote, residual risk contributors, such as for instance pressure vessel failure. The prestressed cast-iron pressure vessel combined with a passive heat sink capability integrated into the reactor silo structure shows promising passive, and forgiving features satisfying a number of presently discussed requirements of nuclear systems of the future. In late 1987, the firm Siempelkamp proposed [2] such a combination of reliable pressure vessel with the advantages of a passive decay heat removal system fully integrated into the reactor silo cell structure. This proposal was based on the design features, overall measures and characteristics developed by SIEMENS/KWU (INTERATOM) for the 200 MWth HTR-modular reactor design, with the idea to substitute the PCI-vessel for the common steel vessel and to replace the silo structure by the composite cast-iron/concrete structure developed by Siempelkamp [2,3], as described in the following.
2. PCl-vessel/passive cast-iron-concrete composite silo structure system Figs. 1 and 2 show axial and transverse cuts through the overall system proposed. All major elements of this passive combination of PCI-vessel with advanced composite structure for the reactor cell structure are numbered and fully itemized in the footnotes of Figs. 1 and 2. The proposed system fits into the overall design of the 200 ~ t h SIEMENS/KWU HTR modular pebble bed reactor design with an operating system pressure of 6 MPa. During normal operating conditions, the energy loss across the pressure vessel of 350 kW is transferred by combined convective and radiative processes to the inside surface of the cavity, where
prestressed cast;iron pressure vessel (~)Reactor cell Axial prestressing tendon~(~ Top s l a b ~ ) Heat exchanger Hoop prestresstng cable ~.~ Ltner wtth support structure Coollng tubes ( ~ Section for test setup
Fig. 1. Vertical cross section of HTR-module with PCI-vessel/cavity design.
it is conducted to the cooling tubes embedded into the cast iron profile. Finally, the majority of the energy is transported away by the water coolant, the rest to the outside surface of the reactor silo structure. In case of an accidental depressurization transient of the system, the energy to be transferred to the cell cooling structure increases to 890 kW. The coolant will rise to the natural draft coolant tower by virtue of buoyant natural circulation. To
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
maintain the heat sink capability under all circumstances, two redundant coolant piping systems are used. During the developmental design of the aforementioned system, it became obvious that a number of input parameters for design and safety analyses were not known with sufficient accuracy. Due to its specific construction, based on a variety of lined-up individual components in series, the vessel component is much more complex than the presently used steel vessels. Therefore, concern was raised, whether the total entity of prestressed components can mitigate the decay heat released over a spectrum of operational and severe accidents as a number of complex combinations of conductive and radiative heat transport mechanisms prevail from the inside vessel through the various components and the special cavity towards the outside concrete surface of the vessel confinement compartment. In order to solve the open questions concerning decay heat transport
411
,/o
I 1 Electric heater and core vessel 2 Liner 3 Pressure vessel 4 Axial prestmcalng tendon 5 Hoop prestressing cable 6 Cover plate
7 Cast Iron block of the reactor call 8 Cooling system 9 Reinforced concrete 10 Sprinkler system 11 Insulation
Fig. 3. INWA test facility (dimensions in ram).
as well as to demonstrate the feasibility of the combined design a 1:1 scale, 20° sector of the vessel/cavity was fabricated and erected, termed INWA-facility, at Siempelkamp. Fig. 2, item 11, indicates the sector element chosen for the experimental demonstration.
3. INWA test facility
~)Core area (~)Ltner (])Liner support structure (~)Slock of cast iron body ..~Axtal prestruss~n9 tendons (~Honp Rrestresstng cable (~T)Reactorce;1 (~)R~nforced concrete "(~Cast iron proftles Q~Cooltng tubes ]Q~ Section for test setup
Fig. 2. Horizontal cross section of HTR-module with PCI-vessel/cavity design.
Fig. 3 gives a perspective view of the INWAFacility [3], which was constructed with all details of the realistic PCI-vessel and cast-iron/concrete composite structure with circumferential tendon reinforcement. The height of the sector was chosen to be about 2 m - only a fraction of the total axial system extension. Therefore, natural circulation
412
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
condition in the embedded tubes had to be simulated accordingly by appropriate forced convection condition. Also, because of the sector shape the hoop prestressing cable could not exert the proper force on the PCI-vessel structure as under normal operating conditions. However, the experimental structure was solidly damped at the floor. In order to provide an ultimate heat sink in case of failure of both redundant, natural circulation piping systems, a film cooling device, see 10 in Fig. 3, was installed to rinse down the outside reactor cell composite surface. The complete INWA-facility was embedded within 500 mm thick thermal insulation. Also, as shown in Fig. 3, the supporting base structure was carefully designed for its isolating properties. Super reflecting metal foil was applied for the floor, ceiling and sidefaces of the reactor cavity to provide proper thermal boundary conditions. All of these precautionary measures were necessary to keep heat losses from the INWA-facility as low as possible. Based on the experience elsewhere, the crucial role of the heat loss control was well
known to affect the successful performance of these types of experiments and their reliable interpretations. The driving thermal power is provided by the electric heater element, see item 1 in Fig. 3. This heating element consisted of 4 independently controlled heating circuits to assure a surface temperature as uniform as possible by properly accounting for the end heat losses. Thermographic control photos indicated maximum temperature deviations of only 4°C located towards the heater plate sides.
4. Measurement systems
In order to reach all objectives of the INWAtests, the complex three-dimensional structure was instrumented with a total of 123 embedded and surface thermocouples, the majority of which were placed along the radial centerline of the sector. Furthermore, to obtain temperatures and heat fluxes across gaps and contact interfaces between components, emphasis in instrumenta-
cooling system (heat sink) radiation radiation and ~ convection
radiation and . heat flux boundary
~
convection conduction
conduction
natural convection or film cooling
Fig. 4. Heat transport phenomena in the INWA-module, MP1-3: Positions of measurement and calculation comparison. Nodalisation of 2-D INWA FE-model.
L. Wolf et al. /Nuclear Engineering and Design 146 (1994) 409-416
tion was put to position a sufficient number of thermocouples at the relevant surfaces and at and into the wire packages. Furthermore, to properly evaluate the heat loss, thermocouples were placed towards the boundary sides of the facility. Special instrumentation was positioned inside the cavity to determine the temperature distribution, velocity fields and radiative transfer. For this purpose, 3 globe thermometers, 4 turbine anemometers and special thermocouples were attached at movable rakes. Four calorimeters were installed into the embedded cooling tubes. After the first preliminary experiment, the importance of correct heat flux control became apparent and a special heat flux sensor was additionally placed at the heater plate surface facing the liner component. This allowed redundant and diverse control of the heat flux boundary condition. All measurement signals were sampled, converted and documented by a PC-based system and software.
5. O~ective and test matrix
Fig. 4 shows a horizontal cut through the INWA-facility and at the same time indicates the
413
nodalization used for the thermal analysis with the FEM-Code TOPAZ [4]. Also, this figure reveals the variety and complexity of heat transport processes occurring in the different components of the PCI-vessel and composite reactor cell structure. The objectives of the INWA-experiments were as follows: (1) Demonstration of reliable and predictable passive heat sink capability under operational and accidental reactor conditions without invalidating thermal limits of individual components and the integrity of PCI-vessel structure. (2) Determination of appropriate values for thermal resistances, thermal conductivities (axial and circumferential packages of cable wires), and heat transfer coefficients at various interfaces. (3) Evaluation of the importances of convective and radiative heat transport mechanisms in gaps, most importantly across the cavity. (4) Determination of the effects of changes in constructive details (addition of a protective cover plate, shielding the circumferential cable package, surface treatment by adding a black coating onto the surfaces facing the cavity). (5) Evaluation of the impact of total failure of
Table 1 Summary of major characteristics of the 5 INWA-experiments Exp. No.
Description
Cooling
Film cooling
0
Heat flux increase (10%) 200 h to 370 h
on
off
1
Start of the DPE at 504 h Experiment with cover plate
on
off
2
Start of the DPE at 408 h
on
off
3
Start of the DPE at 577 h Black coating onto the surfaces facing the cavity
on
off
4
Start of the 1.DPE at 283 h Start of the 2.DPE at 788 h Black coating onto the surfaces facing the cavity
off off
on off
5
Start of the 1.DPE at 100 h Start of the 2.DPE at 350 b Steal vessel instead PCl-vessel
on off
off off
DPE = Depressnrization Event
414
L. Wolf et al. / Nuclear Engineering and Design 146 (1994) 409-416
INWA EXPERIMENT 5"
Tnmsient
i
Sleady-State Ol~'atlen -
L 0.
t00.
200.
300.
i 400.
000.
Fig. 5. Heat fluximposedon liner surface for INWA-tests.
the natural circulation heat sink capability a n d / o r cell outside surface film cooling on the thermal responses of the PCI-structures and the composite reactor cell structure. A total of 5 experiments have been performed and evaluated thus far, testing the PCI-vessel structure under different conditions cited above. Some characteristic features of these experiments are summarized in Table 1. A reference test with a section of a commonly used steel vessel is presently underway. Upon its completion, a direct comparison and demonstration of PCI versus steel vessel behavior is possible. All experiments were performed using the transient heat flux profile depicted in Fig. 5 for controlling the heater plate in order to reach steady-state operational or to simulate accident conditions according to the predicted behavior during a depressurization accident of the modular HTR-reactor designed by S I E M E N S / K W U . The curve shown in Fig. 5 accounts for estimated heat losses (20%) of the INWA-facility.
21)0,
4~.
600.
~,
triO0
R A D I A L C O O R D I N A T E F R O M H E A T P L A T E [mm]
600.
rIME[hl
9.
Fig. 6. Comparisons of cylinder block temperature profiles (4~= IIY, normal operation).
tx~ t-
O
I~0.
500,
1500,
R A D I A L C O O R D I N A T E F R O M C Y L I N D E R R I B [mrn]
Fig. 7. Comparisons of radial temperature profles across cavity(normal operation).
5" L.~ ~e
6. Experimental results
Figs. 6 through 9 present measured radial temperature profiles through the PCI-structure and cavity for steady-state normal operation and the depressurization transient, respectively. These
0.
........ ~0. • .....
4ee.
6e0.
~o.
t~0.
R A D I A L C O O R D I N A T E F R O M H E A T P L A T E [ram] Fig.
8.
Comparisons of cylinder
block
(¢) = 10°, depressurizationtransient).
temperature
profiles
415
L. Wolf et al. ~Nuclear Engineering and Design 146 (1994) 409-416
figures show the results of all experiments without protective cover plate shielding the hoop prestressing cables. The comparisons between individual curves in the respective figures clearly demonstrate the small variance of the experimental results, thereby demonstrating consistency and plausibility of the data despite of thousands of hours of operations of the INWA-facility. Under normal, steady-state operation, maximum temperatures at the liner surface range around 275°C (Fig. 6). The radial temperature profiles across the cavity are essentially fiat (Fig. 7), with steep gradients at the PCI-vessel and inside cavity surfaces, indicating a natural circulating flow pattern along the structural surfaces with a large, essentially stagnant core in the cavity. Changes made for the different experiments have only benign effects on the steady-state temperature profiles. Figs. 8 and 9 show the radial temperature profiles for experiments 2, 3 and 4 at specific instants in time. Again, the variance between the curves is very small. Maximum liner temperatures reach a value of about 470°C. For experiment No. 4 the radial temperature gradient decreases towards the cylindrical fin of the outside PCI-vessel surface because this experiment was performed without the natural circulation heat sink capability in the composite reactor cell structure. The
Table 2 Heat transportby radiation and convectionbetween the PCI and the inside cavitywall Exp. Time Radiation Convection Radiation Convection No. (h) (W/m 2) (W/m 2) (%) (%) 0 1 2 3 4
190 400 650 330 530 513 750 250 450 950
764 378 1170 701 1744 708 1786 672 1722 1712
61,3 185 334 61,4 105 55 91,6 53 54 49
92,6 67,2 78,0 92,0 94,3 92,8 95,1 92,7 97,0 97,2
7,4 32,8 22,0 8,0 5,7 7,2 4,9 7,3 3,0 2,8
simulated failure of this system leads to a temperature increase of the inside cavity surface and as a result to a reduction in heat transport from the vessel surface across the gap. This can be clearly seen from Fig. 9 which shows that the gap temperature level increases from around 1500C for e x p e r i m e n t s 2 and 3 and up to 220"C for experiment 4. A concise error analysis for the temperature sensors resulted in absolute errors in the interval [ + 1,7°C, +3°C] which is too small to be presented in the figures. Table 2 summarizes the evaluation of the special measurement techniques applied in the cavity for the determination of the respective fractions of convective and radiative heat transport across the cavity. The respective values indicate the overwhelming importance of the radiative transport (92-97%) versus the convective transport (3-8%), with the exception of INWA experiment No. I which was performed with a shielding cover plate protecting the azimuthal wire package.
Lrd
7. Accompanying numerical analyses b"
0.
5~.
1000.
151X).
R A D I A L C O O R D I N A T E F R O M C Y L I N D E R R I B [mm]
Fig. 9. Comparisons of radial temperature profiles across cavity (depressurization transient).
All INWA-experiments were accompanied by design, parameter sensitivity-, as well as bestestimate pre- and post-test computational analyses with the Finite Element-code T O P A Z [4]. The T O P A Z - c o d e was extended by including 2-D radiation properties using the FACET-methodol-
L. Wolf et aL /Nuclear Engineering and Design 146 (1994) 409-416
416
be predicted with confidence as depicted in Fig. 11. -'-
MP-3 E:EP
....
~*~-2 CALC
" -"
I4~-3CAL~
8. Conclusions .
...:'2=2'-='2:'.: 2 : . . . . .
g. d
~m,
~0.
TIME [h] Fig. 10. Comparison between data and post-test prediction with TOPAZ.
ogy [5] to derive the multi-dimensional view factors. The nodalization for the two-dimensional TOPAZ computations is shown in Fig. 4 which at the same time indicates the positions MP1 through MP3 for the comparisons between the data and the open post-test predictions shown in Figs. 10 and 11. Figs. 10 and 11 show for INWA-experiments No. 2 and 4, respectively, that after a substantial learning period [3], the optimization of the code input parameters result in excellent agreements between measured data and computed results. Even worst-case scenarios, with failure of all heat sinks, as tested during experiment No. 4, can now
MP-I EXP.
'
l
--- .~:2c,~cI
200,
400.
600.
800.
1000.
[hi Fig. 11. Comparison between data and post-test prediction with TOPAZ.
The large-scale INWA-experiments have convincingly achieved all of their objectives cited above. Most importantly they demonstrated in large-scale the passive heat removal capability of the combination of PCI-vessel and composite reactor cell structure as suggested by Siempelkamp for the specified conditions examined. Detailed insights obtained from the experimental data allow one to properly specify input parameters for code prediction now, which were previously not readily available. Measurement techniques and evaluations were fully tested at high temperature levels and hundreds of hours test time for each experiment. The experience gained from these experiments may well supplement design needs of reactor cavity cooling systems developed elsewhere. The facility is available for more and possibly different experiments in the framework of international cooperation.
9. References [1] Battelle-Institut e.V., Feasibility Study of a Prestressed Cast Iron Reactor Pressure Vessel (in German), Vols. I and II (1973, 1974). [2] B. Beine, Integrated design of prestressed cast-iron pressure vessel and passive heat removal system for the reactor cell of a 200 MWth reactor, Prec. 10th SMiRT Post Conf. Seminar on Small and Medium Sized Nuclear Reactors, Anaheim, CA, USA, Aug. 1989. [3] B. Beine, Large scale test setup for the passive heat removal system and the prestressed cast-iron pressure vessel of a 200 MWth modular high temperature reactor, 3rd Intl. Seminar on Small and Medium Sized Nuclear Reactors, New Delhi, India, Aug. 16-28, 1991. [4] A.B. Shapiro, TOPAZ: a finite heat conduction code for analyzing 2-D solids, University of California, Lawrence Livermore National Laboratory, Rept. UCID-20045 (1984). [5] A.B. Shapiro, FACET: a radiation view factor computer code for axisymmetric, 2D planar, and 3D geometries with shadowing, University of California, Lawrence Livermore National Laboratory, Rept. UCID-19887 (1983).