You are on page 1of 13

Applied Thermal Engineering 24 (2004) 18511863

www.elsevier.com/locate/apthermeng

General characteristics of single shaft microturbine set


at variable speed operation and its optimization
Wei Wang, Ruixian Cai *, Na Zhang
Institute of Engineering Thermophysics, Chinese Academy of Sciences, P.O. Box 2706, Beijing 100080, China
Received 4 July 2003; accepted 9 December 2003
Available online 23 January 2004

Abstract
Microturbine is widely applied now in distributed generation. Here the capacity of microturbine is dened as about 200 kW power. Commonly when the load changes, the generating gas turbine has still to run
at design rotating speed. This results in lower part-load eciency. However, the part-load performance of
microturbine can be improved by variable speed operation since constant frequency of electricity can be
guaranteed through the control of the electronic equipment. This optimum operation mode has not yet
been theoretically studied thoroughly. The paper rstly analytically investigates the general optimum speed
variation for o-design eciency with an analytical approach proposed by the last two authors. Furthermore, the study shows that the optimum part-load performance is basically not inuenced by the design
temperature ratio and pressure ratio. Because of using a typical component performances and analytical
approach of o-design performance analysis, the conclusions are general and typical for common microturbines but not fully accurate for any special microturbine.
2004 Elsevier Ltd. All rights reserved.
Keywords: Microturbine; Part-load performance; Variable speed; Analytical approach

1. Introduction
In recent years, the conception and technology of distributed generation are being
widely concerned and eetly developed [16]. Commercial and industrial energy consumers
now have a wider choice for electric power. At present, the distributed power generation
equipments being researched in the world mainly include internal combustion engine, fuel cell,

Corresponding author. Tel.: +86-10-625-618-87; fax: +86-10-625-759-13.


E-mail address: crx@mail.etp.ac.cn (R. Cai).

1359-4311/$ - see front matter 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2003.12.012

1852

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

Nomenclature
G
N
n
p
T
t
d
g
l
p
r
s
u

mass ow rate
power output
rotating speed
pressure
temperature
time
dimensionless average heat transfer dierence
eciency
mass ow rate ratio
pressure ratio
recuperator eectiveness
temperature ratio
total pressure recovery coecient

Superscripts

divided by design value


.
reduced parameter (ratio of the value at part-load to its design value)
Subscripts
c
compressor
inv
constant speed
rc
recuperative cycle
sc
simple cycle
t
turbine
v
variable speed
0
design value
1
inlet of compressor, atmospheric
exit air of recuperator
20
3
inlet of turbine
4
exit of turbine

micro gas turbine and other small-scale combined power generation systems. Micro gas turbine (ranging about 20500 kW) is an important one. It can produce reliable economical
electric power to satisfy consumers. Due to the adoption of new technologies, the eciency of
micro gas turbine highly increases. For example, the simple cycle eciency of micro gas
turbine with an ecient regenerator reaches from 30% to 33%. The eciency of the hybrid
cycle that consists of micro gas turbine and high temperature fuel cell will reach from 65% to
70%. Moreover, it also has many own merits including simpleness and tightness of conguration, extensive applicability of various liquid and gas fuels and convenience of maintaining,
etc.

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

1853

Most of the power plants often run under o-design situations due to the change of the load or
ambient conditions or both. So the research of o-design situations is even more important than
that of design point, and the former is more dicult than the latter. Most of the design performances have been fully studied. While except individual simple situations, the understanding of
part-load behavior, especially some concise semi-quantitative rules, is far from enough because of
their complexity. Presently there are a lot of ecient numerical methods and computer codes
of various power sets under o-design situations in the world. They have greatly promoted the
performance analysis and the practice of power plants. Their eectiveness and accuracy are
veried usually by the experimental results. However, the practical experimental data are scarce
perhaps due to commercial reasons, they cannot be exactly accurate due to the experimental
errors. Besides, the performance is often needed before producing a new power set, but it is
impossible to attain practical experimental data at the time. If there are typical analytical solutions of power plants performances, the aforementioned problems do not exist. Firstly, analytical
solution is an eective analysis means to check and develop the numerical solution when both
solutions using the same component performance. In addition, some conclusion can also be
summarized to enrich the academic research.
Owing to the constant frequency of turbo generators, the routine single shaft gas turbine
commonly operates at constant rotating speed. However, the rotating speed of micro gas turbine
is so high that it is better to directly drive a high-speed DC generator or a variable frequency AC
generator with frequency control electronic equipment to avoid a heavy and expensive gearbox. It
is well known that the part-load eciency of single shaft gas turbine is better at lower rotating
speed than that at constant speed. Therefore for such microturbines, the rotating speed should be
variable under part-load operation. The paper researches the optimum o-design characteristics
of the gas microturbine composed of radial compressor and radial-inow turbine at variable
rotating speed with an analytical approach proposed by the last two authors. For instance, the
research on the relationship of optimum rotating speed to power output could discover the best
variable rotating speed operating line. Operating along this line under part-load will obtain
highest eciency. The regularities of various parameters summarized according to this way may
not suit to a specic case exactly, but they represent the typical performances of all kinds of gas
microturbine, and have theoretical and practical value.
2. Analytical performance expression of gas microturbine set under o-design situations
The analytical approach of o-design performance analysis proposed by the last authors in
Refs. [7,8] is adopted to give the performance of microturbine. The main ideas of this method are
as follows: expressing the general performance of components (for example, compressor and
turbine) as close as possible to the typical relations with simple analytical functions, then it is able
to derive the analytical typical performances of total set. These analytical performances cannot
accurately simulate any specic practical set but they can simulate the general rules of one kind of
power sets. These ideas have been proven eective.
The main calculation strategy and expressions are generally the same as Ref. [7]. However, there are dierences between the common gas turbines with axial turbomachines and
microturbines with radial turbomachines. Some modications, which are needed to improve the
method, are implemented as follows.

1854

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

The following performance formulas for compressor and turbine are proposed in Ref. [7].
(1) Performance formulas for compressor
p_ c c1 n_ c G_ 2c c2 n_ c G_ c c3 n_ c
i
h
2
g_ c 1  c4 1  n_ c n_ c =G_ c 2  n_ c =G_ c
where:
p_ c pc =pc0 ;
c1 n_ c

.h

G_ c Gc =Gc0 ;

p
Gc T 1
Gc
;
p1

p1  m=n_ c n_ c n_ c  m2

n_ c

nc
;
nc0

nc
nc p ;
T1

g_ c gc =gc0


.h
i
c2 p  2mn_ 2c
p1  m=n_ c n_ c n_ c  m2

.h
i
c3  pmn_ c  m2 n_ 3c
p1  m=n_ c n_ c n_ c  m2
p
c4 is undetermined constants, commonly c4 0:3, 3 p P 2m=3, m and p are undetermined
constants.
(2) Performance formulas for turbine
A modied Fl
ugel formula is used to approximately describe the mass ow characteristics of
turbine:
p q

p2t  1 =p2t0  1
Gt =Gt0 a T30 =T3 
p
where a 1:4  0:4nt =nt0 is adopted to consider the inuence of rotating speed.
2
The eciency characteristic
is dened as: g_ t 1  t4 1  n_ t
n_ t =G_ t 2  n_ t =G_ t , where
p
G
T
n

n
t
G_ t Gt =Gt0 , Gt tp3 3 , n_ t nt0t , 
nt p
, g_ t gt =gt0 , t4 is an undetermined constant, comT3
monly t4 0:3.
(3) Equilibrium condition
Gas turbine steady running should maintain the following equilibrium conditions: pt u  pc ,
_ c =Gc0 . The pressure recovery coecient u
Gt l  Gc , nt nc namely, p_ t u_ p_ c , Gt =Gt0 lG
and the mass ow ratio l are usually chosen to be constants, i.e. u_ l_ 1:0. The part-load
performance calculation can be carried out after two independent variables are given.
From Ref. [7], it can be known that the choice of constants (such as: m, p) in the component
performance formulas is the key to successfully calculate cycle performances. Usually these values
of constants are determined by experience. In this study these values for microturbine are also
determined based on the particularity of microturbine and our experience. The performances of
radial compressor and radial turbine for microturbines are shown in Figs. 1 and 2 with m 1:8
and p 1:8 [9] instead of m 1:06 and p 0:36 for axial ow larger gas turbines.

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

1.0

0.5

n c = 1.0

1.2

0.9
0.8

0.8
0.7
0.4

0.6

Gc
0.0
0.2

0.4

0.6

0.8

1.0

1.2

Fig. 1. Compressor characteristics map.

1.0

1.0
0.8

0.8

0.6

0.6

1.2

n t = 1.0

0.6

0.8

0.4

Gt
0.0
0.4

0.6

0.8

1.0

Fig. 2. Turbine characteristics map.

1.2

1855

1856

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

For recuperative cycle, according to Ref. [10], the classical recuperator eectiveness r, dierent
from the eciency criterion of turbine, compressor and combustion chamber, cannot reect
theoretic perfection and has practical diculty to correctly design the components. In fact, another parameterthe average heat transfer temperature dierence DT T4  T20 or dimensionless
number d DT =T1 in the recuperatoris recommended as the new criterion instead of the recuperator eectiveness. But the part-load data of average heat transfer temperature dierence are
rare in open literature. Moreover, the part-load data of classical recuperative eectiveness r has
more authoritative and concise common formula. Hence, the following o-design formula is used
in our study: r r0 =r0 1  r0 G=G0 0:2
.

3. The comparison between experimental results and variable speed analytical performance
expressions
Nowadays, the part-load performances of variable speed operation can hardly be obtained.
Actually, constant speed operation is a kind of variable speed operation, then we verify the
analytical expressions with constant speed experimental results. Fig. 3 shows the comparison of
experimental results and analytical part-load performances of constant rotating speed small radial
ow gas turbines. It can be seen from the gure, the experimental results are almost on the
analytical curve. It means that the performance expressions are basically valid.

4. Typical part-load performances of microturbine at variable rotating speed


For simulating microturbine part-load performances at variable rotating speed, the values of
design condition chosen for the rst example are as follows: pc0 6:0, s0 4:5, gc0 0:82,
gt0 0:85; Pressure losses are represented by the pressure recovery coecient u 0:97; the
processes of compression, heating and expansion are calculated by means of average specic heat.
1.0

0.8

0.6

0.4

0.2

0.0
0.0

N
0.2

0.4

0.6

0.8

1.0

Fig. 3. Relative analytical cycle eciency curve and comparison with practical data.

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

1857

For recuperative cycle, the values of design condition for the rst example are chosen as pc0 5:0,
s0 4:5, r0 0:9.
The part-load performance curves of single shaft microturbine without recuperation are shown
in Figs. 4 and 5. Accordingly Figs. 6 and 7 show part-load performance of single shaft microturbine with recuperation. Fig. 4 shows the variation of relative speed to relative power output
under maximum eciency part-load condition. It can provide valuable information for control of
microturbine. The variation of relative eciency to relative power output under two dierent
conditions is shown in Fig. 5. The upper curve represents the o-design eciency under optimum
variable speed operation and the lower one under constant speed operation. The obvious difference between variable speed operation and constant speed operation is due to the dierent
variations of the mass ow rate and then the cycle maximum temperature, which has been explained many times in gas turbine textbooks.

1.00

0.95
0.90
Variable rotating speed
Constant rotating speed

0.85
0.80
0.75

N
0.70
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 4. Relative rotating speed variation.

1.0
0.8
0.6
Variable rotating speed
Constant rotating speed

0.4
0.2

0.0

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 5. Relative micro gas turbine eciency variation.

1858

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

n
1.00
0.95
0.90
0.85

Variable rotating speed


Constant rotating speed

0.80

0.75

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 6. Relative rotating speed variation with recuperation.

1.0
0.8
0.6
Variable rotating speed
Constant rotating speed

0.4
0.2

0.0

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 7. Relative micro gas turbine eciency variation with recuperation.

Figs. 6 and 7 are the performance curves of recuperative cycle. From Fig. 7, it can be seen that
if there exists recuperator, the optimum g_ N_ line is much higher and the advantage of variable
rotating speed operation is more evident. It can be explained from the recuperator eectiveness
formula [10] that decrease of the rotating speed and then pressure ratio in part-load operation will
raise the recuperator eectiveness.

5. Typical performance of optimum variable speed microturbine under dierent design situation
From foregoing section, it has been seen that the operation at variable speed is better than that
at constant speed. In this section we are going to analyze microturbine part-load performance of
wider design range at optimum variable speed and to show some conclusions.
Owing to simplicity of analytical solutions and its possibility to satisfy many strict mathematical operations, systematic study can be made very easily and so typical characteristics of part-

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

1859

load operation along maximum eciency variable rotating speed can be summarized through
great amount of calculations under various cases. Fig. 8 shows the relationship between relative
eciency and relative power output for the simple cycle in which the design pressure ratio varies
from 4 to 10 and the design temperature ratio from 3.5 to 5.0. Fig. 9 shows the relationship
between relative eciency and relative power output for the recuperation cycle with the design
pressure ratio varying from 3 to 6 and the design temperature ratio also from 3.5 to 5.0. It can be
found that most curves of relative eciency versus relative power output along the optimum
operation line can almost have the same trend and locate in a narrow area. However, the line most
departing from the average value represents the lowest design parameters and is out-of-date,
actually it can be omitted, then the line concentration is much higher. But the departure degree of
these curves of recuperation cycle is more obvious.
In general, the variation of design temperature ratio has a more strong eect on the curve of
g_ N_ and the variation of design pressure ratio has a weaker eect. That is to say, under the same

1.0
0.8
0.6
0.4
0.2

0.0

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 8. Relative micro gas turbine eciency variation.

1.0
0.8
0.6
0.4
0.2

0.0

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 9. Relative micro gas turbine eciency variation with recuperation.

1860

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

relative power output, when design temperature ratio is higher, the eciency under part-load
condition is higher. For these denser curves in Figs. 8 and 9, the variation of eciency versus
power output can almost be summarized into a single curve. The tting formulas are respectively
as follows:
g_ 4:56N_  13:37N_ 2 23:15N_ 3  20:036N_ 4 6:696N_ 5 for non-recuperation; neglecting s
3:5
g_ 4:8N_  14:02N_ 2 23:88N_ 3  20:412N_ 4 6:752N_ 5 for recuperation; neglecting s 3:5
Dierent design pressure ratio and temperature ratio have more notable eect on the relation
(n_ N_ ) under optimum variable speed operation situation. These curves are shown in Figs. 10
and 11. The optimum variable speed curves are disperse, but similar to previous Figs. 8 and 9, for
common given design range of p and s, the curves of n_ N_ basically have the same trend. In the
same way, comparing with the design pressure ratio, the variation of design temperature ratio has

1.00

0.95
0.90
0.85
0.80
0.75

0.70

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 10. Relative rotating speed variation.

n
1.00
0.95
0.90
0.85
0.80
0.75

0.70
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 11. Relative rotating speed variation with recuperation.

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

T4
1.0
0.9
0.8
0.7
0.6

0.5
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 12. Relative exhaust gas temperature variation.

1.0

0.9
0.8
0.7
0.6
0.5
0.4

0.3
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 13. Relative pressure ratio variation with recuperation.

1.0

0.9
0.8
0.7
0.6
0.5

0.4
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 14. Relative temperature ratio variation with recuperation.

1861

1862

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

Gf
1.0

0.8

0.6

0.4

0.2

N
0.0

0.2

0.4

0.6

0.8

1.0

Fig. 15. Relative fuel consumption variation with recuperation.

a stronger eect on the curve of optimum n_ N_ relation. That is to say, with the same pressure
ratio but dierent temperature ratios, the curves are disperser. Reversely the curves are closer. The
upper curves are ones of lower temperature ratios in Figs. 10 and 11. Because relative speed
dierence between simple cycle and recuperative cycle is much smaller, so these denser curves in
Figs. 10 and 11 can be summarized into a representative curve. The tting formula is approximately as follows:
n_ 0:739 0:521N_  0:446N_ 2 0:186N_ 3
For part-load operation, turbine exhaust temperature is an important parameter, especially for
the cogeneration. The variation trend of exhaust gas temperature T_ 4 T4 =T40 is shown in Fig. 12.
Except for the disperse curve s 3:5, the temperature variation curves are close.
For recuperation set, some other curves are derived and shown Figs. 1315. The variation of
relative pressure ratio p_ under part-load optimum variable speed operation are shown in Fig. 13.
For common given design range of p and s, p_ N_ lines are approximately straight lines but
slightly bent downward.
The variation of relative temperature ratio s_ under part-load optimum variable rotating speed
operation is shown in Fig. 14. For common given design range of p and s, s_ N_ lines are not linear,
but focus in relative narrow eld except for s 3:5. The value of zero-load is about 0.450.54.
The variation of fuel consumption (G_ f ) is shown in Fig. 15. The fuel consumption at zero-load
for operation mode of variable rotating speed, is only about 0.2 0.3. However, the fuel consumption at zero-load for constant speed operation is generally about 0.4 [9], which is higher than
that for variable speed operation. Therefore, variable speed operation is benecial to energy
saving even for standby zero-load operation.

6. Conclusion
The performance of single shaft microturbine running at variable speed for both simple cycle
and recuperative cycle is comprehensively analytically studied in this paper for the rst time. With

W. Wang et al. / Applied Thermal Engineering 24 (2004) 18511863

1863

the analytical method, the optimal speed representing highest eciency at part-load is derived. In
addition, the o-design performance of optimum variable speed microturbine at varied design
pressure ratio and temperature ratio is also analyzed.
Through this study some new and important results are obtained. The variable speed operation
mode of microturbine is obviously better than the constant speed operation mode and the general
dierence is given. The eciency improvement of variable speed operation mode for recuperative
cycle is higher than that for simple cycle. For design parameters, the temperature ratio has much
stronger inuence on the eciency and optimal speed than the pressure ratio does. The variation
of eciency versus power output can almost be summarized into single curves. The variable speed
operation has smaller zero-load fuel consumption.

Acknowledgements
This study project is supported by the NKBRSF project of China (G1999022302) and NSFC of
China (No. 50246003 and 59925615).

References
[1] P. Craig, The Capstone turbogenerator as an alternative power source, SAE Paper 970202, 1997.
[2] J. Carne, A. Cavini, L. Linaaki, Micro gas turbine for combined heat and power, ASME Paper 98-GT-309, 1998.
[3] T. Nakajima et al. The development of the micro gas turbine generation, Proc. 1995 Yokohama Intl Gas Turbine
Congress, Japan, 1995, III 301-III 307.
[4] J.H. Watts, Microturbines: A new class of gas turbine engines. ASME International Gas Turbine Institute, Global
Gas Turbine News 1999.01; (1).
[5] L.S. Langston, Gas turbine industry overview. ASME International Gas Turbine Institute, Global Gas Turbine
News, 1999.06; (2).
[6] L.S. Langston, Gas turbine industry overview. ASME International Gas Turbine Institute, Global Gas Turbine
News, 2000.06; (2).
[7] N. Zhang, R. Cai, Analytical solutions and typical characteristics of part-load performances of single shaft gas
turbine and its cogeneration, Proceedings of ECOS99, Tokyo, 1999, 186191;
N. Zhang, R. Cai, Energy conversion & management 43 (912) (2002) 13231337.
[8] N. Zhang, R. Cai, Analytical solutions and typical characteristics of combined cycle part-load performances.
Proceedings of ECOS 2000, Netherlands, 2000, 469480.
[9] N. Zhang, R. Cai, Analytical part-load performance of single shaft small gas turbine and its cogeneration,
J. Aerospace Power 14 (4) (1999) 382386 (in Chinese with English abstract).
[10] R. Cai, A new analysis of recuperative gas turbine cycles. Pro Instn Mech Engrs Part A, J. Power and Energy 212
(4) (1998) 289296.

You might also like