thepowerindustry’sview ofpast,present,and futuretwo ...figure1 bensontestrig....

HeatTransfer Engineering, 21:5± 17, 2000Copyright C°° 2000Taylor &Francis0145± 7632/00$12.00+ .00

heat in history

The Power Industry’s Viewof Past, Present, andFuture Two-Phase FlowTesting

HOLGER SCHMIDT, WOLFGANG KASTNER, and WOLFGANG KOHLERSiemens AG, Power Generation Group (KWU), Erlangen, Germany

Since 1974, Siemens’ Power Generation Group (KWU) has been operating a high-pressuretwo-phase � ow test loop—called the Benson test rig—which offers a range of operating conditionsthat is unique in the world (1 to 330 bar, 20 to 600±C, and 0 to 2 MW electric heating power). The25th anniversary of the � rst tests performed at this test rig presents a good occasion not only forreviewing the past, but also for contemplating the future of two-phase � ow experiments.

The past was characterized by integral and separate effect tests for power generation usingnuclear, fossil, and renewable energy sources as well as for process industries. This article willpresent examples demonstrating the � exible and broad range of applications for the Benson test rig.The results of the tests have been used to develop algorithms for implementation in computerprograms and also for validating such programs.

Usually these computer programs—so-called analysis tools—are used for analyzing systems orcomponents. From an analyst’s point of view, two-phase � ow experiments serve either to verifyglobal � ow conditions or to supply inputs such as boundary conditions and material laws and/orinitial conditions for the analysis tools. An advanced way of making sure that all availableknowledge can be input into the analysis tools is to collect and store it in a program system fromwhich it can be called up, whenever required, according to the task in hand. Siemens’ KWU Grouphas started developing such a system. Apart from integral tests conducted for new power plants,future two-phase � ow experiments will probably focus on expanding this program system’s database.

Siemens’ Power Generation Group (KWU) has beenoperating the Benson test rig—a test rig that is unique in

A slightly different version of this article was presented at the 37thEuropean Two-Phase Flow Group Meeting in London, 1999.

Address correspondence to Dr. Holger Schmidt, Siemens AG, PowerGeneration Group (KWU), P.O. Box 3220, 91050 Erlangen, Germany.E-mail: [email protected] e

theworld due to its wide rangeof operating conditions—in one of its accredited laboratories since 1974. The testrig’s 25th anniversary is an exceptional event for any testfacility and thus offers a good occasion for reviewingthe past and contemplating the future. As the name ofthe test rig implies, its main purpose is to investigatetopics associated with the design, operation, or further

5

Figure 1 Benson test rig.

development of Benson boilers, i.e., fossil-� red once-through boilers. However, the test facility’s � exibilityalso enables it to be used for other applications in the nu-clear and renewable power industries as well as processindustries. Some of the investigation s have been per-formed in cooperation with universities and researchinstitutes . Based on our 25 years of industrial experi-ence concerning two-phase � ow tests and applications ,the following should also throw light on the purpose offuture two-phase � ow testing.

BENSON TEST RIG

Figure 1 shows a section of the Benson test rig, andFigure 2 its � ow diagram. Its range of operating condi-tions is unique throughout the world:

Pressure 1–330 barTemperature 20–600 ±CPower 0–2 MWMass � ow 0–28 kg/s

The main sections of the test facility comprise a watertreatment system including injection and recirculationsystems, the object to be tested, a pressurizer, and acooling system. Demineralized and deaerated water isinjected into the test loop by a piston pump. A damp-ing vessel is installed directly downstream of the pistonpump to prevent oscillations . The test rig can be oper-ated in the recirculation mode for high � ow rates or inthe once-through mode for low � ow rates. To maintain aconstant water chemistry, a certain amount of the wateris exchanged in the recirculation mode.

The water is heated by being passed through onemain heater and up to two preheaters. Depending onthe size of the object being tested, heating can take theform of direct electric heating—i.e., the wall of the testobject acts as a resistor—or indirect heating via wiresmounted on or in the wall. Since the preheater as well

Figure 2 Benson test rig, � ow diagram.

6 heattransfer engineering vol. 21 no. 4 2000

Figure 3 Data acquisition and visualization.

as the test object are electrically heated, it is possible toestablish any enthalpy at the inlet and outlet of the testsection via energy balances. Heat losses are included inthis balance despite the fact that they are very low dueto the insulation, which is over 50 mm thick. A spray-type condenser installed downstream of the test objectcondenses and subcools the water.

System pressure is adjusted by a large thermal pres-surizer and a throttling valve downstream of the testobject. Assurance of a constant pressure during steady-state tests is very important, especially near the criticalpoint, because pressure oscillations affect heat transferconditions in the test object. With this combination ofvalve and pressurizer it is also possible to perform testswith prede� ned pressure gradients.

The Benson test rig as well as the test object are in-strumented in such a way that all relevant parametersare measured. The data acquisition and visualizationequipment has been constantly adapted to the state ofthe art over the past 25 years. Figure 3 shows the struc-ture of the present data acquisition and visualizationsystems. Visualization on two screens, as shown in Fig-ure 3, ensures that the operator has a clear picture at alltimes of current thermal-hydraulic conditions as wellas measured parameters such as the temperatures in thewall of the test object.

DESCRIPTION AND CHARACTERIZATION OFTYPICAL RESEARCH WORK

Regardless of the speci� c types of experiments thathave been performed at the Benson test rig, all of thetests can basically be broken down into the followingtwo categories:

1. Integral tests. The aim of these tests is to verify thefunctional behavior of components such as heat ex-changers, or to develop measuring equipment.

2. Separate effect tests. The aim of these tests is toinvestigate separate two-phase � ow effects. They canbe subdivided into two groups:a. Stand-alone tests. These tests are interesting in

their own right, such as investigations of erosioncorrosion.

b. Boundary tests. The aim of these tests is to in-vestigate � ow behavior in or around structures.The information gained from the tests is used incomponent analyses. In many cases, this infor-mation comprises boundary or initial conditionsand material laws for analysis tools such as � nite-element (FE) programs, which need heat transfercoef� cients as inputs for temperature distributioncalculations.

Integral Tests

Examples of integral tests are functional tests carriedout on a safety condenser (SACO) and tests performedto develop a compact wet steam measuring system.

In the case of the SACO, the aim of the experimentswas to determine the heat exchange characteristic ofthe component which had been developed for the nextgeneration of pressurized water reactors (PWR). Thedesign of the SACO is based on the idea of using analmost passive-decay heat removal system for the eventof a loss of heat sink on the secondary side. Figure 4shows how the SACO could be installed downstreamof a steam generator. In the event of a secondary-sideloss of the heat sink, the steam would be condensed byevaporating water on the tertiary side. As the tempera-ture differences are expected to be high, heat � uxes ofup to 600 kW/m2 would occur at mass � uxes of less than40 kg/m2 s. The pressure would be near the containmentpressure and thus in the range 1.35–5 bar. The SACOis designed as a tubular heat exchanger with conden-sation taking place inside the tubes. Heat transfer wasinvestigated in two phases. In the � rst phase, only thetertiary side was considered by electrically heating tube.In the second phase, a section (3 £ 3 tube bundle) of theSACO—shown on the right-hand side of Figure 4—wasbuilt into the Benson test rig. The secondary-side steamwas provided from the Benson test rig itself. The heightof all connecting pipes was scaled 1:1 in order to con-sider elevation effects. During the test campaign, heatexchange characteristics as well as transient behaviorwere investigated .

The main result of these tests was that the heat ex-change behavior of the SACO was veri� ed. For ex-ample, Figure 5 shows the exchangeable heat � ow for

heattransfer engineering vol. 21 no. 4 2000 7

Figure 4 Investigation on the integration of a safety condenser (SACO) to the secondary side of a PWR.

steady-state conditions . One interesting effect of theheat exchange characteristic is that the exchangeableheat � ow is independent of the pressure on the sec-ondary side over a wide range. An explanation for thisis that the tertiary side dominates the heat exchangeprocess. As the tests showed, the boiling crisis occursat high steam qualities. In the postdryout region, heattransfer is low compared to the regions with nucle-ate and subcooled boiling, and thus the contributionto heat exchange is also small. The experiments alsoveri� ed that the critical steam quality, which is thesteam quality at which the boiling crisis occurs,depends mainly on the mass � ux and is largely indepen-dent of the heat � ux. In the case of a higher secondarypressure, the temperature difference between the sec-ondary and tertiary sides is larger, and thus the heat � uxin the nucleate boiling region is also higher than whensecondary-side pressure is low. This means that, for highsecondary-side pressures, a smaller heat exchange sur-face area is needed to reach the critical steam quality.In each case, the same amount of heat is exchanged atsimilar mass � uxes, since only the area involved priorto the boiling crisis is relevant for the heat exchangeprocess and the magnitude of heat exchange is simi-

lar at approximately constant critical heat � uxes (seealso [1]).

Another typical kind of integral test is for calibratingmeasuring equipment. For example, Siemens’ KWUGroup developed a mass � ow and steam quality measur-ing device—shown schematically in Figure 6—whichwas then calibrated in the Benson test rig. This systemconsists of a venturi tube, a one-beam gamma densito-meter, a temperature sensor, a differential pressure sen-sor, and an absolute pressure sensor. The � ow throughthe venturi creates a pressure drop that is proportiona lto the square root of the mass � ow for a certain densityand is measured by the gamma densitometer. The � owthrough the venturi leads to such a high degree of ho-mogenization that a one-beam gamma densitometer canbe used. Based on the measured temperature and pres-sure, the properties of the � uid are taken into account.The calibration measurements were used to develop analgorithm for calculating the steam quality, the mass� ow, and the � uid enthalpy (see also [2]).

Use of such a device installed at the end of the boilersection to control the feedwater of a fossil-� red once-through steam generator was compared to the methodpresently employed and was found to have certain


Figure 5 SACO power with gravity and forced feed.

advantages. Up until now, it has been common prac-tice to measure temperature and pressure downstreamof the � rst superheater. However, measurement of theenthalpy at the outlet of the boiler results in faster re-sponse, meaning that the effects of changes in � ringor transient heating effects are detected earlier. Thishas a positive effect on the service life of the high-pressure components installed downstream of the boilersection.

Stand-Alone Tests

The aim of stand-alone tests is not to develop a data-base which can be used to supply boundary conditionsfor component analyses. Instead, the results of thesetests are interesting just in their own right. For exam-ple, tests have been performed at the Benson test rig toinvestigate erosion corrosion, a corrosion mechanismthat is greatly in� uenced by � ow behavior. This type ofcorrosion occurs only if the wall is wetted—for exam-ple, in the case of single-phase � ow or two-phase � owwith bubbly, strati� ed, or annular � ow patterns—and ifthe corrosion resistance of the wall is dependent uponthe formation of a protective oxide layer. The condi-

Figure 6 Wet steam measuring system.

tions promoting this kind of corrosion are so complexthat it is not possible to describe them in a way that en-ables the two-phase � ow conditions to be investigatedseparately from the metallurgical and chemical effects(e.g., as mass transfer coef� cients). Hence, the reduc-tions in wall thickness were measured directly. For thispurpose, various objects were tested in the Benson testrig (plates, tubes, bends, reducers, and increasers). Wallthinning was measured for conditions based on vari-ous � ows, temperatures, water chemistries, and mate-rials. Using this data—a couple of examples are shownin Figure 7, where the experimental determined wallthickness reduction due to erosion corrosion is plot-ted as the wall thinning measured in mm/year versussome typical parameters—a computer program (calledWATHEC) was developed which is employed in powerplant design to avoid signi� cant erosion corrosion. Forexisting power plants, it is also possible to identify lo-cations at which the wall thickness should be measured,meaning that the scope of in-service inspections can bereduced (see also [3–5]).


Figure 7 Wall thinning due to erosion corrosion (effect of thermalhydraulics and water chemistry).

Boundary Tests

Most of the tests performed at the Benson test righave been part of component analyses for the power andprocess industries. In the following, examples will bepresented of individual boundary tests carried out forthe fossil, nuclear, and solar power industries as wellas for a process industry in order to demonstrate the� exibility of the Benson test rig and to explain typicalboundary test applications .

Figure 8 Wall temperature and pressure loss in an uniformly heated, vertical, smooth evaporator tube.

Waterwall Tubes for Benson Boilers

Benson boilers are once-through steam generators.The steam is generated in tubes which are welded to-gether to form gas-tight walls. These walls surroundthe combustion chamber where the fuel is � red. In de-signing the combustion chamber, both the � ue-gas sideand the water side have to be taken into consideration.The steam quality at the outlet of a once-through boileris higher than 1; i.e., the steam is superheated, withthe consequence that all of the � ow patterns shown inFigure 8 as well as the boiling crisis occur in the boilersection. Subcooled boiling also typically occurs, this isindicated in Figure 8 by two different de� nitions of thesteam quality ( Çx and Çxmass ). From the design point ofview this is the energy-balanced steam quality Çx—in thefollowing called just steam quality—which is de� nedas the ratio of the difference between the cross section-averaged enthalpy minus the saturation enthalpy and theheat of evaporation. The mass-balanced steam qualityÇxmass is de� ned as the ratio of the actual steam to thesum of steam and water � ow. The mass-balanced steamquality together with the energy-balanced steam qualitygive a hint as to where a nonequilibrium condition mayoccur in a cross section. For the design of boilers, it isessential to know where the boiling crisis is located andto have information on heat transfer at this position. Itis generally an advantage if the boiling crisis occurs athigh steam qualities and thus at a high position in thewaterwall that is relatively far away from the burners.Another advantage of high critical steam qualities is therelatively good heat transfer behavior due to the highsteam velocities. In view of the relevance of heat trans-fer conditions in waterwall tubes, a variety of different


Figure 9 Test matrix for heat transfer and pressure drop investi-gations.

measurements have been performed in the Benson testrig. The main parameters of these tests are summarizedin Figure 9. Different types of tubes were installed inthe Benson test rig downstream of the heater so that theinlet steam quality could be varied. The electrical resis-tance of their walls was employed for heating the tubes.In some of the tests, the wall thickness was reducedover half of the tube circumference to simulate nonuni-form heating. Pressure drops as well as wall tempera-tures were measured for different � ow conditions . Us-ing the results of these measurements, algorithms weredeveloped which are being applied in steam generatordesign.

As an example taken from the database, Figure 10shows the in� uence of the mass � ux on the inner walltemperature. At low mass � uxes, the boiling crisis oc-curs at lower steam qualities and the waterwall temper-atures are high. In the case of 900 kg/m2 s, they wouldbe much higher than the creep temperature of typicalwaterwall materials. For this reason, mass � uxes for200 bar in the range between 1,400 and 2,000 kg/m2 sare required. The tubes have to be installed at an in-clination of approximately 30± because the size of thecombustion chamber is governed by the design of the� ue-gas side. This inclination requires supporting struc-tures in order to avoid excessive bending stresses in thetubes. The waterwall tubes can only be installed in avertical position if ri� ed tubes are used. The enhancedheat transfer behavior of ri� ed tubes is demonstrated byFigure 11, which shows the inner wall temperatures oftwo uniformly heated tubes, one smooth and one ri� ed.Although � ow conditions are similar, the boiling crisisoccurs in the ri� ed tube at a much higher steam quality

Figure 10 Effect of mass � ux on the inner wall temperature of asmooth tube.

than in the smooth tube. The reason for this effect is theswirl induced by the ri� e geometry. The correspondingcentrifugal forces imparted to the liquid phase inside thetube are responsible for rewetting of the wall and thusfor the high critical steam quality. More than 100,000temperature measurements taken on smooth tubes and150,000 on ri� ed tubes presently form the basis forheat transfer correlations which, together with correla-tions for pressure drop, are key elements of design andanalysis tools for Benson boilers. These tools include,for example, dynamic programs for analyzing startupand shutdown processes, and FE programs for calcu-lating stresses and temperatures in waterwalls (see also[6–10]).

Figure 11 Wall temperatures in smooth and ri� ed tubes.


Figure 12 Test setup heat removal during the accident at TMI-2.

Heat Transfer between a Debris Crust and the ReactorPressure Vessel Wall

During the Three Mile Island-2 (TMI-2)accident, thecore of the reactor was partly melted. The molten de-bris penetrated into the lower plenum. Contrary to mostassumptions, subsequent inspections revealed that thewall of the reactor pressure vessel (RPV) showed nosigns of melting. This discovery could not be explainedby the existing reactor safety codes. One main questionwas how large themaximum removable heat would havebeen if a gap between the debris crust and the RPV wallhad been � lled with water still present in the vessel. Ina joint research program headed by the German Soci-ety for Reactor Safety (GRS), Siemens’ KWU Groupinvestigated the heat transfer conditions in such a gap.For this purpose a vessel was built and integrated intothe Benson test rig as shown in Figure 12. The lowerpart of the vessel had a spherical shape, similar to thebottom head of the RPV. In order to simulate the debriscrust, a heating element was mounted in the vessel, thiselement being movable in order to vary the gap width.Figure 13 shows the measured critical heat � uxes. They

Figure 13 Validation of the heat removal from the molten debrisduring the TMI-2 accident.

are lower than the corresponding heat � uxes during theTMI-2 accident. Although a detailed analysis of theTMI-2 accident would be very complex, comparison ofthe heat � uxes gives a rough explanation of why theRPV retained its structural integrity. In order to obtaina deeper understanding of the accident and to transferthe new knowledge about heat transfer behavior in thegap to safety analyses for existing plants, it is possibleto use an algorithm based on the results of these teststo describe the heat transfer boundary conditions (seealso [11]).

Heat Transfer in Solar Farm Absorber Tubes

Among the various concepts for utilizing solar en-ergy, solar thermal power generation in solar farms isthe most widely used. At present, only farms employ-ing thermal oil in absorber tubes to generate steam in aseparate heat exchanger are in operation. As this ther-mal oil limits the maximum process temperatures andis also dif� cult to handle, new process concepts are un-der development to generate the steam directly in theabsorber tubes. These processes are shown schemati-cally in Figure 14. They are to be tested at Plata FormaSolar in Almeria (Spain). In a step preceding this de-velopment, the heat transfer conditions were investi-gated in the Benson test rig as part of a joint researchproject headed by the German Aerospace Research Es-tablishment. The � ow diagram of these tests is givenin Figure 15, along with a picture showing the tubesin front of the building of the Benson test rig. Varioustube geometries as well as heat � uxes and � ow con-ditions were tested. Pressure drops and temperatureswere measured. As an example, Figure 16 shows theinner wall temperatures. The results of these tests havebeen used to develop algorithms used in design studiesfor the solar power industry. One speci� c feature re-quiring investigation is deformation of the tubes due tononuniform heating in order to avoid positioning out-side of the focus of the mirrors or destruction of thesurrounding glass insulation. Performing these kinds ofinvestigation s with an FE program requires heat trans-fer coef� cients as boundary conditions which can beselected using the results of the tests performed in theBenson test rig (see also [12]).

Pressure Drop Tests

In most of the tests conducted at the Benson testrig, temperatures and pressure drops were measured.The pressure drops are important for component designand analysis as well as heat transfer behavior. Basedon knowledge about pressure drops it is possible, forexample, to analyze mass � ow distributions in piping


Figure 14 Concepts for direct solar steam generation.

networks or to size pumps, etc. Frictional pressure dropsare typically represented as two-phase � ow multipli-ers, which are dependent on � ow conditions and ge-ometry. For instance, it is interesting that the generalcharacteristics of the two-phase � ow multipliers forsmooth and ri� ed tubes are different, as can be seen fromFigure 17.

Pressure drops were the focus of a series of inves-tigations of steam piping networks for the heavy-oil

Figure 15 Test setup for direct steam generation investigations.

industry. These steam networks are used, especially inVenezuela, to inject steam into boreholes to increaseoutput. As these networks are of a large size, a portionof the steam condenses inside thepipes. Many investiga-tions were carried out to develop or validate algorithmsfor the pressure drop, including tests at the Benson testrig. Figure 18 shows the setup used for the test objectand the main parameters of the test matrix. Figure 19shows some selected results for a pipe bend, an inclined


Figure 16 Temperatures at the upper side of the absorber tube.

Figure 17 Effect of mass � ux on pressure drop of smooth andri� ed nonheated tubes.

Figure 18 Test setup for pressure drop measurements in a pipeconveying wet steam.

Figure 19 Two-phase � ow pressure drop in various pipes.

pipe, and a horizontal pipe. The results were integratedinto a simulation program which uses them mainly asboundary conditions (such as frictional pressure drops)or as material laws (such as for two-phase mixture den-sities) (see also [13]).

FUTURE PROSPECTS FOR TWO-PHASE FLOWEXPERIMENTS

From the point of view of the power industry, two-phase � ow investigation s are interesting as long as thesteam cycle is an essential element of power genera-tion. At present, most power plants comprise gas tur-bine plants combined with a steam cycle (combined-cycle plants), coal-� red plants, and nuclear plants. Atcombined-cycle plants, the steam cycle is of minor im-portance. The main requirements to be met by the steamcycle are ease of design and low-cost construction. Atcoal-� red and nuclear power plants, the steam cyclesare very important. Especially in the case of new nu-clear power plant designs, such as the EPR and the SWR1000, it can be expected that integral tests will be nec-essary in order to verify the functional behavior of theirnew systems. These tests are dependent on the stagereached in the design process and it is dif� cult to fore-see what kind of integral tests will have to be performed,and when. The situation is similar with regard to stand-alone tests. The requirements for new stand-alone testswill be governed by new developments in power plantwater treatment, such as injection concepts for reducing60Co activity at nuclear power plants (see also [14]).


Figure 20 Engineering process for systems with two-phase � ow.

Boundary tests are related to analysis tools and to thespecial conditions associated with certain applications .Hence, it can be expected that tests for waterwall tubesin, for example, once-through boilers will continue.From a general point of view, one can see two trends.First, competition in the power plant market is verystrong, meaning that those who want to survive on thismarket will have to optimize their engineering pro-cesses. The second major trend is that the commerciallyavailable analysis tools will become more and morepowerful and user-friendly. Looking at the engineeringprocesses for systems in which two-phase � ow occurs,one � nds a structure of the kind shown schematicallyin Figure 20. The two main steps are general model-ing, which includes description of the geometry and theloads, and two-phase � ow input, which includes the de-scription of material laws as well as initial and boundaryconditions.

One example taken from an engineering process isanalysis of the deformation of a header. For the analysisthe header is a piping system with one big central tubeand a large number of inlets and outlets. Under full-load conditions, it conveys superheated steam. Duringstart-up, some of the inlets on the sides may be closedand the steam may be in a saturated state. This may re-sult in condensation in some sections of the header. Toinvestigate the deformation of such a system it is com-mon practice to use � nite-element programs. In a � rststep the temperature distribution in the header is calcu-lated, which requires heat transfer coef� cients—withand without phase change—as boundary condition, inorder to determine the deformation in a second step withstandard tools. The general procedure as well as tem-perature and deformation distribution for one time stepare shown schematically in the middle of Figure 20. An-other example is the estimation of the change in inven-tory in a drum boiler during load changes. These kindsof investigations are typically performed with simula-tion tools by solving the conservation equations for themass, the momentum, and the energy. Therefore it isnecessary to consider the structure of the boiler as well

Figure 21 Automated two-phase � ow input.

as the control system as it is shown schematically at thebottom of Figure 20. By considering the water–steammixture in the boiler it is possible to neglect the two-phase � ow frictional and the acceleration pressure drop.The remaining two-phase-� ow input is the density ofthe two-phase mixture, which can be regarded as a ma-terial law. In both of the examples shown in Figure 20,two-phase-� ow inputs are required. When one com-pares the modeling and input steps, it becomes clearthat the modeling tools have reached a high standardwhich will become even higher in the future. The two-phase-� ow input step is unique to each task in hand andis often very time-consuming and expensive. In manycases only a small portion of the knowledge available ontwo-phase � ow is taken into consideration, somethingwhich also has an in� uence on quality.

Siemens’ KWU Group has started developing a pro-gram (called In-2-Flow) for automating the input stepin such engineering processes, as outlined in Figure 21.This program will be employed in the general engineer-ing process after speci� cation of the modeling inputsand before performance of the analysis. It will automatethe two-phase � ow inputs. The � rst step is to assess theavailable methods. For tubes, for example, a large vari-ety of different methods are available for calculating thevoid fraction, pressure drop, and heat transfer for var-ious applications . Based on the most suitable method,it will then be possible to prepare the inputs in orderto link them to the analysis tool. As can be seen fromFigure 22, a database is the core component of such

Figure 22 Main structure of the input tool.


a program. Although only heat transfer and pressuredrop are mentioned in Figure 22, it is also planned forother two-phase � ow data such as � ow forces or leak-age mass � ows through small cracks (see also [15, 16])to be considered as well. Relevant two-phase � ow datawill be collected in a database. For the purposes of thisprogram, it will be necessary to store not only equationsbut also all additional information in order that a deci-sion can be made. A search assistant will help the user� nd the most suitable method, although it might turnout that different methods should be used for differentapplications within the same simulation. Depending onthe requirements of the simulation tool, it will be nec-essary to prepare the inputs as single values, tables, orsubroutines . Initially, the knowledge base of this pro-gram will be taken from the available literature. Storinginformation in a database will also provide other advan-tages. One of these is that certain items of knowledgecan be permanently retained. Also, a program of thiskind can provide an overview of all of the knowledgethat is available on a speci� c topic. This can be helpfulfor identifying areas in which there is a lack of infor-mation. Missing information can then be supplied byperforming new boundary tests as part of a continuousprocess for enhancing the quality of the input tool.

CONCLUSION

Although a large variety of different two-phase � owexperiments have been performed in the past, tests ofthis kind will still be needed in the future. These testswill be related to new developments in the � elds ofcoal-� red and nuclear power generation and will takethe form of integral tests or boundary tests for, e.g.,waterwall applications . In addition, it can be expectedthat two-phase � ow tests will also be performed in thefuture for enhancing the quality of a data input programcurrently under development at Siemens, the aim of thisprogram being to automate the input of two-phase � owdata in modern analysis tools.

REFERENCES

[1] Kohler, W., Herbst, O., and Kastner, W., Thermal-HydraulicBehavior of a Safety Condenser, Int. Conf. on New Trends inNuclear System Thermohydraulic, vol. 1, pp. 609–614, 1994.

[2] Kastner, W., Fischer, C., and Kratzer, W., Verbesserte Spe-isewasserregelung durch kompaktes Meßsystem zur Massen-strom- und Dampfgehaltsbestimmung, BWK, vol. 45, no. 12,pp. 510–514, 1993.

[3] Heitmann, H. G., and Kastner, W., Erosion-Corrosion inWater-Steam Cycles—Causes and Countermeasures, VGBKraftwerkstechnik, vol. 62, no. 3, pp. 180–187, 1982.

[4] Kastner, W., Riedle, K., and Tratz, H., Experimental In-vestigations on Material Loss Due to Erosion-Corrosion,VGB Kraftwerkstechnik, vol. 64, no. 5, pp. 411–423,1984.

[5] Kastner, W., Nopper, H., and Roßner, R., Vermeiden von Ero-sionskorrosionsschaden in Rohrleitungen, 3R Int., vol. 33,no. 8, pp. 423–428, 1994.

[6] Griem, H., Kohler, W., and Schmidt, H., Heat Transfer, Pres-sure Drop and Stresses in Evaporator Water Walls—From Ex-periment to Design, VGB PowerTech, vol. 79, no. 1, pp. 26–35,1999.

[7] Kohler, W., Kefer,V., and Kastner, W., Heat Transfer in Verticaland Horizontal One-Side-Heated Evaporator Tubes, Exp. HeatTransfer, pp. 397–409, 1990.

[8] Kefer, V., Kohler, W., and Kastner, W., Critical Heat Flux(CHF) and Post-CHF Heat Transfer in Horizontal and InclinedEvaporator Tubes, Int. J. Multiphase Flow, vol. 15, no. 3,pp. 385–392, 1989.

[9] Kohler, W., and Kastner, W., Heat Transfer and Pressure Lossin Ri� ed Tubes, 8th Int. Heat Transfer Conf., vol. 6, pp. 2861–

2865, 1986.[10] Franke, J., Kohler, W., and Wittchow, E., Evaporator Designs

for BENSON ° R Boilers—State of the Art and Latest Develop-ment Trends, VGB Kraftwerkstechnik, vol. 73, no. 4, pp. 307–

315, 1993.[11] Schmidt, H., Kohler, W., Herbst, O., and Kratzer, W., Exper-

iments on Heat Removal in a Gap between Debris Crust andRPV Wall, 1st European-Japanese Two-Phase Flow GroupMeeting, 1998.

[12] Kohler, W., Herbst, O., and Kastner, W., Thermal Designof Solar Absorber Tubes with Direct Steam Generation, 8thInt. Symp. on Solar Thermal Concentrating Technologies,1996.

[13] Gonzales, M., Kastner, W., Kunstle, K., Lezuo, A., Ostendorf,F. J., Reiter, K., Rodriguez, J., and Thelen, H., Upgrading ofVenezuelan Crude Oil—Continuation of the Orinoco Feasi-bility Study within the Venezuelan/German Agreement (An-nex III B), Final Report BMFT—Reference No. 03E-6400B,1989.

[14] Nieder, D., Schneider-Kuhnle, P., and Wolter, D., Introductionof Zinc Injection in NPP Biblis, Unit B, VGB PowerTech, no. 6,pp. 61–63, 1999.

[15] Kefer, V., and Kastner, W., Leckagen bei unterkritis-chen Rohrleitungsrissen—Ausstromraten und ihre Gerausche,BWK, vol. 39, no. 6, pp. 310–315, 1987.

[16] Kefer, V., Kastner, W., Westphal, F., John, H., Reimann, J.,and Friedel, L., Vorhersagegenauigkeit von Modellen furLeckraten aus Rissen in druckfuhrenden Komponenten,DECHEMA—Monographien, Band 111, Fortschritte derSicherheitstechnik II, 1988.

Holger Schmidt is currently a manager in theDepartment of Thermal Hydraulics and Fluid Dy-namics of Siemens AG Power Generation Group(Siemens/KWU) in Erlangen, Germany. He is aMechanical Engineer and received his Ph.D. fromthe University of Darmstadt, Germany. For morethan 13 years he has been involved inexperiments,modeling and developing programs for thermal

hydraulics and stress analysis.


Wolfgang Kastner is currently a director atSiemens AG Power Generation KWU, where heheads the Thermal Hydraulics and Fluid Dynam-ics Department in Erlangen, Germany. For morethan 25 years he has been active in nuclear safety(UFTF and PKL), in research on fossil-fueledboilers (Benson), and on solar-powered steamgeneration. He is also responsible for product de-

velopment in the � eld of � ow-accelerated corrosion (WATHEC and DASY).

Wolfgang Kohler is a Scienti� c Advisor in theThermal Hydraulics Laboratories of Siemens AGPower Generation KWU. He is a Mechanical En-gineer and received his Ph.D. from the TechnicalUniversity in Munich, Germany. For more than25 years he has been involved in experimentaland theoretical work concerning the thermal hy-draulics of nuclear reactors, fossil-fueled boilers,

and solar-powered steam generators.


thepowerindustry’sview ofpast,present,and futuretwo ...figure1 bensontestrig....

Documents