1

THE DEVELOPMENT OF HIGH EFFICIENCY HEAT EXCHANGERS FORHELIUM GAS COOLED REACTORS. (PAPER NO. 3213)

Stephen John Dewson Heatric Division of Meggitt (UK) Ltd.46 Holton RoadHolton HeathPooleDorset BH16 6LTEnglandPhone: +44(0) 1202 627000Fax: +44(0) 1202 632299Email: stephen.dewson@heatric.com

Bernard Thonon CEA (Commissariat à l'Energie Atomique) - GRENOBLE17 Rue Des Martyrs38054 Grenoble Cedex 09Phone: 33 4 38 78 30 79Fax: 33 4 38 78 51 72

1. ABSTRACT.

This paper considers the requirements and development of a suitable heat exchanger for bothrecuperator and intermediate heat exchanger HTR applications.

For power generation there is an expectation for a global net plant efficiency of around 50%.Consequently intermediate heat exchangers and recuperators being considered will require aneffectiveness of 95% or higher.

Thus the heat exchangers will play a key role in projects such as the European Union “HighTemperature Reactor Components and Systems,” and “Rational and Objective for Integrated ProjectHigh Temperature Reactor Nuclear Technologies” and other non European Union Projects, whichcouple a helium cooled reactor to helium turbine plant.

Key issues will include:

• Thermal design.• Hydraulic design• Mechanical design• Containment and safety considerations• Layout and compactness• Cost

In particular we will examine the development of the Printed Circuit Heat Exchanger (PCHE) toaddress these issues. Traditionally the PCHE is a high efficiency, high integrity compact plate typeexchanger used primarily for gas applications on offshore facilities. From this a purpose designedrecuperator and intermediate heat exchanger configuration have been developed, suitable for hightemperature gas cooled reactor applications.

We conclude that the PCHE permits the key issues to be successfully addressed. The latest PCHEdesigns permit a simple, compact, high integrity, low cost arrangement. This high effectivenessexchanger is expected to be substantially cheaper than arrangements hitherto considered.

2

2 AIMS

The aim of this paper is to investigate theheat exchanger requirements for hightemperature gas cooled reactors. Focus isgiven to the principle exchanger the gas /gas exchanger, although many of theconsiderations and conclusions are validfor the coolers and secondary exchangers.

Finally the paper aims to show thatalthough there are several exchanger typesand configurations that appear to meet therequirements for gas cooled reactors, inpractice there is only one. The onlyexchanger that is able to meet the thermal,hydraulic, mechanical, compactness andeconomic requirements of current projects,is the Printed Circuit Heat Exchanger(PCHE). Further it is only the PCHE thathas the potential to be developed to highertemperatures and pressures to meet futurerequirements.

It is not the aim of this paper to makecomparison between GT-MHR andPBMR, or direct cycle and indirect cycle,although from the comments about theheat exchangers some conclusions may bedrawn.

3 INTRODUCTION

If nuclear energy is to provide a viablecost-effective alternate to fossil fuels thenglobal net plant efficiencies in the order of50% must be achieved. Additionally safetyrequirements must be enhanced, and theassociated cost of that safety must bereduced.

At present Light Water Reactors (LWR)using the steam or Rankine cycle achievean efficiency of around 33%. However, theefforts required to assure safe operationare considerable.

The medium temperature indirect gasturbine cycle (MHTGR) offers someincrease in efficiency up to 38%. Themajor weak point is the steam generator.The low heat transfer coefficients ofhelium when compared with water leads tothe need for large steam generation units.

High temperature gas reactors have theunique ability to use the Brayton cycle.With the single phase Brayton cycleefficiencies of 48% can be easily achieved.

3

There are variants on the gas cooledreactor, for instance both GT-MHR andPBMR use the direct gas turbine cycle.The current GT-MHR concept has areactor using conventional fuel and all ofthe rotating equipment associated with thepower conversion system is on a singleshaft. PBMR has a less conventionalpebble bed reactor and the rotatingequipment is split between three separateshafts.

However, except for differences in the sizeand overall duty, the requirements for therecuperator are very similar.

Another variant is the high temperaturegas reactor with an indirect cycle.

The reactor is again cooled with helium,which is now on a separate primary loop.The secondary loop is also a gas loop;usually helium, although other fluids underconsideration include air and nitrogen. Forpower generation, efficiencies in the orderof 50% can again be achieved.

4

For this indirect cycle the recuperator isreplaced by an intermediate heatexchanger and the operating conditions aresomewhat different. Some perceive theseparating of the primary and secondaryside to be safer. However, it is easy to seethat with this separation of primary andsecondary sides it is possible to replace thepower generation system with a hydrogenproduction facility, or any other facilitywhere large quantities of heat are required.

Whether GT-MHR or PBMR, direct cycleor indirect cycle all of the hightemperature gas cooled reactors have onething in common. If a global net plantefficiency of around 50% is to be achievedthere is a need for high efficient heatexchange. This paper looks at those heatexchange requirements, givingconsideration to the following aspects:

• Thermal• Hydraulic• Mechanical• Containment and Safety

Considerations• Space considerations• Fouling• Reliability• Availability• Economics• Future developments and

needs.

4 TYPICAL SPECIFICATIONS

Before starting to define the requirementsof a suitable heat exchanger it is necessaryto better understand some of the operatingparameters. The table below summarisesthe recuperator specifications for twodirect cycle exchangers, the PBMR andthe GT-MHR Projects. In addition twotypical specifications for indirect cyclesare given.

Parameter Units Symbol PBMR GT-MHR Indirect 1 Indirect 2

Thermal Duty MW Q 360 635 600 200Hot Side Fluid - Helium Helium Helium HeliumHot Side Flow Rate kg/s MH 198.0 316.0 277.2 63.2Hot Side Inlet Temperature °C T1 494.0 510.0 787.8 950Hot Side Outlet Temperature °C T2 143.8 125.0 371.1 340Hot Side Inlet Pressure MPa P1 3.00 2.63 6.21 4.0Cold Side Fluid - Helium Helium Air HeliumCold Side Flow Rate kg/s MC 183.0 313.0 1061.7 63.2Cold Side Inlet Temperature °C T3 102.0 105.0 243.3 290Cold Side Outlet Temperature °C T4 481.4 490.0 760.0 900Cold Side Inlet Pressure MPa P2 9.00 7.07 2.07 3.0LMTD °C 24.3 20.0 65.5 50.0Thermal Effectiveness1 % E 97 95 95 92Pressure Drop2 % dP 1.0 1.6 4.0 2.9

Note:

1 Thermal effectiveness is defined as:

E = (T4 – T3) / (T1 – T3)

2 Pressure Drop is defined as:

dP = [(DP1 / P1) + (DP2 / P2)] * 100

where DP equals inlet pressure minus outlet pressure.

5

5 DESIGN CONSIDERATIONS

5.1 Size and Containment.

The thermal duty of all the exchangersspecified above are relatively large. Thesmallest is 200MW. This implies a largeexchanger, especially whenconsideration is given to the thermaleffectiveness (typically 95%) and therelatively low pressure drops. All of thedirect cycle projects require therecuperator to be fitted inside thecontainment vessel. For PBMR therecuperator is to be installedimmediately underneath the powergeneration turbine. For GT-MHR therecuperator is to be installed around thesingle rotating equipment shaft. Space isat a premium so the direct cyclerecuperator must be a compact design.Mechanical, economic and safetyconsiderations dictate that thecontainment vessel should be as smallas possible.

For the indirect cycle safetyconsiderations will probably dictate thatthe intermediate heat exchanger is againinside the containment vessel. Reasonsfor this include:

• Contaminant in the form of carbonparticulate, picked up by theprimary side helium as it flowsthrough the reactor.

• Mechanical considerationsdiscussed in section 5.2.

Without too much in-depth analysis, theneed for a small compact exchanger hasbeen established. Clearly, conventionalshell and tube technology will not besuitable for either the direct cyclerecuperator or the indirect cycleintermediate heat exchanger. Aconventional U tube 10 MW HTRhelium intermediate heat exchanger wasbuilt in Germany in the 1980s.1Although satisfying the specifiedparameters including an operatingtemperature in excess of 900°C this10MW exchanger weighed some 135tonnes. Given current requirements for a200 to 650MW exchanger, the weightalone demonstrates that shell and tubetechnology is not a viable option.

5.2 Material & MechanicalConsiderations

a Direct Cycle

With the direct cycle it is possible todesign the recuperator assuming adifferential pressure design. For bothPBMR and GT-MHR the containmentvessel atmosphere is helium under thehot stream outlet conditions. Hence therecuperator needs to be designed with acold side design pressure of:

(P2 – P1) * safety margin

Assuming the safety margin is about10%, a cold side design pressure in therange 5.0MPa to 6.6MPa is anticipated.On the hot side the design pressureneeds to be little more than the streampressure drop.

Based on normal operation designtemperatures on both the hot and coldsides will be in the range of 550°C.Consideration must also be given to nondesign conditions, such as totalunexpected loss of load on the powergenerator. Brief temperature excursionsperhaps as high as 800°C may occur,under these non-design conditions.Under these transient conditions as thetemperature increases, so the differentialpressure falls. The increasingtemperature and consequent reduction indesign stress is offset by the fall inpressure.

It therefore seems reasonable to assumethat an austenitic stainless steel such asSS316 will satisfy the design criteria forthe direct cycle recuperator. Pressurevessel design codes such as ASME VIIIDivision 1 already allow the use ofaustenitic stainless steels such as SS316to 815°C (1500°F). Although themajority of nuclear codes have notqualified austenitic stainless steelsbeyond the temperature requirements ofLight Water Reactors.

Material and mechanical considerationstherefore restrict exchanger selection tothose types that can be fabricated inaustenitic stainless steel or equivalentand withstand pressures of up to 6.6MPa. This almost certainly prohibits the

6

use of those plate type exchangerswhere support is only provided at theedge of the plate.

There is a further consideration undertransient conditions where temperaturesof up to 800°C might be expected. Thisprobably prohibits the use of brazedstructures. The brazing process uses lowtemperature melting point alloys to joinsthe different exchanger components.What is the braze integrity at 800°C?There seems to be little information inthe public domain. Brazed productshave been developed for hightemperature aerospace applications.However, there can be little doubt thatthe brazed joint will have strength andcreep characteristics that are inferior tothe parent material properties. A brazedproduct will require specificdevelopment.

Given that the recuperator is expected tohave a 30 year life or longer, is brazinga suitable joining process? Would it notbe prudent to use an all weldedconstruction, or a diffusion bond? Witha diffusion bond there is no braze orfiller material. Further, as the bondedstructure is all parent material, it isreasonable to expect, and can bedemonstrated, that parent metalproperties are achieved. Therefore if theparent metal is suitable for use, thediffusion bonded core is also suitable.

However, the questions about brazeintegrity go unanswered. Brazed plateexchangers in a variety ofconfigurations are widely used in a vastnumber of applications. Theseapplications are almost exclusively atlow temperature, often cryogenic. Thereis little literature or operationalexperience to support their use attemperatures above 200°C.

Obviously there is considerableliterature on the precautions and designconsiderations to achieve good weldintegrity.

b Indirect Cycle

With the in-direct cycle intermediateexchanger, operating pressures aresimilar to or lower than those present inthe direct cycle recuperator. Theexamples specified in section 4 have a

maximum operating pressure of6.2MPa, although others haveconsidered operating pressures as highas 9.0MPa. As the two sides of theexchanger are independent loops the useof differential pressure design is lessappropriate. Therefore design pressuresof up to 10.0MPa can be expected.

Operating temperatures are also higher,with hot side inlet temperatures as highas 950°C being specified.

For this exchanger, material selectionand mechanical design philosophybecome significant factors. As stated inthe previous section, few nuclear codesqualify materials much beyond therequirements of light water reactors.Conventional pressure vessel designcodes such as ASME VIII Division 1allow the use of Incoloy 800HT (UNSN08811) and Alloy HX (UNS N06002)to design temperatures as high as 898°C(1650°F). It can further be expected thatalloys such as Inconel 617 (UNSN06617) could be used to designtemperatures as high as 1000°C.

This does not provide the total solution;typical allowable design stresses at900°C are as follows:

• Incoloy 800HT 6.27MPa²(based on a creep life of 105 hours)

• Alloy HX 8.27MPa²(based on a creep life of 105 hours)

• Inconel 617 17.45MPa(based on a creep life of 104 hours)

Assuming a design life of 30 years isrequired at 900°C normal alloys wouldnot permit a design stress much beyond6MPa. Applying normal design criteriathis restricts design pressures to about50% of this value. To exceed thesetemperature and strength limitationsthere will probably be a need to employoxide dispersion strengthened alloys.

If intermediate heat exchangers are to beof normal metallic construction we canconclude the following, irrespective ofexchanger type:

1 Design temperatures and pressuresmust be restricted such that theallowable design stress attemperature is at least twice thedesign pressure.

7

2 Consideration must also be given todifferential pressure design, if theoperating conditions are not to belimited to more moderatetemperature and pressurecombinations.

As the fluids are independent,precautions will be necessary tomaintain the correct pressure differentialin the event of unscheduled upset.Whereas an upset in the direct cyclegives rise to thermal transientconsiderations, an upset in the indirectcycle will also result in pressuretransient considerations.

What becomes apparent is that given thematerial and allowable stressconsiderations, only an exchanger typewhere all joints have parent (or nearparent) properties should be considered.Brazed structures would almostcertainly not be fit for purpose if properconsideration is given to mechanical anddesign life considerations.

5.3 Thermal and HydraulicConsiderations

For both the direct and indirect cycles,the exchanger specifications require athermal effectiveness of 95% or more;typically between 10 and 20 NTUs.Further, allowable pressure drops arelow. This is particularly true for thedirect cycle recuperator.

Hence, irrespective of exchanger type,the hot and cold fluid contactarrangement will need to bepredominantly counter current. Further,the flow length will almost certainly belimited by pressure drop.

As discussed in section 5.1, non-compact heat exchangers for this type ofduty will be excessively large 135 tonneper 10MW, implying a 8,000 tonnerecuperator for a 600MW (thermal)reactor.

The majority of compact heatexchangers tend to be plate type, orplate variant. Therefore we can concludethat the heat exchanger will be a plate

Alloy 800 HT

0

10

20

30

40

50

60

70

0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950

Temperature (°C)

Max

imum

Des

ign

Pres

sure

(MPa

)

8

type exchanger with counter currentcontact arrangement. However, no platetype exchanger can be truly countercurrent, as the plate must have width aswell as length and the fluid must beintroduced across the full width of theplate. Therefore every plate exchangermust have some cross flow component.

Fluid introduction and the cross flowcomponent can be handled in severalways. With conventional plate fin typeexchangers a distributor is used on oneor both of the streams. Examples aregiven in the table below.

Typically with this type of exchangerthe distributor is a relatively smallproportion of the plate, 20% or less. TheIngersoll-Rand plate fin also uses adistribution system.

Less is known about the Ingersoll Randplate fin exchanger. From drawingsavailable in the public domain thedistributor area is however largecompared to the plate area and activeflow length.

With a distributor it is usual for the findensity to be lower than in the heattransfer area; typically the hydraulicmean diameter in the distributor is 2.5 to3.0 times greater than in the heattransfer area. As a consequence, heattransfer and pressure dropcharacteristics in the distribution systemare low compared with the heat transferarea. However, a flow path going acrossthe plate prior to a change intemperature will have different pressuredrop characteristics to a fluid changingtemperature and then crossing the plate.

This results in flow maldistribution, thesmaller the aspect ratio the greater themaldistribution effect. As the figurebelow indicates, due to the high

9

effectiveness, flow maldistributioncauses significant under performance ofsome passages but there is very limitedopportunity for other passages to overperform. The result is an operationaleffectiveness below that calculated ormodelled. Exchangers such as theIngersoll Rand plate fin, where theeffective width seems to be large incomparison to the flow length mayprove to be very susceptible to thismaldistribution mechanism.

The normal geometry diffusion bondedPCHE, offers an alternative todistributors.

The active passage geometry is used inthe cross flow region and there is nodistributor. Effectively the exchangerhas a counter current contactarrangement with cross flow entry andexit regions. All of the plate is active,but the efficiency of the cross flowregions is small, for high effectivenessduties. Further, there is flowmaldistribution similar to that foundwith the plate fin and their distributors.Increasing the plate aspect ratio andchanging the passage geometry indifferent parts of the plate can reducebut not eliminate this phenomenon.

One significant advantage of having nodistributors is reduced headermaldistribution. As PCHE headers areoften longer than with other exchangertypes, it is often wrongly assumed thatheader maldistribution is greater. In factplate type exchangers with distributorshave significantly greater headermaldistribution, even though they haveshorter headers.

If correctly analysed, the following canbe concluded about header pressurevariations and potential maldistributionmechanisms:

• There will be a pressure profilealong the header due to the dynamichead effect (the difference betweentotal and static pressure). Near thenozzle the dynamic head is large;remote from the nozzle the dynamichead tends to zero. This effect isindependent of length. For anygiven header diameter, thedifference in inlet (static) pressurebetween the first and last plate isthe same irrespective of the numberof plates between.

• Any gravitational effect isnegligible.

• Frictional losses are present, but,they are also negligible; at least anorder of magnitude smaller than thedynamic head effect.

The only significant component is thedynamic head within the header. Forthose exchangers that have nodistributors, such as PCHEs, the headerdiameter is at least equivalent to theexchanger effective width. For thoseexchangers with distributors, such asplate fins, the header diameter is justsufficient to cover the distributor, whosewidth is a fraction of the effectivewidth. The dynamic head is proportionalto the square of the header diameter.Therefore, assuming a distributor whoseeffective opening is 50% of the plateeffective width, a plate fin headerdynamic head would be 4 times greaterthan that found in the equivalent PCHE.The potential maldistributionmechanism will equally be 4 timeslarger.

10

To conclude, the normal geometryPCHE has some advantage over platefins in that header maldistribution issignificantly smaller in a PCHE. Thetraditional PCHE and the plate fin areboth subject to plate maldistribution.Further, whether because of distributorsor cross flow region both havesignificant areas (20% of the plate ormore) where heat transfer is not veryeffective. For these reasons and others,Heatric decided to develop a newcontact arrangement and exchangerconfiguration for their product thePCHE.

At this point it is not possible to disclosethe exact internal geometry of the “NewConcept PCHE.” However, the newdesign exhibits the following thermaland hydraulic features:

• All active heat transfer passages arein pure counter currentconfiguration.

• The fluids are distributed to theactive passages via internal ports.Further, the plate active length towidth ratio is greater than 7:1. As aconsequence there is no significantplate maldistribution.

• The herringbone (zig-zag) passagegeometry has been reconfigured toenhance turbulence, and thereforeheat transfer coefficients. This givesa 60% increase in overall U.

• The reconfigured plate and passagegeometry offers reduced pressuredrop characteristics (Up to 55%reduction in overall pressure drop).

5.4 Fouling Considerations

All compact exchangers are to someextent susceptible to fouling. This isinevitable, as they are compact withreduced passage size and increasedpassage tortuosity to enhance heattransfer coefficients and thus reduce theexchanger size.

For this application the service will beessentially clean, particularly whencompared with many of the hydrocarbonapplications where both plate fins andPCHEs are widely employed. The onlyfouling agent will be carbon particulatefrom the reactor. This particulate will beless than 1 micron in size and theanticipated particulate load will be inthe order of 0.1kg /MW at the end of thereactor life.

With a minimum exchanger operatingtemperature in the order of 100°C nomoisture will be present to act as abinding agent. Particle agglomeration istherefore improbable. Therefore theonly question is whether the particlespass through the exchanger.

In the case of the PCHE the answer isyes. The fluid passages areapproximately semicircular with aradius of 0.6mm. It is thereforereasonable to assume that a particlesmaller than one third of the passageradius will pass through the exchanger.Hence we can determine that particulatewith a maximum dimension less than0.2mm (200 micron) will pass throughthe exchanger. A particle some 2 ordersof magnitude greater than the actualgraphite particles would pass throughthe exchanger.

PCHE passages are also continuous.Any passage can be followed from inletto outlet; there are no breaks ordiscontinuities. Hence, there are no deadareas or areas of low velocity and thereis high wall shear throughout theexchanger internals. There is nopossibility of particles settling or ofadhesion to the wall.

With exchangers of plate fin variety theminimum dimension is not the finheight but the distance between fins. Ahigh performance fin would have a findensity of about 850 fins per meter. The

11

foil thickness would be of the order ofabout 0.2mm, hence the clearancebetween fins is about 1mm. Howeverfor serrated fins, the next fin interruptsthis clear way. It is therefore necessaryto divide this distance by two anddeduct a further foil thickness. Theactual clearance is therefore about0.4mm. Again assuming a particlesmaller than one third of the passageminimum dimension will pass throughthe exchanger we find that particlessmaller than 0.13mm (130micron) willpass through the exchanger. Thisclearance is smaller than for the PCHE,but still some 2 orders of magnitudegreater than the actual graphite particles.

To avoid particulate fouling within theexchanger, emphasis needs to be put oncontinuous passages. Exchangers withmultiple fin pads and or serrated finshave a greater potential fouling5;consequently they are probably notsuited to this application.

5.5 Containment and SafetyConsiderations

No heat exchanger is designed to leak.Consideration should however be givento the consequence of a leak. For eitherthe direct or the in-direct cycle, it isprobable that the exchanger, irrespectiveof type will be installed inside thecontainment vessel. As this vessel willalmost certainly have an atmosphere ofeither the hot side outlet or the cold sideinlet, both internal leaks* and externalleaks* will have the same effect; theywill represent a bypass and fall inefficiency. Therefore it is onlyimportant to give consideration to thelikely leak magnitude.

The PCHE has one feature that makes itunique among compact exchangers, itscontinuous passages. Each passage oneach plate can be regarded as a separatepressure vessel. Therefore a leak withina PCHE can be regarded as a leak froma semicircular passage of radius 0.6mm.

For a shell and tube exchanger if thesame criterion is applied, one mustconsider the leak of a ruptured tube.Assuming standard 15mmNB tube, atchoke velocity this leak will be abouttwo orders of magnitude greater than theequivalent PCHE leak.

If a normal brazed plate type exchangeris considered, the passages are non-continuous, due to different fin pads andserrated fin, so consideration must begiven to a leak from a whole plate.Dependent upon plate size and finheight this would again be a leak at leasttwo orders of magnitude greater than theequivalent leak in a PCHE.

The PCHE core having no brazed jointsis less likely to leak than moreconventional exchanger types, which aredependent upon joints of less thanparent metal properties.

* An internal leak is regarded asa leak between side A to Side B. Anexternal leak is regarded as a leak fromone side of the exchanger to theatmosphere of the containment vessel.

5.6 Cost and Economics

It is difficult to obtain realistic prices formany of these exchangers, but estimatedweights are available. For any givenservice and operating parameters, it isalso likely that materials of constructionwill be similar. Conclusions as to pricecan therefore be drawn from thefollowing table.

Exchanger Type Weight / Duty(tonnes / MW)

Shell & Tube1 13.5Brazed Plate Fin6 0.2PCHE 0.2

12

Assuming weight is an accurateindicator of cost, the above shows shelland tube to be more than an order ofmagnitude more expensive. This ignoresthe additional cost associated with largercontainment vessels, supports etc.

Based just on weight there is little tochose between PCHE and brazed platefin. There is a need to look at associatedcosts, pipework costs etc. For instance a360MW PBMR recuperator could besupplied as approximately 48 brazedExchangers or 6 PCHEs, yet both aredesigned to the same specification, to fitin the same containment vessel.Pipework, manifolds, supports andassembly for a configuration of 6exchangers will be considerably lessthan with 48.

One can also conclude that an alldiffusion bonded and welded structuresuch as the PCHE, where parent metalproperties are achieved in every jointwill have a longer design life than anexchanger containing weaker brazedjoints.

6 CONCLUSIONS

1 There are many and varied reasonswhy the heat exchangers for hightemperature gas cooled reactorsneed to be compact. Compactexchangers offer a smaller, safercheaper exchanger, with equal orsuperior thermal and hydraulicperformance.

2 Of all of the compact exchangersthe Printed Circuit Heat Exchanger(PCHE) is superior in nearly everyaspect.

• The PCHE is the only compactexchanger that offers parentmetal strength and propertiesthroughout the entireexchanger. It is not dependenton brazed joints.

• Developments of theherringbone or zig-zag fluidpassage, gives heat transfercoefficients that are equal to orgreater than all other surfaces,including serrated fin.

• The same fluid passage givessubstantially lower pressuredrop characteristics than canbe associated with highperformance fins.

• Heatric’s new contactarrangement offers a platewith pure counter currentflow. Unlike otherarrangements with or withoutdistributors this flowarrangement is virtually freefrom maldistributionmechanisms.

• As the PCHE has noconventional distributorsheader losses andmalistribution mechanisms aresmaller than for other compactexchanger types.

13

• The PCHE’s continuouspassage geometry results in anexchanger that is lesssusceptible to fouling thanother compact types. ThePCHE has larger passageclearance and is free fromdead areas.

• The PCHE’s continuouspassages results in theconsequence of a leak beingtwo orders of magnitude lessthan other exchanger types,compact or non compact.

• The weight of a PCHE is lessthan or equivalent to othercompact exchangers.However, the number ofsections is considerably less.This will result in competitivepurchase prices, together withreduced containment,pipework, manifold, supportand assembly costs.

• The all diffusion bonded andwelded structure will also havea longer life, than a brazedstructure.

3 In the case of the direct cyclerecuperator the PCHE is fullydeveloped and manufacturecould begin immediately. Forthe indirect cycle intermediateheat exchanger, dependentupon design temperature andpressure, some developmentwork is required. However thiswork is on materialdevelopment and designphilosophy not the exchanger.

7 REFERENCES

1 R. EXNER, “HTR PlantComponents” HTR/ECS 2002High temperature ReactorSchool, Cadarache, November2002.

2 ASME II Part D Table 1B3 The Standards of the brazed

Aluminium Plate Fin HeatExchanger Manufacturers’Association.

4 CHILD et al, “United StatesPatent 6,305,079 B1”

5 R. Clarke, “Fouling – A factorwithin control for compact heatexchangers offshore.” GasProcessing AssociationConference, Aberdeen,September 1994.

6 C. Wang et al, “Design of aPower Conversion System foran Indirect Cycle HeliumCooled Pebble Bed ReactorSystem.”