Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Aerodynamic Characteristics of A Centrifugal Compressor Working in Supercritical Carbon Dioxide 2012 Energy Procedia

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com
Energy
Procedia
Energy Procedia 00 (2011)14000–000
Energy Procedia (2012) 1149 – 1155
www.elsevier.com/locate/procedia

2011 2nd International Conference on Advances in Energy Engineering (ICAEE 2011)

Aerodynamic Characteristics of a Centrifugal Compressor


Working in Supercritical Carbon Dioxide
Motoaki Utamuraa,*, Taro Fukudab, Masanori Aritomia
a
Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan
b
Daiichi System Engineering Co. Ltd., 1-16-30 Meiekiminami, Nakamura-ku, Nagoya 450-0003, Japan

Abstract

Development of a closed cycle gas turbine with supercritical carbon dioxide as a working fluid is underway to
generate power from low-range or intermediate-range waste heat sources. A demonstration test using a reduced scale
turbomachine was conducted and the aerodynamic characteristics of a compressor were examined. A compressor was
selected as centrifugal with 30 mm outer diameter and rated rotational speed of 1.7 kHz and mass flow rate of 1.2
kg/s. To reduce compression work, the operating condition at the inlet to compressor was chosen in a supercritical
state close to the critical point of 7.38 MPa, 304 K where the compressibility coefficient z becomes markedly small
and real gas effect dominant. The experimental range in terms of z is 0.16<z<0.6. The measured pressure ratio and
adiabatic efficiency of the compressor are compared with calculations conducted using the Meanline method. The
compressor performance in the supercritical liquid-like phase becomes highest in the experiment, which is well
simulated by the method. The calculated pressure ratio shows excellent matching with experimental data in the
supercritical liquid-like phase. However overestimation is recognized at the off-design point in the supercritical gas-
like or subcritical region. Experiments also show that the compressor performance improves with reduction of the
compressibility coefficient, which the Meanline method has well predicted.
© 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the organizing committee
of©2nd
2011 Published Conference
International by Elsevieron
Ltd. Selection
Advances and/or peer-review
in Energy under responsibility
Engineering (ICAEE). of [name
Open access under organizer]license.
CC BY-NC-ND

Keywords: supercritical state, carbon dioxide, centrifugal compressor, turbine, power cycle

* Corresponding author. Tel.: +81357343293; Fax: +81357343293.


E-mail address: utamura@nr.titech.ac.jp.

1876-6102 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the organizing committee of 2nd International
Conference on Advances in Energy Engineering (ICAEE). Open access under CC BY-NC-ND license.
doi:10.1016/j.egypro.2011.12.887
21150 Motoaki Utamura
Author name et al.\Procedia
/ Energy / Energy00
Procedia
(2011) 14 (2012) 1149 – 1155
000–000

1. Introduction

Gas in a supercritical state close to the critical point might engender the reduction of compression
work compared to ideal gas because of the real gas effect. Moreover, carbon dioxide has a critical point at
around room temperature. Therefore, it is possible to configure an efficient gas turbine power cycle that
functions even at a low or intermediate temperature range. Theoretical studies were reported in the
literature of the late 1960s [1, 2].
The cycle would facilitate power conversion from unused energy in a low-temperature region such as
waste heat from industry or renewable energy. Because the power cycle is a closed regenerative one, it
necessitates unusually high efficiency at the regenerative heat exchanger involved in the cycle. Then
realization of the power cycle necessitates a high-performance compact heat exchanger with high
temperature and also with high mechanical strength that is resistant to supercritical pressure up to 20 MPa.
To meet those requirements, Tokyo Tech developed an X-shaped fin microchannel heat exchanger that is
made of plural metal plates with flow channels engraved and integrated by diffusion bonding [3].
Based on these fundamental studies, compressor tests were conducted, confirming stable operation
close to the critical point [4]. Then, a demonstration test plant of 10 kW power cycle was built [5] and
implemented. Continued power generation of 200 W was realized [6]. The cycle compressor plays a vital
role in enhancing cycle thermal efficiency. To this end, it operates in a supercritical state near the critical
point of carbon dioxide with critical pressure of 7.38 MPa and critical temperature 304 K, where thermal
properties change considerably according to a small change of thermodynamic state. This paper describes
the performance characteristics of a centrifugal compressor. Applicability of the Meanline aerodynamic
design method [7] to widely various operations from a subcritical to a supercritical state was discussed.
2. Experiment
2.1 Supercritical CO2 power cycle
Figure 1 presents a carbon dioxide phase diagram. The critical point at which a phase change
disappears is located at pressure 7.38 MPa and temperature 305 K. The whole of the present power cycle
is to be formed in a supercritical state so that the state at the compressor inlet should be located as close as
possible to the critical point. That is true because compressor work is reduced most there, which would
result in enhanced thermal efficiency of the power cycle. The thermodynamic rationale of selecting
carbon dioxide as the working fluid is that the critical temperature exists near room temperature, which
benefits easy cooling and effective utilization of degraded heat as a heat source. The typical configuration
of the cycle is presented in Fig. 2, called a closed regenerative Brayton cycle. It was practiced in the
present demonstration test. Working fluid recirculates the whole cycle, undergoing compression, heating,
expansion, cooling, and also generating electricity.

Fig. 1 Supercritical power cycle in phase Fig. 2 Schematic of test facility


diagram of carbon dioxide
Author name / Energy Procedia 00 (2011) 000–000 3
Motoaki Utamura et al.\ / Energy Procedia 14 (2012) 1149 – 1155 1151

2.1 Test apparatus


Figure 3 portrays a cross section of a rotor, which comprises a permanent magnet, a centrifugal
compressor, and a radial inflow turbine. A compressor and a turbine are aligned co-axially back-to-back
to reduce net thrust. A synchronous motor/generator was driven by an inverter up to 60 000 rpm (1 kHz).
Because the rated speed of the rotor was designed as 100 000 rpm (1.7 kHz), CO2 gas bearings were
adopted as journal and thrust bearings. Differential pressure acting on the wheel was estimated as around
4 MPa. Because the thrust is too great for a tapered-land thrust gas bearing to withstand, some
compressed gas was introduced between the thrust bearing and the thrust collar to withstand it. The gas
passed around the rotor to reach the back face of the turbine wheel to be sucked at the turbine impellor
inlet. Incidentally it functioned to reject heat from the electromagnetic loss of the permanent magnet as
well as from windage loss i.e. fluid shear around the rotating cooling rotor. The highest allowable rotor
temperature was 180°C because the existence of high-density CO2 windage loss in a typical operating
condition was estimated as 5 kW: 100 times more than that of atmospheric air. Pictures of fabricated
turbomachines are presented in Fig. 4. Outer diameters of the aluminum compressor and titanium turbine
are 30 mm and 35 mm, respectively. The blade height is 1 mm at the compressor exit, where the diffuser
is connected with the collector. All compressor-related measurements were made at external piping that
was the inlet for total temperature and total pressure, and the exit for static pressure and temperature.
Table 1 shows basic design specifications. Compressor inlet operating conditions were determined to
account for both aspects of seeking minimum possible compressibility coefficient and avoidance of liquid
formation at the mouth of the compressor, where flow is accelerated. A tradeoff study showed tolerance
of 4 K. Given this and the turbine inlet temperature of 550 K and compressor outlet pressure of 12 MPa,
the compressor inlet pressure was determined by cycle calculation to obtain maximum thermal efficiency.
The CO2 recirculating flow rate was 1.2 kg/s necessary to give 10 kW power output. Aerodynamic
performance was designed to become maximum at the rotational speed of 1.7 kHz.
2.2 Measurement results
Unexpected large windage loss prevented the system from reaching self-sustaining operation. Further
access to the critical point at the compressor inlet condition finally led to continuous power generation.

Table 1 Major design


specifications
Compressor Stator Turbine

CO2 Flow

Thrust
bearing

Water Thermal
jacket Rotor insulator

Fig. 3 Cross section view of a rotor Fig. 4 Pictures of impeller blades


41152 Motoaki Utamura
Author name et al.\Procedia
/ Energy / Energy00
Procedia
(2011) 14 (2012) 1149 – 1155
000–000

Table 2 Work balance

Generation (kW) Consumption (kW)


Turbine 17.4 Compressor 10.6
Windage 5.2
Load 0.1
Miscellaneous 1.5
Sum 17.4 Sum 17.4

Fig. 5 Trend of parameters

Figure 5 exhibits a time history of parameters during power generation. At the time of 945 s, the
rotational speed of 1.17 kHz produced 1.11 kg/s compressor mass flow rate and pressure ratio of 1.44 at
turbine inlet temperature of 264.9°C (538 K). Most are close to design values. However, the power output
was 206 W at most. Work balance showed that this arose primarily from windage, as shown in Table 2.

3. Meanline analysis of compressor aerodynamics

3.1 Results

Meanline calculations were conducted based on the experimental data. In general, this calculation
predicts performance in each component section along the mean streamline using empirical modeling
parameters. Refining such parameters to match measurement data and thereby predict performance with
high accuracy is an important task. In this task, the slip factor and impeller relative Mach number ratio
(MR2) were investigated as the major impeller modeling parameters evaluated at the design point. MR2
was defined by the ratio of the relative Mach number at the impeller inlet tip and impeller exit (primary
zone [7]). This task was executed using the Meanline program “COMPAL” developed by Concepts ETI
Inc. The slip factor of 0.9 and MR2 of 1.21 were selected as design points. Figure 6 shows that the
predicted efficiencies lie in the band due to temperature measurement error of ∓0.1K, although the
prediction of pressure ratio agreed well with measurements in widely differing mass flows. Red color
shows the condition where model tuning was made.
0.190<z<0.237

Fig. 6(a) Pressure ratio vs. design mass flow ratio, (b) Efficiency vs. design mass flow ratio
Motoaki Utamura
Author nameet/ al.\ / Energy
Energy Procedia
Procedia 14 (2012)
00 (2011) 1149 – 1155
000–000 1153
5

The occasion of the efficiency difference between


prediction and experiment is regarded as deriving
from the fact that the compressor exit
thermodynamic state was located near the pseudo-
critical line, where, as portrayed in Fig. 7, a small
temperature difference produces a large exit
specific enthalpy change. Because of both the
measurement error of the compressor exit
temperature and limitation of accuracy of Meanline
analysis, further improvement of the performance
prediction at the design point remains as a future
design issue. Figure 8 presents the off-design
performance curve, which is divided to four Fig. 7 p–h diagram in supercritical
regions: a subcritical region (z=0.49–0.6),a
supercritical gas-like region (z=0.3–0.47),a supercritical liquid-like high pressure region (z=0.253–
0.29),and a supercritical liquid-like low pressure region (z=0.215–0.251) in which compressibility
coefficient z is defined as z=pv/RT, where p stands for pressure, v signifies the specific volume, R denotes
the gas constant, and T represents temperature. Actually, v was calculated using the thermophysical
property library PROPATH [8]. Results show that the pressure ratio prediction tends to depart from
experimental values as the thermodynamic state moves apart from supercritical to subcritical as the mass
flow rate decreases. However, the efficiency shows only a very slight change with the flow rate.

3.2 Discussion

The reason for the discrepancy between calculations and experiments in Fig. 8a is regarded as
follows. The fluid gains specific work through an impeller passage, as expressed by Euler’s equation :
U 22 − U 12 C 22 − C12 W 22 − W12 (1)
E= + −
2 2 2
In that equation, U signifies the impeller tip speed with sufficies 1 and 2 denoting inlet and exit
respectively, C denotes the absolute velocity, and W stands for the relative velocity in a rotational
coordinate in which static pressure rise depends on the first and third terms. The first term signifies the
centrifugal force to raise static pressure at the exit. With an increased amount of the compressibility
coefficient, it was observed that W2 is larger than W1, so that the third term in Eq. (1) became negative.
This caused a decrease in MR2, which means that the impeller relative velocity at the exit is greater than
that at the inlet tip. Consequently, the pressure ratio in the subcritical phase (MR<1) is smaller than that in
the supercritical liquid-like phase (MR>1).

55000rpm 55000rpm

Fig. 8(a) Pressure ratio vs. design mass flow ratio, (b) Efficiency vs. design mass flow ratio
61154 Motoaki Utamura
Author name et al.\Procedia
/ Energy / Energy00
Procedia
(2011) 14 (2012) 1149 – 1155
000–000

Next, it must be discussed whether the compressibility coefficient affects aerodynamic performance
or not. Figure 9 presents an evaluation result of the compressibility coefficient under two constant flow
coefficients. Both the pressure ratio and efficiency decreased concomitantly with increased
compressibility coefficient z. This fact implies that the compressor performance is enhanced in a
supercritical state.

Fig. 9(a) Pressure ratio vs. design mass flow ratio, (b) Efficiency vs. design mass flow ratio

4. Conclusion
Verification testing of a gas turbine with supercritical fluid of carbon dioxide was conducted. Also
conducted were refinement and evaluation of the performance using the Meanline program. The
compressor performance in the supercritical liquid-like phase achieved the highest performance in both
experiments and predictions by the simulation program. The Meanline program also predicted better
matching with experimental data in the supercritical liquid-like phase. However overestimation of the
pressure ratio develops as a thermodynamic phase of carbon dioxide shifts from supercritical phase
through subcritical phase at compressor inlet. Experiments also showed improvement of compressor
performance with reduction of compressibility coefficient, which the Meanline program well simulated.
Acknowledgements

Work partly performed under the program of “Strategic Development of Energy Conservation
Technology Project” and “Development of Gas Turbine with Super-critical Carbon Dioxide as Working
Fluid,” sponsored by NEDO.
References

[1] Angelino G. Real gas effects in carbon dioxide cycles, ASME Paper No. 69-GT-103, (1969) American Society of
Mechanical Engineers.
[2] Frutschi HL. Closed-cycle gas turbines, ASME Publications; 2005, p. 164–9.
[3] Utamura M, Kajita R, Yamamoto T, Hasuike H. Development of Supercritical Carbon Dioxide Closed Cycle Gas
Turbine, 3rd Report: Development of Regenerative Heat Exchangers (in Japanese), Proc. 39th Gas Turbine Periodical
Symposium, C-9 (Matsumoto, Japan); 2011.
[4] Aritomi M, Ishizuka T, Muto Y, Tsuzuki N. Performance test results of the supercritical CO2 compressor for a new gas
turbine generating system, Proc. 18th Intn’l. Conf. on Nuclear Engineering, ICONE18-29371, Xi’an China; 2010.
[5] Utamura M, Hasuike H, Yamamoto T. Demonstration Test Plant of Closed Cycle Gas Turbine with Supercritical CO2 as
Working Fluid. Journal of Strojarstvo 2010; 52 No. 4: 459–65.
[6] Hasuike H, Ogawa K, Utamura M, Yamamoto T, Fukushima T. Development of Supercritical Carbon Dioxide Closed
Cycle Gas Turbine, 4th Report: Results of Power Generation Experiments in Bench Scale Test Facility (in Japanese), Proc. 39th
Periodical Symposium of Gas Turbine Society of Japan C-3 (Matsumoto, Japan); 2011.
Motoaki Utamura
Author nameet/ al.\ / Energy
Energy Procedia
Procedia 14 (2012)
00 (2011) 1149 – 1155
000–000 1155
7

[7] Japikse D. Centrifugal compressor design and performance, Concepts ETI Inc.; 1996.
[8] PROPATH group, PROPATH: A Program Package for Thermo-Physical Properties of Fluids, Version 10.2,
http://gibbs.mech.kyushu-u.ac.jp/~akasaka/propath_man/,1990.

You might also like