Proceedings

Enhanced, Compact and Ultra-Compact Heat

Exchangers: Science, Engineering and
Technology
Engineering Conferences International Year 2005

High Temperature Compact Heat

Exchangers: Performance of Advanced
Metallic Recuperators for Power Plants
D. Aquaro M. Pieve

This paper is posted at ECI Digital Archives.

http://dc.engconfintl.org/heatexchangerfall2005/25
 Aquaro and Pieve:

Proceedings of Fifth International Conference on Enhanced, Compact and Ultra-Compact Heat Exchangers: Science,
Engineering and Technology, Eds. R.K. Shah, M. Ishizuka, T.M. Rudy, and V.V. Wadekar, Engineering Conferences
International, Hoboken, NJ, USA, September 2005.
CHE2005 – 30

HIGH TEMPERATURE COMPACT HEAT EXCHANGERS:

PERFORMANCE OF ADVANCED METALLIC
RECUPERATORS FOR POWER PLANTS
D. Aquaro1, M. Pieve2

1
Dipartimento di Ingegneria Meccanica, Nucleare e della Produzione - Via Diotisalvi, 2 - 56100 Pisa; aquaro@ing.unipi.it
2
Dipartimento di Energetica “Lorenzo Poggi” - Via Diotisalvi, 2 - 56100 Pisa; m.pieve@ing.unipi.it

ABSTRACT deeply because of the possibility to reach very high thermal

efficiency, up to 46% in simple cycles and 62% in combined
An overview is given of recent high temperature ones (GE Power Systems, 2003, Franco and Casarosa,
compact heat exchangers technology developments, both in 2002).
the thermo-fluid dynamic innovative solutions and in the The main drawback for the application of recuperative
materials. Gas turbine recuperative cycles and microturbine heat exchangers in such cycles lies in the high volumetric
systems are the investigated fields where this kind of heat flow ratio between the streams, due to the high pressure
exchangers is used. Some common technological features ratio, generally between 12 and 30, combined with an
join all of them, namely: new surface geometry design and almost equal massive flow on both sides of the exchanger.
burdensome material requirements. In the following the consequences of such different
A sizing procedure is developed for a plate-fin type and volumetric stream flows will be shown.
an innovative printed circuit heat exchangers to be used in a Microturbines are a relatively new sort of distributed
gas turbine cycle. Starting from this calculation, the most generation technology being used for stationary energy
important difficulties arising from the specific conditions of generation applications. There are no established guidelines
each heat exchange system are pointed out, together with to categorize small gas turbines, it is arbitrarily assumed that
their possible solutions. A comparison among them is machines in the 50–500 kW range are classed as
presented and some suggestions are proposed about microturbines. For simple cycle microturbines, it is difficult
expanding the boundaries of the individual research fields, to get a thermal efficiency much in excess of 20% based on
to get a wider understanding of common critical issues. today’s technology, because of the small pressure ratios,
Finally, the paper focuses on the main current thus an exhaust heat recovery exchanger is mandatory to
challenges to be faced in the near future in order to get a achieve efficiencies values of 30% and higher (McDonald,
competitive development of such systems. 2000).
In developing heat exchangers for microturbine systems
it should be considered that their application has to comply
1. INTRODUCTION with space and layout restrictions that markedly affect even
the conceptual design of the recuperator. In addition, they
High temperature heat exchangers (arbitrarily meaning have to handle the constant temperature differentials across
temperatures above 650°C) are required in many industrial the surface area, and match an operating life of several
systems for process reasons or aiming to attain high thousands hours with hundreds of start-up/stop cycles. It is
efficiency. Latter one is the case of power plants. Among also very important to balance the need for higher
these, the most promising in terms of development chances performance and more durable materials with the need to
are: make them as cost-effective as possible.
- Recuperative gas turbine and combined cycles
- Recuperative microturbine systems
Each one of these systems has to comply with specific 2. HEAT EXCHANGER CURRENT EQUIPMENT
requirements, related to its function.
In recuperative gas turbine cycles the feasibility of high The industrial systems that have been briefly reviewed
temperature heat exchangers should be analysed more can be analysed on the basis of their main technological

239
Published by ECI Digital Archives, 2005 1
 Enhanced, Compact and Ultra-Compact Heat Exchangers: Science, Engineering and Technology, Art. 25 [2005]

issue. Two major research areas are identified. The first one as desired to make a cross flow, counter flow or multipass
is related to the development of heat transfer surface cross-counter flow exchanger. Besides, the passage shape
geometries, the second to the material issue (namely: high may be corrugated or straight, depending on a number of
temperature mechanical properties, creep properties, and factors such as the fluid used, the heat load and the pressure
environmental resistance). drop requirements. After bonding, several blocks can be
Starting from gas turbine cycles, largely different flow welded together to provide flow capacity to any duty.
areas on the sides need to be accommodated. A numerical Headers and nozzles are welded directly onto the final core
example is reported below to better explain such a problem. block. Figure 1 illustrates a typical PCHE core configured
A reference gas turbine combined cycle is considered, for a counter-current flow.
whose typical data are listed in Table 1. A thermo-fluid
dynamic sizing calculation has been performed, showing the
aforementioned trickiness of arranging the two flow
streams. Two compact heat exchangers have been selected
for this purpose, a printed circuit heat exchanger (PCHE)
and a plate-fin (PFHE).

Tab. 1 - Typical reference data for a heat recovery system in

a combined gas turbine cycle

Streams mass flow (kg/s) 100

Lower cycle pressure (Pa) 1.08·105
Streams pressure ratio 24.3
Inlet air temperature (°C) 521
Inlet gas temperature (°C) 825
Maximum ∆p/p allowable (%) 5 - 5 (gas - air)
Thermal duty (W) 31·106 Fig. 1 - Printed circuit heat exchanger core configured for a
Required effectiveness, min. 0.91 counter-flow arrangement

In this sizing problem a criterion is followed that is the

2.1. Sizing high temperature gas-gas heat exchangers minimization of the overall frontal area. In a PCHE frame it
depends, among other things, on the ratio between the two
A “sizing” problem means the design of a heat stream diameters: the greater the latter, the smaller the
exchanger, selecting an exchanger type, its flow former. So, the highest ratio between the channels diameters
arrangement, the materials and the physical size, in order to should be chosen to minimize the frontal area. Moreover,
meet the specified requirements on heat transfer and the core flow length decreases with the diameters, once their
pressure drops. When the first three are selected, the ratio is fixed. But some constraints have to be set, a
problem reduces to the determination of the physical size, minimum flow length, equal to 0,8 m, to get a true counter-
e.g. length, width, height and surface areas on each side. In current flow and a maximum value for the diameters ratio,
order to get a very high value of efficiency only the counter- which here is 2,5. Generally speaking, geometry with
current flow configuration and the multipass one should be different diameters on the two sides is not preferable, for
considered. The single-pass crossflow configuration is not a manufacturing reasons, so their ratio should be kept closest
valid choice, since the heat exchanger would work in such a to one. (Gezelius, 2004).
way that a very large part of it doesn’t actually contribute to The f Fanning factor and the Colburn j factor to be used
the overall thermal exchange, because of the small in the heat transfer and pressure drops calculation are
difference between the stream temperatures. applied, assuming that the streams flow within straight
The PCHE is a rather novel heat exchanger type, smooth channels.
formed by diffusion bonding of a stack of plates with fluid The sizing procedure is based on the core mass velocity
passages photo-chemically etched on one side of each plate equation (Kays and London, 1984). It couples the heat duty
by using a technique derived from that employed for and the pressure drops, enabling the computation of the first
electronic printed circuit boards - hence the name. The approximate value of the fluid mass velocity, G. As a result,
diffusion bonding process allows getting an interface-free the iterative solution to the problem converges relatively fast
join between the plates, giving it the base material strength with this estimated value of G.
and very high pressure containment capability. The fluid For duties involving high gas flow rates with pressure
passages have approximately a semicircular section, being drop constraints it’s quite logical to expect a short and fat
typically 0.5 to 2 mm wide (but larger diameters are shape. This is the present case: the sizing calculation of a
available, up to 5 mm). There are different etching patterns, PCHE core gives the results summarized in Table 2.

240
http://dc.engconfintl.org/heatexchangerfall2005/25 2
 Aquaro and Pieve:

Tab. 2 - Thermo-fluid dynamic sizing results for a counter- taken in/out in at least partial crossflow (Hesselgreaves,
current flow Printed Circuit Heat Exchanger and a Plate 2001).
Fin Heat Exchanger core All of the listed drawbacks arise from the high
volumetric flow ratio between the two streams. It yields a
Counter-current flow configuration too much burdensome constraint on the gas side pressure
PCHE PFHE drop, due to its low pressure, compared to the air one.
Gas channel D (m) 4·10-3
Air channel D (m) 1.6·10-3 2.2. Feasible solutions
Pitch/Diameter ratio 1.3
Plate Thick./Diameter 0.75 A two-pass cross-counter flow model with two identical
Gas side fin surface Plain 5.3 modules would be a valid alternative, for two main reasons:
Air side fin surface Strip 1/9 – 24.12R a - The single cross-flow modules have an efficiency
Core cross section (m2) 29.2 10.3 value close to the maximum practically allowable
Stream flow length (m) 0.8 2.5 for such a configuration.
Effectiveness 0.91 - Their inlet and outlet headers can be made and
Outlet T (°C) gas/air 570/798 installed as in most heat exchangers.
Air side h (W/m2K) 271.3 1061 A sizing calculation for a two-modules core, both
Gas side h (W/m2K) 106.7 98.6 PFHE and PCHE, leads to the results shown in Table 3, with
Overall U (W/m2K) 76.6 89.2 reference to the single module. The surfaces are the same
Gas pressure drop (%) 5.00 3.88 previously used.
Air pressure drop (%) 0.123 5.00
Compactness (m2/m3) 941.7 876.2 Tab. 3 - Thermo-fluid dynamic sizing results for a PFHE
and a PCHE core in cross-counter flow
As one could expect, a very large frontal area to comply
Cross-Countercurrent configuration,
with the required gas pressure drop characterizes the
2 identical modules
resulting overall core shape. A practical PCHE counter-flow
PCHE PFHE
configuration can be obtained only by allowing a crossflow
Gas flow length (m) 0.45 1.89
heat exchange zone near to the exchanger ends, as Figure 1
shows. So, the width should be kept one half the flow length Air flow length (m) 2.00 1.75
at most. In this case, it would mean to get a (73 x 0.4 x 0.8 m) Core height (m) 17.05 6.66
heat exchanger. Consequently, a true counter-current flow is Single module effectiveness 0.789
not feasible. Overall effectiveness 0.91
Regarding the PFHE model, a variety of methods are Outlet T (°C) gas/air 570/798
described in the literature for plate-fin surfaces performance Air side h (W/m2K) b 739.5 931.3
comparison (Shah, 2003). Among these, the Surface Flow Gas side h (W/m2K) b 103.7 87.5
Area Goodness Factor, defined as j/f, is used to get the Overall U (W/m2K) b 91.0 80.9
minimum frontal area for the exchanger. According to this Gas pressure drop (%) 5.00 5.00
principle, the Plain 5.3 surface is adopted for the gas side, as Air pressure drop (%) 5.00 5.00
it’s the critical one. A modified air side surface is selected,
by increasing the fin thickness, to allow the withstanding of As is obvious, the resulting overall volume increases
the fluid pressure at the maximum operating temperature. compared to the counter-current one. But the proportional
By this way a quite cheap material as Alloy 800H could be rise is not the same for the two heat exchangers. It’s about
used, its cost being only 2.5 times the AISI 347, while the 31.4 percent for the PCHE model and 71 percent for the
others heat resistant alloys are 4÷5 times more expensive. PFHE. This could be explained by looking at the air side
The f and j factors are expressed as polynomial functions of heat transfer coefficient. In the cross-current PCHE its value
Re, whose coefficient were deduced from a fit of available almost triples, as the fluid conditions are changed, from
literature data (Kays and London, 1984). laminar to turbulent, due to the reduction in the free-flow
A PFHE type would be possible, if carefully designed area.
distribution headers are assembled that could guarantee
against excessive flow maldistribution and avoid 2.3. Microturbine systems
unacceptable pressure drops. Usually, external header tanks
with nozzles welded into them are applied to the exchanger. In microturbine systems the two parameters that have
Internal distributors are required, since one stream must be potential for efficiency advancement are increased values of
turbine inlet temperature and higher recuperator
a
The fin thickness is increased by a factor 1.5 to withstand the
b
fluid pressure at the maximum temperature relating to the coldest module
241
Published by ECI Digital Archives, 2005 3
 Enhanced, Compact and Ultra-Compact Heat Exchangers: Science, Engineering and Technology, Art. 25 [2005]

effectiveness. Both of them have a strong impact on the strains, leading to individual failures that, in turn, connect
recuperator. A higher turbine inlet temperature means that and propagate from the brazed joint into the parent material
the recuperator hot gas inlet temperature increases, this of the plate. This cell failure results in leakage of the
necessitating the use of higher cost superalloys for a pressurised air into the exhaust gas stream, reducing the
component that is already the most expensive in the system. turbine power and increasing the fuel consumption.
Recuperator technology is not new - recuperators are
used in larger industrial gas turbines - but it is also not Fig. 2 - Compact heat exchangers surfaces:
perfected, particularly at this new size level. One of the a) Primary surface sheets
chief challenges in designing recuperators is to obtain a b) Brazed Plate-Fin Recuperator
surface area that is both large enough to absorb the heat and
able to absorb the thermal shock due to temperature change.
This is particularly difficult to obtain given the space
restrictions of a microturbine. Recuperators also operate in a
hostile environment, undergoing thermal stress as they
exchange heat varying from room temperature to
approximately 600-650°C.
Differential pressure ratios inside the recuperator also
contribute to weaken it: compressed air and high-
temperature exhaust gases push against the inner surfaces of
the recuperator from opposite directions, deforming the
metal components. These effects combine to make the
recuperator susceptible to heat leakage and to creep.
In manufacturing microturbine recuperators two main
design have entered into use: primary surface design (PSR)
and plate-fin design (PFHE).
The first one attempts to mitigate the heat exchange
stress by maintaining a constant surface area. This is
achieved through alternating layers of thin (less than 0.1 mm
thick) metal sheets folded into a corrugated pattern, which
provides the requisite surface area without the large number
of joints (Figure 2a). Pairs of these sheets are welded
together around the perimeter to form air cells, which are
then pressure checked before being welded into the
recuperator core assembly. There are no internal welds or
joints within the air cell. Heat flows through the corrugated
layers such that hot air proceeds in a cross-counter flow
through the recuperator. PSR design is inherently resistant to cyclic thermal
The PSR layouts are divided into annular recuperators, fatigue because it can flex to relieve thermal and mechanical
which are wrapped around the turbine, and rectangular strains. Its construction allows the individual sheets to move
recuperators installed behind the rotating machinery. Each relative to each other and relieve the thermal strain. An
of them may be attractive for different applications. additional benefit is the intrinsic natural damping
The Plate Fin heat exchangers use thin sheets of metal characteristics derived from stacking cells together with
lined up parallel to one another, like the gills of a fish. This multiple friction interfaces. These several contact points
provides a large surface area, but there are corresponding absorb displacements from vibration sources (Ward, 1995).
disadvantages. The most important one is its minor
reliability, compared to a PSR.
The brazed plate-fin recuperator is a rigid structure. As 3. MATERIALS ISSUES
the size of the plate-fin recuperator increases, the structure
becomes less capable of handling thermal gradients resulting All the previously investigated high temperature heat
from sudden changes in operating conditions, including cold transfer systems have in common high materials
starts and shutdowns. The continuous brazed joints between performance requirements.
the two plates and air fins and between adjacent plates and Currently, for the majority of gas turbines operating at
exhaust gas fins, lock the structure firmly together (Figure temperature lower than 650°C, PSR and PFHE are made of
2b). It is too rigid to accommodate the temperature gradient thin sheet of Type 347 stainless steel (18Cr-10Ni-1Nb),
between the sides from cell to cell and across the entire which provides excellent oxidation resistance. Its
recuperator structure. performance up to this temperature is derived from the
The brazed joints distort to comply with high thermal presence of a thin, adherent chromium oxide (Cr2O3) film,
242
http://dc.engconfintl.org/heatexchangerfall2005/25 4
 Aquaro and Pieve:

which forms in service at a predictable rate. It inhibits Tab. 4 - Composition of some heat resistant austenitic
further transport of oxygen to the underlying material, thus stainless alloys processed into foils (Maziasz and
avoiding excessive oxidation damage. At higher Swindeman, 2003)
temperature, namely above 700°C, film instabilities lead to
high oxidation rates and possibly cracking with spallation 347 All. All. HR HR HR All. 602
and Cr depletion, which could breach the protective layer SS 803 740 120 230 214 625 CA
and accelerate oxidation (Ward, 1995). For that reason, at Fe 68.7 40 2 33 3* 3 3.2 9.5
temperatures higher than 650°C or where severe corrosion Cr 18.3 25 24 25 22 16 22.2 25
may be anticipated, alternative materials with improved Ni 11.2 35 48 32.3 52.7 76.5 61.2 63
properties meeting these specific needs are to be used. Mo 0.3 n.a. 0.5 2.5 * 2 9.1
Before an alloy can be adopted for the fabrication of a Nb 0.64 n.a. 2 0.7 3.6
PSR or a PFHE surface, several questions must be
C 0.03 0.05 0.1 0.05 0.1 0.02 0.18
answered. First, the alloy must be capable of being rolled
Si 0.6 n.a. 0.5 0.6 0.4 0.2
into thin (<0.1 mm) foils that have adequate uniform
elongation so that they can be formed into the required Ti 10-3 n.a. 2 0.1 0.23 0.15
corrugated pattern. The foil-forming process must lend itself Al 3·10-3 n.a. 0.8 0.1 0.3 4.5 0.16 2.2
to continuous operation in order to minimize cost. Second, 3 Co, 5 Co*,
0.2 20
the fine-grained alloy foil must possess the necessary creep Oth. n.a. 3 W, 14 W, Y
Co Co
strength at the maximum operating temperature. The creep 0.2 N La
strength is only one measure of lifetime of the alloy; the * means maximum allowable element content
other is its resistance to oxidation and other corrosive
species present in the exhaust gas. Third, the final product
must be the most cost-effective combination of the two
previous properties (Omatete et al., 2000).
Intensive research activity has been carried out during
last ten years, especially by Oak Ridge National Laboratory
(ORNL), that allows getting a wide overview of the
candidate alloys for meeting the performance requirements
(Stinton and Raschke, 2003).
Type 347 stainless steel has been shown capable of
being modified to operate up to 750°C with creep resistance
superior to the standard commercial material, at a cost less
than twice the standard 347 SS. Nevertheless, at
temperatures above 700°C, corrosion resistance and the
effects of moisture remain the dominant limiting factors for
Type 347 steels. A breakaway oxidation in water vapour
could still happen, but the delay in such attack would
provide a substantial added benefit of these modified alloys
for recuperator applications (Pint et al., 2001, Pint and
Peraldi, 2003).
Coming to better, and also more expensive alloys the
following two groups could be identified, with Alloy 740
representing a sort of dividing line:
- Alloys comparable with 347 stainless steel:
modified Alloy 803, Alloy 602 CA, Haynes HR120
and Haynes HR230.
- Alloy considerably stronger than 347 SS: Alloy
625, Haynes HR214, Hastelloy X and Plansee alloy
PM2000 (ODS alloy).
Table 4 shows their nominal composition, while Figures
3a and 3b illustrate the creep strains versus time for many of
the alloys. The creep data refer to creep-rupture tests in air
at 750°C and 100 MPa. This aggressive creep testing was
chosen as an accelerated screening condition for advanced Fig. 3 - Creep strain versus time for some foils of heat
alloys, as most commercial recuperators made from standard resistant alloys in air at 750°C and 100 MPa. a) alloys
Type 347 SS operate at low stress below 700°C. comparable with Type 347 SS; b) Stronger alloys (Maziasz
and Swindeman, 2003, Pint et al., 2001)
243
Published by ECI Digital Archives, 2005 5
 Enhanced, Compact and Ultra-Compact Heat Exchangers: Science, Engineering and Technology, Art. 25 [2005]

It should be pointed out that the results apply to foil gained in less than 4,000 hours, whereas the thinner (50 µm)
creep, where foil is a few tenths of millimetre thickness Al2O3-forming foils show minimal weight gained after
sheet. At such a reduced thickness, a maximum value of 10 12,000 hours. This is due to the limited reservoir of Cr in
µm for the grains diameter should be expected, compared to those stainless steel foils and the fast Cr consumption
50-100 µm for typical tubular and plate products (Maziasz (oxidation) at 900°C compared with the rate of formation of
et al., 1999). As the creep resistance is dependent on the Al2O3 on the alloys that contain aluminium (Pint et al.,
grain size, (a greater one providing a higher strength) a 2001).
different behaviour will be obtained for the same materials
depending on their form. That’s the reason why some alloys,
for example HR230, show a worse creep resistance than
others that usually are weaker.
It’s clear that there is a wide range of more heat
resistant and corrosion resistant alloys and superalloys
commercially available, but many of them cost significantly
more than Type 347 stainless steel. That could be seen in
Table 5, where some relative costs are compared setting the
347 SS as a baseline.
From these limited data, Alloy 625 is one of the most
cost effective improvements relative to 347 steel, but at a
substantially greater initial cost. The same is true for
HR214, which has a longer rupture life and a much higher
cost too. By contrast, alloys HR230 and 602CA cost about 7
and 9 times more than Type 347 steel respectively, and
HR230 lasts only 1.6 times longer to the 10% strain limit,
while 602CA shows less creep resistance than Type 347
steel. These alloys are not cost effective replacements for
347 steel, simply based of their creep behaviour.
Fig. 4 - Total mass gain during 500-hour cycles at 900°C in
Tab. 5 - Relative costs for some alloys using Type 347 laboratory air for Cr2O3 and Al2O3 forming specimens (Pint
stainless steel as a baseline (Maziasz and Swindeman., et al., 2001)
2003, Pint et al., 2001)
This indicates that alloys that are protected by Cr2O3
347 Alloy HR Alloy HR Alloy HR ODS formation can be used roughly up to 800°C; beyond that,
SS 803 120 625 230 740 214 PM2000 only alloys protected by Al2O3 formation will survive.
1 3 3.5 5 7 9 9 10 In many power plant combustion systems the
combination of high temperature and aggressive
The stainless alloys with 25%Cr and 35%Ni, modified environment, which may contain oxidizing, sulphidizing and
Alloy 803 and HR120, have about twice the creep resistance chlorinating elements, obliges to give up the typical solution
of 347 steel, and cost about 3-3.5 times greater. Therefore, strengthened alloys, either Fe or Ni based, starting from
both of them would represent possible performance 750-800°C. Compared to such environments, up to now two
upgrades relative to Type 347 steel for using to about 750°C ways only have been followed: ODS alloys and ceramics.
or slightly more, while Alloy 625 and HR214 would be the ODS means Oxide Dispersion Strengthened alloy,
very high performance alternative up to 800°C and possibly referring to the distribution of fine particles of an inert oxide
higher. throughout the alloy matrix and the development of a
microstructure in which large grains are developed.
3.1. Environmental resistance performances Typically, ODS alloys contain 0.5 % of an oxide, such as
Y2O3 or ThO2, dispersed through the material. Table 6
As previously said, a fundamental requirement for the shows the nominal compositions of some ferritic ODS
materials is the oxidation and corrosion resistance at high alloys.
temperatures. The oxidation of alloys protected by the The reason for the interest in ODS alloys is summarized
formation of a chromium oxide (Cr2O3) film (such as SS in Figure 5, which compares the average 100000 hours
347, 20/25/Nb, 253 MA and HR230) has been compared stress rupture strength for all available wrought or cast high-
with alloys protected by the formation of an aluminium temperature alloys with the theoretical maximum rupture
oxide (Al2O3) such as HR214 and ODS alloy PM2000. strength that could be obtained if the strengthening
Figure 4 shows the total mass gain in foils of alloys mechanisms used were fully optimised. The typical
heated at 900°C in laboratory air for greater than 10,000 longitudinal stress rupture strength for ODS-FeCrAl alloys,
hours. The Cr2O3-forming foils show excessive weight also shown on the diagram, illustrates their strength
244
http://dc.engconfintl.org/heatexchangerfall2005/25 6
 Aquaro and Pieve:

advantage. Essentially, ODS alloys have potential for use to cost and a high directionality.
temperatures where otherwise ceramic materials would have ODS tubes have been developed for the US HiPPS heat
to be considered (Wright et al., 2003). exchanger with MA754, a Ni-Cr based alloy, whereas the
European EFCC program heat exchanger is planned to be
Tab. 6 - Composition of some ferritic ODS alloys (Wright et built with ceramic tubes, hence the difference between the
al., 2003) two projects in the maximum temperature achievable
through the heat exchanger. Unfortunately, up to now no
ALLOY Fe Al Cr Mo Ti Dispersoid laminar products have been manufactured with ODS alloys,
ODS-Fe3Al Bal. 15.9 2.2 0.07 Y2O3 so their application in the field of the extended surfaces is
unlikely to take place in the near future.
MA956H Bal. 5.77 21.66 0.4 Y2O3
MA956 Bal. 4.46 19.64 0.39 Y2O3
PM2000 Bal. 5.5 20 0.5 Y2O3 4. CONCLUSIONS
ODM751 Bal. 4.5 16.5 1.5 0.6 Y2O3
A general review of high temperature compact heat
exchangers has been performed. They are required in many
industrial processes for inherent reasons or in order to
improve the efficiency. This is particularly true in the case
of the power production plants.
A representative group of high temperature systems has
been critically analysed and some common technological
issues have been pointed out.
The difficulty of managing highly different massive
flows in high-pressure gas turbine recuperators has been
shown through a numerical example, suggesting some
possible solutions.
As regards the material issue three reference ranges
have been identified, depending on the temperature.
In order, one could consider current developments in
microturbine recuperator materials. A modified version of
Alloy 803 and HR120 alloy have been found the most
suitable from a cost effective standpoint up to about 750°C.
In succession, the nickel-based superalloys are the most
promising up to 800-850°C and also above, if a properly
aimed design of microstructure and mechanical properties is
planned. Finally, the more suitable classes of materials for
Fig. 5 - Schematic comparison of the current average and using in the high temperature indirect firing have been
theoretical maximum creep strength of some high- identified: ODS alloys and ceramic materials. Many
temperature alloys (Wright et al., 2003) differences could be highlighted between latter two, for
example the maximum allowable temperature, which for
The dispersed oxide particles help to pin grain ceramics may be up to 200-250°C higher than ODS alloy. It
boundaries reducing their ability to creep at high should be considered that, when properly applied, ODS-
temperature and stabilize oxide scales. In fact, they are FeCrAl alloys could be usable to temperatures up to 1150°C
much more stable than the typical strengthening particles in components at the lower end of the temperature range
such as γ' and carbides, present in conventional high where ceramics typically are thought to be required.
temperature alloys and the oxide scales possess better As concerns the main current challenges to be faced in
spallation resistance (Quadakkers et al., 1989). the near future, the followings should be recalled:
It is expected that ODS alloys in service will be - Even if it’s currently possible to find suitable
subjected to mechanical loading such that their rate of creep metallic materials on the entire temperature range
would be quite slow, therefore their service life is likely to between 700°C and 1150°C, each one of them has
be determined by the oxidation behaviour. Currently, it to comply with manufacturing limitations. For
limits the maximum use temperature at 1050-1150°C, instance, it is not easy to process ODS in foils, and
depending on the specific environment aggressive it’s still quite difficult to join them.
constituents (Wright et al., 2003). - Many alloys are at a developmental stage yet,
The ODS microstructure requires a powder which causes their projected cost to be very high.
metallurgical production process to be used and also poses It’s essential that such items be deeply investigated, in
problems in joining these alloys. Further drawbacks are the order to get a successful development of high temperature
245
Published by ECI Digital Archives, 2005 7
 Enhanced, Compact and Ultra-Compact Heat Exchangers: Science, Engineering and Technology, Art. 25 [2005]

compact heat exchange systems. In this perspective, a Shah, R.K, Sekulic, D.P., 2003: “Fundamentals of heat
stronger connection between the investigations regarding exchanger design”, John Wiley and Sons, 2003.
seemingly different fields would be probably very Maziasz P.J., Swindeman R.W., 2003: “Selecting and
profitable. Developing Advanced Alloys for Creep-Resistance for
Microturbine Recuperator Applications”, ASME Journal of
Engineering for Gas Turbines and Power, pp. 310-315,
NOMENCLATURE AND ACRONYMS 2003.
Omatete O.O., Maziasz P.J., Pint B.A., Stinton D.P.,
D Tube or channel diameter, m; 2000: “Assessment of recuperator materials for
f Fanning friction factor, dimensionless; microturbines”, Report No. ORNL/TM-2000/304, Oak Ridge
G Fluid mass velocity, kg/m2s National Laboratory, Oak Ridge, TN, December 2000.
h, U Heat transfer coefficient, W/m2K; Stinton D.P., Raschke R.A., 2003: DER Materials
j Colburn factor, dimensionless; Quarterly Progress Report, June-September 2003, DER
∆p/p Pressure drop, %, dimensionless; Materials Research, Oak Ridge National Laboratory, Oak
GE General Electric Company Ridge, TN, 2003.
HiPPS High Performance Power System Pint B.A., Peraldi R., 2003: “Factors affecting corrosion
ODS Oxide Dispersion Strengthened (alloy) resistance and recuperator alloys”, ASME Paper 2003-GT-
PCHE Printed Circuit Heat Exchanger 38692, Am. Soc. Mech. Engin., New York, NY, 2003.
PFHE Plate-fin Heat Exchanger Pint B.A., Swindeman R.W., More K.L., Tortorelli P.F.,
PSR Primary Surface Recuperator. 2001: “Materials selection for high temperature (750-100°C)
metallic recuperators for improved efficiency
microturbines”, ASME Paper 2001-GT-0445, Am. Soc.
REFERENCES Mech. Engin., New York, NY, 2001.
Franco A., Casarosa C., 2002: “Thermoeconomic
Maziasz P.J., Swindeman R.W., Montague J.P., evaluation of the feasibility of highly efficient combined
Fitzpatrick M., Browning P.F., Grubb J.F., Klug R.C., 1999: cycle for power plants”, Proceedings of ECOS 2002, Berlin,
“Improved creep resistance of austenitic stainless steel for 2002, Vol 2, p. 801-812.
compact gas turbine recuperators”, Materials at High Ward M.E., 1995: “Primary surface recuperator
Temperatures 16(4), pp. 207-212, 1999. durability and applications”, Turbomachinery Technology
McDonald C.F., 2000: “Low-cost compact primary Seminar paper TTS006/395, Solar Turbines, Inc., San
surface recuperator concept for microturbines”, Applied Diego, CA., 1995.
Thermal Engineering, 20 (2000), p. 471-497. Wright I.G., Pint B.A., McKamey C.G., 2003: “ODS
Quadakkers W.J., Holzbrecher H., Briefs K.G., Beske alloy development”, 17th Annual Fossil Energy Materials
H., 1989: “Differences in Growth Mechanisms of Oxide Conference, Baltimore, Maryland, April 2003.
Scales Formed on ODS and Conventional Wrought Alloys”, GE Power Systems, 2003: “LMS100 Gas Turbine. GE’s
Oxidation of Metals 32 (1,2) 1989, p. 67-88. New Gas Turbine System: Designed to Change the Game in
Hesselgreaves J.E., 2001: Compact heat exchangers. Power Generation”, Brochure GEA13640.
Selection, Design and Operation, Pergamon Press, 2001. Gezelius, K., 2004: “Design of compact intermediate
Kays, W.M., London, A.L., 1984: “Compact heat heat exchangers for Gas Cooled Fast Reactors”, Master
exchangers”, McGraw-Hill, 1984. Thesis, MIT, 2004.

246
http://dc.engconfintl.org/heatexchangerfall2005/25 8