Sei sulla pagina 1di 15

Small recuperated ceramic microturbine demonstrator concept

Colin F. McDonald
, Colin Rodgers
McDonald Thermal Engineering, 1730 Castellana Road, La Jolla, CA 92037, USA
ITC, 3010 N. Arroyo Drive, San Diego, CA 92103, USA
Received 15 October 2006; accepted 5 January 2007
Available online 7 February 2007
It has been about a decade since microturbines rst entered service in the distributed generation market, and the eciencies of these
turbogenerators rated in the 30100 kW power range have remained essentially on the order of 30%. In this time frame the cost of fuel
(natural gas and oil) has increased substantially, and eorts are now underway to increase the eciency of microturbines to 40% or
Various near-term means of achieving this are underway by utilizing established gas turbine technology, but now based on more com-
plex thermodynamic cycles. A longer-term approach of improving eciency is proposed in this paper based on the retention of the basic
recuperated Brayton cycle, but now operating at signicantly higher levels of turbine inlet temperature. However, in small low pressure
ratio recuperated microturbines embodying radial ow turbomachinery this necessitates the use of ceramic components, including the
turbine, recuperator and combustor.
A development approach is proposed to design, fabricate and test a 7.5 kW ceramic microturbine demonstrator concept, which for the
rst time would involve the coupling of a ceramic radial ow turbine, a ceramic combustor, and a compact ceramic xed-boundary high
eectiveness recuperator. In a period of some three years, the major objectives of the proposed small ceramic microturbine R&D eort
would be to establish a technology base involving thermal and stress analysis, design methodology, ceramic component fabrication tech-
niques, and component development, these culminating in the assembly and testing to demonstrate engine structural integrity, and to
verify performance. This would provide a benchmark for more condently advancing to increased size ceramic-based turbogenerators
with the potential for eciencies of over 40%. In addition, the power size of the tested prototype could possibly emerge as a viable prod-
uct, namely as a natural gas-red turbogenerator with the capability of meeting the total energy needs of an average house.
2007 Elsevier Ltd. All rights reserved.
Keywords: Ceramic microturbines; Gas turbine; Turbogenerator; Recuperator; Regenerator
1. Introduction
First generation microturbines that entered service
around 1995 were based on proven technology, conserva-
tive operating parameters, and the use of existing materials.
The motivation for this was to produce turbogenerators
with high reliability, and requiring minimum maintenance,
and this was achieved. Based on the above criteria,
machines rated in the 30100 kW power range operate with
modest levels of eciency on the order of 30%. This,
together with their high cost, and some distributed genera-
tion institutional issues, has likely contributed to the smal-
ler number of units being manufactured annually than had
been projected at the onset of microturbine deployment.
Currently ICEs, burning both gasoline and diesel fuel,
dominate the small gen-set market. Other variants includ-
ing Stirling engines and fuel cells are being developed for
this growing market. Technological and economic aspects
of these approaches are being addressed by other special-
ists. Only microturbines are covered in this paper, where
1359-4311/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +1 858 459 9389; fax: +1 858 459 6629.
E-mail addresses: (C.F. McDonald), (C. Rodgers).
Applied Thermal Engineering 28 (2008) 6074
focus is on increasing the eciency of small low pressure
ratio recuperated turbogenerators by utilizing ceramic
components to facilitate operation at temperatures higher
than can be accommodated by non-internally cooled
metallic radial ow turbines and existing superalloy
Several decades of ceramic development work have been
undertaken particularly in support of vehicular gas tur-
bines rated in the 10s and 100s of kilo Watts. However,
ceramic technologies for the size of the components needed
in these engines (namely the turbine and heat exchanger)
have so far not demonstrated the necessary integrity and
reliability for commercial service in the vehicular or power
generation markets. In taking advantage of existing cera-
mic technology the proposed recuperated ceramic micro-
turbine concept is, by virtue of its small size (i.e.,
7.5 kW), viewed as a reduced cost and lower risk modest
rst step towards the demonstration of a structurally viable
APU auxiliary power unit
AGATA advanced gas turbine for automobiles
CHP combined heat and power
DG distributed generation
hp horsepower
HX heat exchanger
ICE internal combustion engine
ICR intercooled and recuperated
kW kilo Watt
LOM laminated object manufacturing
NTU number of transfer units
ORBC organic rankine bottoming cycle
PMG permanent magnet generator
R compressor pressure ratio
SFC specic fuel consumption
SOFC solid oxide fuel cell
TIT turbine inlet temperature
UAV unmanned aerial vehicle
Fig. 1. Eciency comparison of candidate small power conversion systems.
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 61
ceramic based turbogenerator. The lower risk projection is
based on two factors: (1) the use of a ceramic radial ow
turbine about the same size as those used in turbochargers
that have been in service since the mid 1980s in high perfor-
mance cars in Japan; and (2) the emergence of a compact
high eectiveness ceramic micro-channel recuperator. The
condence gained from the successful demonstration of a
small recuperated ceramic gas turbine would contribute
towards the technologies needed for the eventual deploy-
ment of increased size ceramic based microturbines with
eciencies of 40% and higher.
For the proposed ceramic recuperated microturbine
demonstrator concept, germane technology bases, compo-
nent design considerations, performance data, turbogener-
ator conguration layout, cost considerations, and a
tentative development schedule to accomplish the project
in less than some three years, are addressed in this paper.
2. Small power generation systems overview
In the power range from a few kilo Watts to say over
about 500 kW a variety of power generation systems will
nd niches in the marketplace. It is not the purpose of this
paper to discuss in detail dierences between options since
the emphasis here is on microturbines; however, it was felt
to be meaningful to include an overall comparison of the
type shown on Fig. 1. This plot is by no means comprehen-
sive, and since there are so many variables involved, the
data shown is regarded as tentative, and specialists in the
various elds could well show diering boundaries. Indeed,
performance data for the various power generation systems
shown, and numerous others, have been included in many
published papers and books. In the foreseeable future ICEs
with established manufacturing, user and service infrastruc-
tures will continue to be dominant in the small gen-set eld.
Diesel and gasoline engines have beneted from over a cen-
tury of continuous evolution, and development continues
with emphasis on emissions reduction. Stirling engines
and fuel cells and other hybrid systems are expected to nd
gradual acceptance in power generation and CHP markets.
These prime-movers are not discussed here since they are
viewed as being beyond the scope of this paper. However,
the portrayal of data in the formshown on Fig. 1 is included
here primarily to put into perspective the relative eciencies
of rst generation state-of-the-art metallic microturbines,
and projected future generation advanced ceramic variants,
and this topic is discussed in Section 3.
3. Small gas turbine developments
3.1. State-of-the-art metallic engines
Several thousand microturbines, dominantly natural
gas-red, are currently in service in the 30100 kW power
range. These microturbine turbogenerators have ecien-
cies on the order of 30%. With interest now being expressed
in somewhat larger size turbogenerators, the eciency of
machines rated at 100 kW and above can be increased by
operating with higher turbine inlet temperatures, but still
retaining simple radial ow turbomachinery. Increasing
the compressor pressure ratio and turbine inlet temperature
to values that can be accommodated by a non-internally
cooled metallic radial turbine, and using a superalloy recu-
perator, an upper eciency of about 35% is felt to be a
practical limit for todays fairly low pressure ratio all-
metallic turbogenerators operating at moderately hot day
conditions [1].
3.2. Higher eciency metallic engines
For higher pressure ratio engines the realization of close
to 40% eciency necessitates the use of more complex ther-
modynamic cycles. One approach is to use an ICR cycle.
The engine is more complex, but this approach has the
advantage that the additional metallic components are of
proven technology. Recent studies [24] have shown that
all-metallic ICR microturbines in the 300400 kW power
range have the potential for an eciency close to 40% using
a metallic turbine with an inlet temperature of 1100 C.
Again using proven technology, another approach is to
add a bottoming Rankine cycle power conversion system
[57]. The temperature of the gas turbine exhaust is su-
cient to heat a low boiling point liquid in a heat exchanger
to drive a vapor turbine generator. This is regarded as a
state-of-the-art solution since ORBC technology is well
understood. On the basis of the same eciency, it is not
clear at this stage how the cost of these two state-of-the-
art gen-sets would compare with a more advanced technol-
ogy ceramic microturbine.
3.3. Ceramic gas turbine developments
Prior to initiating the proposed ceramic microturbine
demonstrator development project it is germane to take
advantage of the extensive R&D eorts that have been
undertaken over the last ve decades on ceramic gas tur-
bine components. Two recently published books give very
comprehensive accounts of ceramic engine design and test
experience [8] and component development and character-
ization [9]. Additional valuable data on materials and cera-
mic component development can be gleaned from recent
papers [1015]. We would be remiss if we did not mention
the signicant accomplishments made in the small ceramic
gas turbine eld with details of the engines highlighted on
Table 1, and briey discussed as follows.
The European AGATA program was focused on the
development of three critical ceramic components; a cata-
lytic combustor, a radial turbine, and a xed boundary
recuperator for a 60 kW turbogenerator in a future hybrid
vehicle. After ve years of work the AGATA program
achieved signicant accomplishments in the areas of cera-
mic component design, fabrication and full scale testing.
The feasibilility testing concluded that it was possible to
meet the design specications [16].
62 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
While a larger machine embodying an axial ow turbine,
a 300 kW ceramic engine development project carried out
over a 10 year period in Japan is noteworthy because of
the high eciency demonstrated. The ceramic components
included a can combustor liner, turbine nozzle and axial
ow turbine and turbine scroll. The lower heat exchanger
gas inlet temperature associated with the selection of a
higher compressor pressure ratio allowed the use of a
metallic recuperator. The engine tested, designated CCGT
302 demonstrated an exceptional thermal eciency of
42.1% (without a generator) at a turbine inlet temperature
of 1350 C [17]. If the generator and frequency converter
were included with say a 90% overall electrical eciency,
the thermal eciency would have been 37.4%.
The Japan Automotive Research Institute undertook a
seven year program to develop a 100 kW automobile sin-
gle-shaft ceramic engine with radial ow turbomachinery.
The components included a ceramic can combustor, radial
ow turbine and nozzle, and twin ceramic high eectiveness
rotary regenerators. A very creditable eciency of 36.5%
(non-electrical) was demonstrated with a turbine inlet tem-
perature of 1350 C [18]. As in other small gas turbine pro-
jects the sealing and durability of the ceramic rotary
regenerator presented a major challenge.
4. Small recuperated ceramic microturbine demonstrator
4.1. Genesis
The eciency of very small gas turbines (i.e., in the 10s
and 100s of kilo Watts) is impacted by factors which
include small blade heights, low Reynolds number eects,
tip clearance eects, manufacturing tolerances, surface
nish, and engine-to-engine variations. This was apparent
in a study done by the authors over ve years ago on a
recuperated microturbine rated at 5 kW [19]. Based on
state-of-the-art technologies the projected eciency for this
Table 1
Operated small ceramic gas turbines
Application AGATA CCGT 302 Automotive
gas turbine
Modied J-850 Model
aircraft turbojet
Proposed ceramic
microturbine demonstrator
Project/activity Ceramic component development for
60 kW automotive gas turbine
300 kW
100 kW
Ceramic turbine test Small 7.5 kW recuperated
ceramic turbogenerator
Year 19931998 19881998 19901997 2005 2009
Recuperated Recuperated Regenerated Simple cycle Recuperated
Pressure ratio 4.0 8.0 5.0 3.0 3.0
Turbine inlet
1350 1350 1350 1280 1170
Eciency goal
40 42 40 30
Single shaft Two shaft Single shaft Single shaft Single shaft
Radial Radial Radial Radial Radial
Rotor speed
125,000 48,000 110,000 140,000 160,000
Turbine type Radial Axial Radial Radial Radial
Ceramic Ceramic Ceramic Ceramic Ceramic
Turb tip speed
650 480 665 410 585
Combustor type Catalytic Can Can Annular Can
Ceramic Ceramic Ceramic Metallic Ceramic
Heat exchanger
Platen Platen Regenerator Micro-channel
HX material Ceramic Metallic Ceramic Ceramic
HX eectiveness 0.90 0.82 0.93 0.92
Generator type PMG

Turbine inlet
1350 1350 1350 1280
Power (kW) NA 312 NA
Eciency NA 42.1
36.5 NA
Excludes generator eciency.
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 63
conservatively designed radial ow machine with a turbine
inlet temperature of 900 C was 21.5%.
With the eciencies of the major components (i.e., com-
pressor, turbine and generator) near plateauing after sev-
eral decades of development it was concluded that a
signicant gain in eciency could only be realized by
increasing the turbine inlet temperature. In such a small
machine size there was not an existing technology base,
and this was the motivating factor for proposing a small
ceramic microturbine concept with the power increased
to 7.5 kW.
The authors are encouraged that the proposed essen-
tially ground-up demonstrator approach has already
started in Japan in a very unique way. A small turbojet
engine used in model aircraft was modied to incorporate
a ceramic radial ow turbine [20]. The advantage of start-
ing with such a small demonstrator included fabrication,
facilitation, lower cost, and decreased probability of fail-
ure. The small test engine embodied a 55 mm diameter
radial ow turbine (about the same size as ceramic turbines
used in automobile turbochargers in Japan), and for this
application was fabricated in SN235 ceramic material by
Kyocera [21]. While some problems were encountered a
rotational speed of 140,000 rpm was demonstrated at a tur-
bine inlet temperature of 1280 C [22]. Continuing develop-
ment led to a 1-h test run at a temperature above 1000 C
without severe damage [23], and further duration tests are
planned. It was this innovative approach in Japan to
explore the capability of a small ceramic turbine in a gas
turbine environment that also encouraged the authors to
suggest the proposed demonstrator concept. The proposed
7.5 kW ceramic demonstrator is viewed as an obvious next
step, and made much more meaningful by incorporating a
ceramic xed-boundary recuperator, and extracting power
via a high speed generator.
4.2. Thermodynamic cycle
It is dicult for small simple-cycle gas turbines to
achieve eciencies much above 20% [24], thus from the
onset the inclusion of a high eectiveness recuperator is
mandatory for small microturbines. In small low pressure
ratio radial ow turbomachinery the impact of component
eciency is very signicant, and compared with existing
microturbines this is aggravated by the very small size of
the components in the proposed ceramic demonstrator.
It is of interest to note that two very small turbogenera-
tors with turbocharger-based turbomachinery were built
and operated in Japan. With a metallic radial ow turbine
diameter of 52 mm a small gas turbine engine for a
Fig. 2. 2.6 kW Micro gas turbine generator set (courtesy IHI).
Fig. 3. Small ceramic microturbine concept cycle diagram.
64 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
mini-cogeneration system was built and tested using natu-
ral gas fuel in Japan [25]. Utilizing an exhaust-heated cycle
the target electrical power was 3 kW at 20% eciency.
The other small microturbine (shown on Fig. 2) rated at
2.6 kW, again based on turbocharger technology, was pro-
duced as a commercial product [26] and several hundred
units were sold. While the target eciency when burning
kerosene was 20%, it fell far short of this, and an investiga-
tion was undertaken to determine the performance
This detailed investigation of the 2.6 kW turbogenerator
focused on the complex interfaces between the small com-
ponents, and the eects of uid and heat leaks, and seal
bypass ows were analysed. To a large extent the eciency
discrepancy between estimated performance and test data
was rationalized, and simple changes such as thermal isola-
tion and leak sealing were suggested as a means of restor-
ing the target eciency [27]. Lessons learned from this
investigation would be taken advantage of in the design
of the proposed 7.5 kW turbogenerator.
The simplistic nature of the proposed microturbine dem-
onstrator system is shown in the cycle diagram on Fig. 3.
While the major goal of the ceramic-based engine is to
demonstrate the integrity and performance of the complete
turbogenerator, a possible waste heat recovery module for
CHP service is shown.
5. Major component design considerations
5.1. Compressor
A single stage centrifugal compressor is the inevitable
choice for microturbines as a consequence of its cost, sim-
plicity, compactness and performance characteristics such
as wide surge margins with high inlet ow distortion toler-
ance. Typical eciency characteristics have been discussed
previously [28]. The attainable eciencies for small mass
ow compressors with impeller tip diameters on the order
of 60 mm, as would be the case for the proposed demon-
strator, would be about 75% (see Fig. 4) and largely depen-
dent on design choice of specic speed and Mach number.
Specic speed is a function of rotational speed, volume
ow, and adiabatic head. Mach number is a function of
pressure ratio and is particularly critical for small compres-
sor entry blading, necessitating very thin blades on the
order of 0.25 mm. Moderate pressure ratios are therefore
to be preferred for microturbines with power ratings less
than approximately 10 kW.
To keep the rotating assembly simple, a back-to-back
aluminum centrifugal compressor and ceramic radial tur-
bine rotor assembly is proposed. With this arrangement
careful design attention must be given to the interface
between the two, plus adjacent back shrouds to minimize
the heat soak back from the hot ceramic turbine to the
cold metallic compressor. The topic of heat transfer
between the turbine and compressor has been addressed
previously [29], and could either be a turbine cooling asset
or a performance penalty. Its eect on engine performance
can be minimized by designing the compressor near opti-
mum specic speed (small surface area/mass ow) , but it
is aggravated by higher compressor exit to turbine inlet
temperature dierentials. The long-term stability of the
ceramic-to-metallic interface must also be demonstrated.
5.2. Turbine
Inward ow radial and mixed ow turbines have estab-
lished a prominence in small turbomachines because of
their simplicity, low cost, relatively high performance,
and ease of installation. The predominant application of
these turbines are in small gas turbines and turbochargers
in the 0.052.0 kg/s ow range.
Extensive development of both radial and axial ow tur-
bines with outputs of 100 kW or less has shown the superi-
ority of the radial type particularly including eciency and
manufacturing cost. Small ceramic turbocharger radial tur-
bines have proven reliability records in the automobile
industry as a consequence of small rotor mass extending
Wiebull stress related life.
Single stage axial turbines are being used in small turbo-
jets, and an example of a small turbojet with a proposed
application as a 7 kW turbogenerator embodying a ceramic
axial turbine [30] is shown on Fig. 5. Small gas turbines
with turbine expansion ratios above 3.0 can result in over-
loading the single stage axial turbine, and for higher
Fig. 4. Size eects on eciency of radial ow compressor and turbine.
Fig. 5. Small gas turbine (7 kW equivalent rating) with ceramic axial ow
turbine (courtesy Innotech Europe BV).
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 65
eciencies two stages would be necessary. Design studies
have indicated that for the same overall turbine eciency
the single stage radial can oer lower cost, but requires a
larger diameter, higher tip speed rotor, concomitant with
increased stresses.
The attainable eciencies for rotor tip diameters on the
order of 70 mm are about 83% as shown on Fig. 4. This
radial turbine eciency is representative of levels demon-
strated in gas turbine applications with optimum blade
solidities and thickness, together with vane nozzles. Small
turbochargers utilize more robust blades to cater for
increased blade loads during pulse blow down operation,
and coincidentally benecial for ceramic casting processes.
Automobile turbocharger rotor tip speeds rarely exceed
450 m/s. Turbine entry volutes are being superseded by
variable nozzles, some with ceramic vanes, to improve
acceleration response and transient emissions.
In the automobile turbocharger eld various technolo-
gies have been exploited to reduce the characteristic
response dwell time during rotor spool-up. Reducing the
turbine rotor inertia by utilizing a ceramic radial ow tur-
bine has been adopted for some applications, particularly
in the last two decades mainly in Japan [3138].
Ceramic radial ow turbines manufactured by Kyocera
for small turbochargers (as shown on Fig. 6) have been
deployed in thousands of premium market performance
cars in Japan since the mid 1980s [39], and are still in pro-
duction today [40]. The ceramic rotor for the proposed
7.5 kW ceramic microturbine concept would be similar in
size to the one shown.
The design of microturbines with single stage radial
compressors and radial turbines is focused towards optimi-
zation of overall engine thermal eciency, thus the selec-
tion of rotational speed becomes a compromise between
both compressor and turbine aerothermodynamic design
criteria. The predicted thermal eciency variation with
rotational speed for a microturbine rated at 7.5 kW with
a turbine inlet temperature of 1170 C, and pressure ratios
of 2.53.5, as portrayed on Fig. 7 was computed using a
cycle optimization computational technique described pre-
viously [41]. The optimum rotational speed is shown to
encompass the range of 150,000200,000 rpm. A rotational
speed of 160,000 rpm with a pressure ratio of 3.0 was cho-
sen to meet the target eciency of 30%. The compressor
and turbine tip diameters are on the order of 60 and
70 mm, respectively. The turbine tip speed of 585 m/s
would be higher than current state-of-the-art automobile
ceramic turbines requiring progressive advancements in
aerodynamic design to optimize the degree of reaction
together with improved ceramic materials and brittle stress
analysis techniques.
5.3. Generator
Improvements in magnetic materials have resulted in
lighter and more ecient permanent magnet generators
(PMGs) than wound eld generators. The eld excitation
is provided by permanent magnets that are capable of oper-
ating at temperatures up to 250 C. Power output is deter-
mined by the attainable generator tip speed, diameter and
length-to-diameter ratio. Approximate generator sizing is
illustrated on Fig. 8 as limited by a rotor tip speed limit
of 250 m/s. Generator tip speed, diameter, and length
parameters would therefore bracket rotating speeds
between 150,000 and 200,000 rpm for the 7.5 kW rating,
ideally matching the aerothermodynamic optimum speed.
Generator cooling and heat rejection is a major consid-
eration and may incur additional parasitic power losses.
An air-cooled generator with ambient air being drawn over
external ns surrounding the generator casing is conceptu-
alized with a conservative overall eciency of 88%, includ-
ing frequency conversion.
5.4. Bearings
Small gas turbines currently operate with rotational
speeds from 60,000 to 150,000 rpm with both conventional
Fig. 6. Ceramic radial ow turbine from automobile turbocharger
(courtesy Kyocera International Inc). Fig. 7. Optimal rotating speed for 7.5 kW microturbine concept.
66 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
antifriction and air bearings. Some microturbines in service
today have demonstrated trouble free operation with air
bearings, and this type is proposed for the 7.5 kW ceramic
microturbine. Air bearings require no lubrication or associ-
ated lubrication cooling system, plus minimal parasitic
drag during starting. Air bearings however possess low
thrust bearing load capacity, and are sensitive to thermal
gradients, and shock loading under high g accelerations.
Air bearings do incur reduced power losses, especially the
thrust bearing, which may be as large in diameter as the
compressor impeller.
5.5. Combustor
Scaling techniques for the design of mini combustors are
less well dened due in part to the eects of the following.
Surface area/volume changes with size, increased eects of
wall quenching, low fuel ows resulting in tiny injectors,
and increased eect of leakage gaps on pattern factor.
In an earlier 5 kW turbogenerator design concept [19] an
annular combustor was selected from the gas ow path
standpoint since it was symmetrically compatible with an
annular prime-surface metallic recuperator. As will be dis-
cussed in a Section 6, having a rear-mounted cube-shaped
ceramic modular recuperator in the proposed 7.5 kW
engine design concept has a major impact on the engine
architecture. A compact overall engine arrangement, with
acceptable gas ow paths, was established with the selec-
tion of a single can combustor with a ceramic liner. The
design and fabrication of this component can take advan-
tage of an established ceramic combustor technology base
[11]. Emissions and carbon dioxide release are rapidly
becoming a dominant criterion in the design of small
microturbines, and a development eort on this small com-
bustor is foreseen. The proposed ceramic microturbine
demonstrator would be natural gas-red, but could be
engineered to operate with a variety of liquid fuels, includ-
ing biodiesel and ethanol which are currently receiving a lot
of attention. However, the use of liquid fuels can increase
the likelihood of build up of carbon deposits in the com-
bustor. When they become dislodged from the combustor
or fuel injectors they could cause damage to the ceramic
nozzle and rotor blades. Radial impellers are more suscep-
tible to damage because the interspace centrifugal forces
tend to reverberate the particles back against the nozzle
trailing edge causing erosion. One remedy to this problem
is the use of axial type turbine nozzles.
Natural gas is a fuel of choice for small business and
domestic microturbines, but most probably would require
compression from essentially atmospheric pressure to levels
exceeding microturbine compressor delivery pressure.
Selection of microturbine cycle pressure ratio therefore
requires consideration being given to the gas supply sys-
tem. The design of the very small gas compressor would
take advantage of units currently operating and the use
of emerging technologies.
6. Ceramic recuperator
6.1. Background
A separate section is devoted to the recuperator since in
the microturbine eld it is the component that has received
the least development attention, and because the proposed
heat exchanger represents a major ceramic recuperator
technology advancement. For several decades there has
been interest in ceramic recuperators and regenerators for
a variety of small gas turbine applications, and the merits
of each type have been discussed previously [42,43]. Devel-
opment activities were primarily focused on vehicular gas
turbines and a wide range of surface geometries and types
of construction were tested [44,45]. Organizations world-
wide have undertaken ceramic recuperator development,
but over the years there has been no coordinated eorts
or continuity. In general, technical progress has been mod-
est, and today it is recognized in the microturbine eld that
further ceramic recuperator development activities are war-
ranted. In their current metallic form recuperators are
expensive and represent about 30% of the overall turbogen-
erator cost.
Until recently the authors had felt that ceramic recuper-
ators would have as their genesis the same types of surface
geometry and construction as metallic heat exchangers
(e.g., platen, primary surface, tubular, etc.), but an
emerging ceramic heat exchanger technology has made this
thinking essentially obsolete as discussed below.
6.2. Ceramic micro-channel recuperator
Utilizing laminated object manufacturing (LOM) meth-
ods a compact ceramic recuperator for use in high temper-
ature microturbines is being developed by Ceramatec [46].
The selection of silicon carbide was based on the following
considerations; cost, fabrication and compatibility with the
gas turbine environment. The LOM process begins with the
blending of ceramic powders and organic binders. This
blend (i.e., ceramic slip) is then cast onto a mylar lm that
Fig. 8. Permanent magnet generator sizing.
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 67
dries and cures into a exible form, and in the green-state
micro-channels are formed. These are then sintered to form
monolithic structures amenable to mass production. For
gas turbine recuperator applications these micro-channels
typically have air side and gas side hydraulic diameters of
1 and 2 mm, respectively. With such a compact surface
geometry, the engine would be run initially with natural
gas as the fuel. In forming the matrix a thin dense layer
of silicon carbide that acts as the primary surface is posi-
tioned between the channels.
The micro-channels are designed into a planar or plate
that can be installed as part of the shell-and-plate type of
construction. An example of the basic ceramic plate is
shown on Fig. 9, and these are the primary building blocks
of the recuperator and include features for gas manifold-
ing, and macro features to be integrated into the gas head-
ers to give the counterow conguration as shown by the
arrows on Fig. 9. By stacking and bonding these plates
with spacers, narrow gaps are formed and dene the gas
channels for the turbine exhaust gas. The plates are then
stacked to form the recuperator matrix as shown on
Fig. 10. This cube-shaped module is manifolded for instal-
lation on the rear of the microturbine.
To achieve the targeted engine performance goal a
demanding requirement is put on the recuperator in terms
of eectiveness. The impact of recuperator eectiveness on
engine eciency is shown on Fig. 11. For the selected tur-
bine inlet temperature, an eectiveness of 0.92 is required
to meet the turbogenerator eciency goal of 30%. The high
surface compactness of the silicon carbide micro-channel
recuperator results in a very small heat exchanger size (vol-
ume of approximately 10,000 cubic centimeters) that inte-
grates well with the rotating machinery to give a compact
and light weight turbogenerator package. In establishing
the size of the counterow recuperator the eectiveness-
NTU relationship for both metallic and ceramic variants
is the same. The part load characteristics would be slightly
dierent since longitudinal conduction eects are related in
part to the matrix material thermal conductivity.
The utilization of such a compact high temperature
recuperator with the type of construction amenable to
low cost high volume production is regarded as a key fac-
tor for the future deployment of the next generation of
higher eciency microturbines. The design, fabrication,
development and testing of the proposed ceramic microtur-
bine demonstrator is viewed as the rst step towards this
7. Proposed engine design layout concept
The layout of the 7.5 kW microturbine concept is shown
in a simple form on Fig. 12. The compact turbogenerator
has the following major features: (1) single stage centrifugal
compressor; (2) can combustor with a ceramic liner; (3)
ceramic radial ow turbine and nozzle; (4) ceramic
micro-channel recuperator; (5) ceramic interconnecting
Fig. 9. Ceramic micro-channel recuperator element and ow conguration (courtesy Cermatec Inc).
Fig. 10. Ceramic micro-channel gas turbine recuperator module (courtesy
Cermatec Inc).
68 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
hot gas ducts and scrolls; (6) high speed rotor supported on
hydrodynamic air bearings; and (7) direct-drive air-cooled
permanent magnet generator.
The basic turbogenerator design concept shown on
Fig. 12 is very compact, with a height of about 300 mm
and an overall length on the order of 500 mm. Several
components not shown on the layout include the control
system, gas compressor, instrumentation, other small
accessories, electrical equipment and thermal insulation.
When assembled within a packaged enclosure the envelope
Fig. 11. Impact of recuperator eectiveness on small ceramic microturbine performance.
Fig. 12. Ceramic microturbine demonstrator concept layout.
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 69
of the unit would be about the same size as a domestic dish
8. Projected performance of ceramic microturbine
The proposed 7.5 kW turbogenerator concept is a fol-
low-on to a 5 kW machine design suggested previously by
the authors [19]. The design of this small turbogenerator
was based on state-of-the-art component technology and
the use of existing materials. For this all-metallic engine
designed with conservative parameters (i.e. pressure ratio
3.0, turbine inlet temperature 900 C, and recuperator
eectiveness 0.85) the estimated eciency was 21.5%.
While this concept was aimed at meeting the total energy
needs of an average house [47] this natural gas-red micro-
turbine did not attract any interest. The low eciency com-
pared with existing ICEs and emerging SOFCs was felt to
be a factor. Since focus in this paper is on a ceramic micro-
turbine concept, in-depth comparisons with other systems
including ICEs and fuel cells are not really appropriate
at this stage.
Using the component eciencies discussed in a previous
section the power rating of the aforementioned all-metallic
machine was increased to 7.5 kW and means of improving
the eciency by raising the turbine inlet temperature was
investigated. Advantage was taken of previous work done
on the eects of engine size on performance [48], particu-
larly an assessment of system losses. Typical overall cycle
pressure losses for small recuperated microturbines range
from 5% to 11% basically dependent on the type of recu-
perator and combustor. The design overall cycle pressure
loss selected was 10%. The overall generator eciency
selected was 88% including power conditioning and bear-
ing losses. Engine output would be reduced approximately
6% if an electrically driven fuel gas compressor was
required. A representative performance array relating the
salient parameters for the 7.5 kW machine is shown on
Fig. 13. Superimposing the eciency of the earlier 5 kW
all metallic turbogenerator on this gure is not valid
because of the dierent levels of recuperator eectiveness
used for the two machines.
The ceramic turbine inlet temperature is essentially
determined by the turbine material Wiebull strength, high
cycle fatigue strength, oxidation resistance, and duty cycle.
Another important cycle parameter is the recuperator hot
gas inlet temperature, which in the past has been limited
by metal oxidation considerations together with life
requirements for the selected material. The pressure ratio
is determined by the compressor type and the material
used. The rotational speed is initially selected during cycle
optimization studies as described in Section 5.2 and subse-
quently rened by rotor dynamics, bearing loads, and gen-
erator considerations, together with rotor stress limitations
commensurate with the machine life requirement.
Based on various aerodynamic, thermal, and stress load-
ing criteria an ambitious eciency goal of 30% was tar-
geted for such a small turbogenerator. The selection of
30% eciency was based on what was felt to be realizable
with a very small ceramic machine, and to avoid confusion
is not meant to be compared with existing metallic micro-
turbines in the 30100 kW power range that are operating
Fig. 13. Small microturbine performance array.
70 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
with eciencies on the order of 30%. From Fig. 13 it can be
seen that this is projected with a pressure ratio of 3.0 and a
turbine inlet temperature of 1170 C. To achieve this a
large demand is placed on the recuperator, necessitating a
high eectiveness value of 0.92. Compared with the previ-
ous 5 kW design [19] the temperature of the hot gas enter-
ing the recuperator mandates the use of a ceramic heat
exchanger. The major parameters and features of the
7.5 kW ceramic microturbine demonstrator concept are
highlighted on Table 2. At this conceptual stage it is recog-
nized that these parameters are tentative and a comprehen-
sive trade study involving the inuence of individual
component eciencies and duty cycle on engine perfor-
mance would be undertaken as part of the conceptual
9. Cost considerations
There have been many microturbine papers published in
the open literature in recent years, but because of the pro-
prietary nature of the business very little cost data has been
reported by the gas turbine industry. This is disappointing
to researchers since cost data to be really meaningful must
come from the microturbine manufacturing industry. The
few numbers available for microturbines range from
$1000/kW for fully equipped units being produced in lim-
ited quantities [49], to a projected cost of $400/kW for a
30 kW unit with an annual production of 100,000 units
In a previous paper by the authors a preliminary cost
breakdown of a 5 kW turbogenerator was reported [19].
In updating this cost generated six years ago a cost target
of $250/kW for a metallic turbogenerator in the power
range of 510 kW has been reported [51]. This estimate is
based on turbogenerators being mass produced in very
large quantities like automobile turbochargers that are fab-
ricated in automated factories in Europe.
The above cost data can be regarded only as a target for
a small ceramic turbogenerator since an established data
base does not exist to permit a realistic cost estimate. Sim-
ilarly, more in-depth studies are required to establish the
overall project cost, involving the analysis, design, fabrica-
tion, test facilities, component development and testing of
a small recuperated ceramic microturbine demonstrator
of the type proposed in this paper.
10. Ceramic microturbine demonstrator tentative schedule
In an eort to put into perspective the time-frame for a
program leading to the demonstration of a 7.5 kW recuper-
ated ceramic-based turbogenerator prototype, a tentative
and simplistic schedule was prepared and is shown on
Fig. 14. The development of the metallic components
would draw heavily on technologies already deployed in
existing microturbines, particularly experience gained in
the fabrication of very small gas turbine components [52].
The deployment of air bearings in such a small machine
is recognized to be a formidable task, and a development
eort would be initiated right from the start of the project.
The pace of the program would be determined by the
fabrication and integrity of the ceramic components. Pro-
totype ceramic fabrication would benet greatly by the
advent of green and bisque machining which would reduce
lead and iteration time over production methods used for
automotive turbocharger rotors. Ducting, sealing and
attachments will also be key to successful deployment.
While full advantage would be taken of established ceramic
technology bases, there would be a learning process and
inevitable failures along the way. The design and fabrica-
tion of the radial ow turbine would have as its genesis
the know-how from similar sized turbines used for almost
two decades in automobile turbochargers in Japan. It being
recognized from the onset, of course, that there is a large
dierence in operating conditions between that of a turbo-
charger and a gas turbine. The ceramic micro-channel recu-
perator is new, and the extensive development and testing
required could well be the pacing item, since assurance of
its structural integrity and performance is a key factor
towards the success of the proposed demonstration project.
From the onset it has to be recognized that the 30%
eciency goal is ambitious for such a small turbogenera-
tor. If at the end of the envisioned three year program,
the prototype has demonstrated mechanical integrity with
Table 2
Salient features of small recuperated ceramic microturbine demonstrator
Component/parameter Feature
Turbogenerator type Advanced small ceramic
Power (kW) 7.5 (11 hp for mechanical drive)
Demonstrator eciency target/goal
Thermodynamic cycle Recuperated
Turbomachine type Single-shaft rotor
Rotational speed (rpm) 160,000
Bearing type Air bearings
Turbine type Single stage radial
(70 mm diameter)
Turbine inlet temperature, C (F) 1170 (2137)
Turbine eciency 83
Turbine material Monolithic ceramic
Compressor type Single stage radial
(60 mm diameter)
Compressor inlet temperature, C (F) 30 (86)
Compressor pressure ratio 3.0
Compressor eciency (%) 75
System pressure loss (%) 10.0
Generator type PMG air-cooled
Generator eciency (%) 88
Combustor eciency (%) 98
Recuperator type Compact counterow
Recuperator material Silicon carbide
Recuperator eectiveness 0.92
Recuperator gas inlet temperature,
(C) (F)
906 (1663)
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 71
an eciency close to 30% it will have been judged a success.
Further nessing of the prototype into the fourth year
could be foreseen to fully explore its eciency potential,
and for it to meet demanding emission goals, together with
rigorous thermal cycling to prove durability.
11. Summary
In this paper a ceramic microturbine demonstrator con-
cept has been presented, the development and operation of
which could provide the basis for the eventual development
of microturbine eciency to 40% or higher. While full
advantage would be taken of earlier gas turbine develop-
ment activities based on the use of ceramic components,
a new approach based on a much smaller turbogenerator
is suggested. The rationalization for this was based on fac-
tors which included low risk, modest cost eort, plus a
good chance of being successful to get a complete recuper-
ated ceramic microturbine operational in less than about
three years.
A conventional turbogenerator conguration layout has
been discussed this being based on a single shaft approach
using radial ow turbomachinery, a can combustor, rear-
installed recuperator, direct drive air cooled generator with
the high speed rotor supported by air bearings. For the rst
time it represents a microturbine involving the coupling of
a ceramic combustor, ceramic radial ow turbine, and a
ceramic xed boundary recuperator. The selection of the
7.5 kW machine rating was strongly inuenced by two of
the major ceramic components, namely the radial turbine
and the counter-ow modular recuperator. The size of
the ceramic turbine is similar to those used in automobile
turbochargers that have been in service in Japan since the
mid 1980s. The ceramic micro-channel recuperator is new
and initial development has been encouraging for gas tur-
bine service. Such a high eectiveness heat exchanger con-
tributes signicantly towards the goal of demonstrating an
eciency of 30%.
Two of the major objectives of the proposed R&D eort
are to demonstrate the performance potential and struc-
tural integrity of a small microturbine embodying ceramic
components. Additional data will be acquired in the fol-
lowing areas; verication of analytical and design method-
ologies, application of new ceramic manufacturing
processes, the ability of small ceramic components to with-
stand thermal transients, observe the impact of foreign
material ingestion, materials-related eects when running
on dierent gaseous and liquid fuels, characterization of
emissions, and means of acoustic attenuation.
The ultimate goal of the proposed microturbine demon-
strator unit will be to establish a technology base that
would provide a benchmark for the eventual realization
of microturbines rated up to about 100 kW with eciencies
of over 40% based on the use of ceramic components.
In closing, it is germane to project beyond the successful
operation of the demonstrator to potential applications of
such a small recuperated microturbine including: (1) a
natural gas-red CHP system to meet the total energy
needs of an average house; (2) small energy systems using
indigenous fuels in the developing nations; (3) it could be
coupled with a SOFC to give a very high eciency and
low emissions hybrid power source [53] and (4) perhaps
meet propulsion needs for future extended endurance
Fig. 14. Tentative schedule for ceramic microturbine demonstrator.
72 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074
In preparing this paper the authors would like to express
their thanks to the following for helpful discussions,
providing material, and for their constructive review
comments: Merrill Wilson (Ceramatec Inc.), Jay Scovie
(Kyocera Intl. Inc.), Takero Fukudome (Kyocera, Japan),
Bryan Seegers (M-Dot), Dr. Hiro Yoshida (AIST), Peter
Kuijpers (Innotech Europe BV), and consultants Dr. John
Mason, David Carruthers and David Richerson. This
paper has been enhanced by the inclusion of hardware
photographs and the authors are appreciative to all con-
cerned, with credits being duly noted.
[1] C.F. McDonald, C.Rodgers, Ceramic recuperator and turbine. The
key to achieving a 40% ecient microturbine, ASME Paper GT2005-
68644, 2005.
[2] C. Rodgers, Conceptual engine study for a 400 kW ICR turbogen-
erator, ASME Paper GT2005-68043, 2005.
[3] K.Takase,, A preliminary study of an intercooled and recuper-
ative microturbine below 300 kW, ASME Paper GT-2002-30403,
[4] C. Rodgers, A. Stone, D. White., A gas turbine selection issue-
recuperated or ICR, ASME Paper GT2007-27910, submitted for
[5] C. Invernizzi et al., Bottoming micro-rankine cycles for micro-gas
turbines, Journal of Applied Thermal Engineering 27 (1) (2007) 100
[6] J. Shi, Preliminary design of ceramic components for the
ST5 + advanced microturbine, ASME Paper GT-2002-3057, 2002.
[7] L.Y. Bronicki, D.N. Schochet, Bottoming cycles for gas turbines,
ASME Paper GT2005- 68121, 2005.
[8] M. VanRoode et al., Ceramic gas turbine design and test experience,
Progress in Ceramic Gas Turbine Development, vol. 1, ASME Press,
N.Y, 2002.
[9] M. Van Roode et al., Ceramic gas turbine component development
and characterization, Progress in Ceramic Gas Turbine Development,
vol. 2, ASME Press, N.Y, 2003.
[10] W.D. Carruthers et al., Advances in the development of silicon nitride
and other ceramics, ASME Paper GT-2002-30504, 2002.
[11] J. Price, Advanced materials for gas turbine combustion systems:
program summary, ASME Paper GT2004-54250, 2004.
[12] S. Tsuruzona et al., Development and evaluation of ceramic
components for 8000 kW class hybrid gas turbine, ASME Paper
2001-GT-0516, 2001.
[13] B. Treadway et al., Design of ceramics for an advanced micro-turbine
engine, ASME Paper GT2004-54205, 2004.
[14] T. Fukudome et al., Development and evaluation of ceramic
components for gas turbines, ASME Paper GT-2002-30627, 2002.
[15] D.W. Richerson, Historical review of addressing the challenges of use
of ceramic components in gas turbine engines, ASME Paper GT2006-
9033, 2006.
[16] R. Lundberg, M. Ferrato, Ceramic component development for
AGATA, ASME Paper 99-GT-392, 1999.
[17] I. Takehara, Summary of CGT302 ceramic gas turbine research
and development program, ASME Paper 2000-GT-644, 2000.
[18] T. Nisiyama et al., Status of automotive ceramic gas turbine
development program-seven years progress, ASME Paper 97-GT-
383, 1997.
[19] C.F. McDonald, C. Rodgers, The ubiquitous personal turbine a
power vision for the 21st century, ASME Paper 2001-GT-100. 2001.
[20] H. Yoshida et al., Micro gas turbine with ceramic rotor, ASME Paper
[21] Personal communication with Dr. Hiro Yoshida (AIST), November
15, 2005.
[22] T. Matsunuma et al., Micro gas turbine with ceramic nozzle and
rotor, ASME Paper GT2005-68711, 2005.
[23] I. Norihiko et al., Micro gas turbine with ceramic nozzle and turbine
Part 2, ASME Paper GT2006-90328, 2006.
[24] C. Rodgers, Microturbine cycle options, ASME Paper 2001-GT-0552,
[25] M. Sakakida et al., A study of a small gas turbine engine for mini-
cogeneration system, in: Proceedings of the International Gas
Turbine Congress, Kobe, Japan, November 1419, 1999, pp. 1069
[26] T. Nakajima et al., The development of the 2.6 kW micro gas turbine
generator, Paper Presented at International Gas Turbine Congress,
Yokohama, Japan, October 2227, 1995.
[27] M.A. Monroe et al., Component integration and loss sources in 3
5 kW gas turbines, ASME Paper GT2005-68715, 2005.
[28] C. Rodgers, The eciencies of single stage centrifugal compressors
for aircraft applications, ASME Paper 91-GT-77, 1991.
[29] C. Rodgers, Design and development of a monorotor gas turbine
auxiliary power unit, ASME Paper 78-WA/GT-2, 1978.
[30] Personal communication with Peter Kuijpers, November 28,2005.
[31] R.P. Walson et al., Performance of a ceramic rotor in a cummins T46
turbocharger, SAE Paper 840014, 1984.
[32] Y. Okazaki et al., Ceramic turbine wheel developments For Mitsu-
bishi, SAE Paper 850312, 1985.
[33] M.F. Lasker, Ceramic turbine wheel for turbochargers, SAE Paper
861130, 1986.
[34] K. Katayama et al., Development of Nissan high response ceramic
turbocharger rotor, SAE Paper 861128, 1986.
[35] I. Matsuo, F. Nishiguchi, The development of second generation
ceramic turbocharger, SAE Paper 880703, 1988.
[36] C.C. Baker, Garrett experience in ceramic turbocharger turbine
wheels, SAE Paper 890426, 1989.
[37] Y. Katano et al., Application of ceramics to turbocharger rotors for
passenger cars, ASME Paper 91-GT-264, 1991.
[38] H. Kawase et al., Development of ceramic turbocharger rotors for
high temperature use, ASME Paper 91-GT-270, 1991.
[39] Personal communication with Jay Scovio, Kyocera Corporation,
September 20, 2005.
[40] Personal communication with Takero Fukudome, Kyocera Corpo-
ration, R&D Centre, Kagoshima, Japan, February 22, 2006.
[41] C. Rodgers, Thermo-economics of small 50 kW turbogenerator,
ASME Paper 97-GT-260, 1997.
[42] C.F. McDonald, D.G. Wilson, The utilization of recuperated and
regenerated engine cycles for high eciency gas turbines in the 21st
century, Journal of Applied Thermal Energy 16 (8&9) (1996) 635
[43] P. Avran, S. Boudigues, Comparisons of regenerators and xed heat
exchangers for automotive gas turbine applications, Paper Presented
at Yokohama International Gas Turbine Congress, Yokohama,
October 27November 1, 1991.
[44] C.F. McDonald, Heat recovery exchanger technology for very small
gas turbines, International Journal of Turbo and Jet Engines 13 (4)
(1996) 239261.
[45] C.F. McDonald, Ceramic heat exchangers. The key to high eciency
in very small gas turbines, ASME Paper 97-GT-463,1997.
[46] M.A.Wilson et al., Design and development of a low-cost high
temperature silicon carbide micro-channel recuperator, ASME Paper
GT2005-69143, 2005.
[47] S. Mraz, A Microturbine in every house, Machine Design, March 7,
2002, p84.
[48] C. Rodgers, Some eects of size on the performance of small gas
turbines, ASME Paper GT2003-38027, 2003.
[49] S. Gelsi, Capstone Turbines Eyes PC-Like Role, Market Watch
Publication, 2000.
[50] S. Dunn, Micropower the next electrical era, World Watch Paper
151, July 2000, p22.
C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074 73
[51] C. Rodgers, C.F. McDonald, Small recuperated gas turbine APU
concept to abate concern about emissions, high fuel cost, and noise,
ASME Paper GT2007-27913, submitted for publication.
[52] M-DOT Web site <http://www.m-dot>., December 2005.
[53] L. Magistri et al., A hybrid system based on a personal turbine
(5 kW) and a solid oxide fuel stack: a exible and high eciency
energy concept for the distributed market, ASME Journal For Gas
Turbines and Power 24 (2002) 850857.
74 C.F. McDonald, C. Rodgers / Applied Thermal Engineering 28 (2008) 6074