Paper ID: 164, Page 1

IMPLEMENTAION OF A SMALL SCALE ORGANIC RANKINCE

CYCLE TEST BED SYSTEM USING STEAM AS HEAT SOURCE

Muhammad Usman ‡†, Muhammad Imran ‡†, Dong Hyun Lee † and Byung-Sik Park*‡†

‡Korea University of Science and Technology, 217 Gajeong-ro, Yuseong-gu, Daejeon 305-350, South
Korea
[email protected] / [email protected]
Web page: http://www.ust.ac.kr/

†
Korea Institute of Energy Research,
152 Gajeong-ro, Yuseong-gu, 305-343, Daejeon, South Korea
*
Corresponding Authur e-mail: [email protected]
Web page: http://www.kier.re.kr

ABSTRACT

Organic Rankine cycle based power systems are well known for waste heat recovery application due to
their adaptability to follow heat source variations. Industrial exhaust steam has an appreciable potential
for the installation of waste heat recovery units. Korea Institute of Energy Research has developed waste
heat recovery units which can generate power in the range of hundreds of kilowatts. In order to,
rigorously test new cycle configurations and control strategies with least cost for heat source, a small-
scale organic Rankine cycle test bed was implemented which a has steam condensing heat exchanger
for using steam as a heat source in similar configurations as of larger units. The test bed was equipped
with data logging and standalone control system and was configured for the electrical output around
1kW using R245fa as a working fluid. The system is composed of plate type heat exchangers, scroll
type expansion machine, screw type working fluid pump and control valves with actuators. This work
will present the difficulties, solutions and operational results in terms of design, equipment selection,
fabrication and operational experience of system for small-scale power generation with efficiency over
5.2 % for a temperature difference of 120oC. Complexities involved in superheat control of working
fluid for the system powered by steam will also be discussed.

INTRODUCTION

Organic Rankine cycle system (ORC) is the accepted viable technology for low-temperature heat
conversion to electricity (Lecompte et al. 2015). Statistical analysis suggests that low-grade waste heat
accounts for more than 50% of the total heat generated in the industry (Hung et al. 1997). Utilization of
this waste heat energy to harness electrical power output will be a necessity of future to stay within
energy budget allocations, to compete with the growing economies of the world. Low-grade waste heat
power generation has been experimentally tested in many studies, where (Zhang et al. 2014), (Wenzhi
et al. 2013), (Peris et al. 2015) & (Minea 2014) are the few works to mention in low-grade waste heat
recovery works. Most of the published works presented their heat source as hot water or exhaust gas.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 2

Korea Institute of Energy Research has implemented 100kW class ORC systems for low-grade
industrial waste heat. The implementation of a very small-scale ORC system was to be tested for
performance, controllability, and net power output using low-temperature exhaust steam (<135oC), as
a heat source. As this system will be operated for 1kW power output with condensing steam as a heat
source, already published literature rarely contain information for such configuration of a small power
system with such heat source. The implemented test bed system will be used to test thermodynamic
parameters and control strategies by using negligible operational costs as compared to the already
installed 100kW scale system. It is suggested that for a small-scale power generation system (under
5kW) scroll machine is the best choice (Quoilin et al. 2012). Following the previous findings, a scroll
machine is used in this work. The working fluid is an important factor of choice for ORC power system
design, R245fa is known for its safety levels and environmental impacts for waste heat application
(Wang et al. 2011). R245fa is also the best choice based on the criteria of net power and suitable working
pressure (He et al. 2012). Thus, R245fa is selected as the suitable working fluid for the proposed system.
This work describes the design, fabrication and technical difficulties involved in the implementation of
ORC system of such small power out using available components from the market. The test results are
presented with performance evaluation for different degrees of superheating of the working fluid at
expander inlet.
Materials & Method

The system’s operational success is dependent upon its design and implementation. For small scale
systems, it is customary to design the thermodynamic cycle for optimum performance at given source
and sink conditions to obtain required power output and then finding the equipment (expansion machine,
pump, heat exchangers, etc.) from the market, which can operate at required conditions of the power
cycle. Contrarily, for bigger systems, it is common to design the components to strictly fulfil the
performance requirements. For smaller systems, it is not viable to design and manufacture every single
component as per requirements. In that case, components which match closely are selected and can be
operated at slightly off design, compromising on performance. Component sizing & selection is difficult
when the unusually small power requirement and cycle configurations are imposed. The mismatched
components cause a lower overall performance of the system.
The system was designed to be operated by steam in the pressure range of 0~2 bar gauge pressure and
the sink temperature at 12oC was available. Numerous working fluids have been under investigation for
the usage in organic Rankine cycle based system for low-grade waste heat recovery. R245fa was
selected for its closest match of required thermodynamic properties for available heat source and sink
conditions and favorable environmental impact. The design procedure was initiated by fixing the pinch
point temperature difference in heat exchangers (evaporator and condenser). Having fixed the pinching
temperatures allowed to fix evaporation and condensing temperature of the organic Rankine cycle. The
pressure ratio was calculated from evaporation and condensing pressures. Although when optimizing
the performance of the waste heat ORC system, an optimum evaporating temperature exists that
maximizes the output power (or overall efficiency) (Quoilin et al. 2011) but for prototypes and small
systems it is more feasible to follow the equipment specifications. In current case, the expander inlet
pressure for required output was selected as evaporation pressure. Isentropic enthalpy change was
calculated at the calculated pressure ratio. Isentropic efficiency of the scroll type expansion machine
was estimated from published literature (Song et al. 2014). The mass flow rate for the proposed system
can be calculated from Equation (1).
𝑊̇𝑠ℎ𝑎𝑓𝑡 = 𝑚̇. (ℎ1 − ℎ2𝑠 )𝜂𝑖𝑠𝑒𝑛 (1)

Power output for net electric power includes the mechanical losses in expander, coupling and generator.
Also the generator efficiency reduces the generator output. In order to obtain 1kW electric output,
𝑊𝑠ℎ𝑎𝑓𝑡 is estimated to be higher than 1kW. After performing the design calculations for the system, the
scheme was modelled in cycle tempo software (Cycle Tempo) for verification and ease of study for

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 3

various operational parameters. Major thermodynamic parameters at design point are presented in Table
1.
Table 1: Thermodynamic parameters at design point from design calculation and verified by cycle-
tempo software simulation

Parameters Value
Evaporator pressure 12.5bar
Condenser pressure 1.1bar
Expander Isentropic efficiency 50%
Working fluid mass flow rate 0.051kg/s
Heat source temperature (saturated steam) 120oC
Sink Temperature 16oC
Sink mass flow rate 0.73kg/s
Expander inlet temperature 101.5oC
Condenser subcooling 3oC

Table 2: Details of Equipment selected for implementation of test bed

Sr.
Equipment Details
no
Scroll Type expander (Airsquared Inc.), 12cc/rev, 1kW Nominal output, Magnetic coupling linked
1 Expander
to 60Hz Generator 110V AC generator

Braze plate heat exchangers (Janghan Engineers, Inc.) 60 & 50 plates, area of 6.5 & 5.38 square
2 Evaporator & Condenser
meter for evaporator & condenser, respectively

Screw type pump (Tuthill Pump Company). Displacement of 2.6ml/rev Magnetic coupling linked to
4 Feed Pump
0.75kW 3 phase 380V motor.
Condenser
Electric Bulbs of various power ratings were connected to the generator using single push single
throw (SPST) type relays. Bulbs for 110V, were used in different power rating combination. 200W,
5 Load Bank
100W, 60W, 30W watt bulbs are used to obtain various step of resistive load connected to the
generator

6 Heat Sink Cooling tower already installed in facility

7 Heat Source LNG based boiler with steam generation capacity of 500 kg/hr maximum

Working fluid with Coriolis type flow meter, sink flow rate measurement with electromagnetic flow
8 Flow rate meters
rate meter, sink flow rate measurement with vortex type flow meter

9 Valves Butterfly valves and globe valve for steam flow control (actuator controlled)
T -type thermocouples for temperature measurement & RTD (PT100) for evaporator and expander
10 Temperature Measurement
inlet/exit
11 Pressure Transduces Pressure Transducers with range of 0~16bar

12 Data Acquisition & Control NI - cRIO 9074 with 8 slot integrated controller, NI I/O C-series model and Labview software

Table 2 presents the detail of selected equipment for test bed construction. Very few off the shelf
expansion machines are available in the market, which can be used directly without alteration, but in
this work a scroll machine designed for ORC application was available. Pump selection is another tough
choice because of low flow rate, and large pressure head, no suitable centrifugal machine was found. A
screw pump was able to cater our requirement with confidence as far pressure and flow rate
requirements were required, but resulted in poor efficiency (<30%). Load bank was made by connecting
electric bulbs of various power ratings to generator circuit and connections were controlled by relay
operation.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 4

Figure 1 presents the piping and instrumentation diagram for the test rig which was constructed
according to the CAD diagram of system presented in Figure 2. Figure 3 presents the test rig in final
fabricated form and under organic Rankine cycle system operation for power output.

Figure 1: P&ID Scheme for proposed system

Expander
Generator

Condenser

Evaporator

Receiver Tank

Pump Coriolis flow

meter

Figure 2: Isometric view of implemented system Figure 3: System under operation

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 5

Results & Discussion

Figure 4 presents organic Rankine cycle test bed operational parameters for 20 minutes of its operating
time. The system was operated in a way that the expander speed was kept below 3600RPM to follow
safety recommendations from the manufacturer. Electric bulbs were connected to the load circuit one
after another until a total of load of 1060W was imposed to the electric generator. After increasing the
electric load, the pump speed controlled by variable frequency drive was increased to increase the mass
flow rate and increase the evaporator pressure which increase the expander speed. After attaining the
expander speed around 3600RPM, next electric bulb was connected which would reduce the expander
speed as the torque on generator increased and for the same power output the rotational speed was
decreased. Increasing the mass flow rate further, increased the evaporator pressure which allowed
higher power output from the expander. The Figure shows that the evaporator pressure follows the
pump speed (Mass flow rate). Generator power output increased to a maximum value of 1020W, when
evaporator pressure is around the maximum operating design point of 11.5 barg. Expander speed was
3400 RPM while the generator was connected to a circuit of 8 bulbs (200W x 4, 100W x 2, 30W x 2)
rated at 110V which summed up to yield 1060W resistive load.

Figure 4: Organic Rankine cycle system operational parameters for different power output range

The sharp spikes in generator power output data occurred at the same time when expander speed was
reduced at constant pump speed, increasing the imposed torque reduced the expander speed for the same
power output. This was due to the connection of the relay with bulb. When the applied load increased
as a step change the expander speed was reduced, to get back to recommended expander speed, pump
speed was carefully adjusted to carry higher loads.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 6

Figure 5: Electric Generator Output and Net Output Figure 6: Electric Generator Power
vs Expander Inlet Pressure Output power vs Expander Speed

Figure 5 presents the electric output of generator with respect to the evaporator pressure while condenser
conditions were held constant at 0.3barg. The figure also presents the net electric power output from
the system. The pump electric power consumption was measured and used for evaluation. Both data
plots are fitted with a 2nd order polynomial for analysis. Figure 6 presents generator output at various
electric loading conditions. An applied load of 1000W refers to the connected electric bulbs with a
combined power rating of 1000W to the electric generator circuit.

Figure 7: Variation of degree of super heat of working fluid at evaporator exit with respect to steam
(heat source) pressure at heat source inlet side.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 7

Figure 8: Effect of degree of super heat of working fluid at expander inlet vs electrical efficiency for
same power output
An experiment to investigate the effects of superheat of the working fluid at expander inlet was
performed. Extensive research and theoretical background suggested the lesser the degree of
superheating, the better would be the performance for a system without recuperator using R245fa
working fluid. In reality to maintain, superheat value of 0 would be ideal but is very challenging task to
maintain a stable working fluid output with all vapor exit quality without any superheating. The ability
to maintain the minimum superheating degree depends on various equipment and control characteristics.
Evaporator type, size& geometry, plays an important role in this regard. In case of plate heat exchanger,
the number of channels, aspect ratio, port design & thermal inertia influence the performance in such a
way that to obtain a superheat of 0 is very difficult. The heat source type is also an important factor
when considering the controllability of superheating of the working fluid at evaporator exit. In case of
current system the heat source is steam, which is condensed in evaporator and subcooled to certain level.
Control of heat source was performed by using a globe valve which has been carefully sized and custom
manufactured for the current conditions of steam pressure & flow rates. The actuator controlled valve
allowed control of heat source for a steam requirement of less than 40 kg/hr. Figure 7 presents the
controllability of working fluid superheat at evaporator exit. The globe valve which has an upstream
steam pressure of 1.95 ~ 2barg was opened in such a way that the downstream pressure of globe valve
was increased. As a result, condensing pressure in evaporator (heat source side/steam) was increased.
Superheat at evaporator exit (Working fluid side) was observed as 10OC when source pressure was
0.4barg, when steam pressure was increased to 1.1barg the super heat value was observed around 21OC.
While the source temperature was observed around 133OC in both cases. Figure 8 presents superheating
degree at expander inlet and cycle efficiency for same electrical power output. Figure 7 and Figure 8
present data from same experiment, but the Figure 8 presents superheat value at expander inlet while
Figure 7 presents the superheat value for evaporator exit. The difference in super heat values is due to
heat loss in piping network when the working fluid flows from evaporator to expander. The data in
Figure 9 and 8 is used to plot 2nd order polynomial fits for analysis. Electrical efficiency of system was
obtained by equation (2).
𝐸𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝐺𝑒𝑛𝑒𝑟𝑎𝑡𝑜𝑟 𝑃𝑜𝑤𝑒𝑟 𝑂𝑢𝑡𝑝𝑢𝑡 (𝑘𝑊)
𝜂𝑒𝑙𝑒𝑐 = 𝑄̇𝑒𝑣𝑎𝑝𝑜𝑟𝑎𝑡𝑜𝑟. 𝑤𝑜𝑟𝑘𝑖𝑛𝑔 𝑓𝑙𝑢𝑖𝑑 (𝑘𝑊)
(2)

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 8

Where

𝑄̇𝑒,𝑤𝑓 (𝑘𝑊) = 𝑚𝑤𝑓

̇ (ℎ𝑒𝑜 − ℎ𝑒𝑖 ) (3)

Figure 8 data results used to obtain 2nd order polynomial relation which suggest the drop in efficiency
of system with increasing super heat value. The relation is presented as Equation (4)

𝜁 = 5.55084 − 0.00542(𝜃) − 0.0005245(𝜃)2 (4)

Where
𝜁 − 𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 (𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒) 𝑎𝑠 𝑎 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑜𝑓 𝑠𝑢𝑝𝑒𝑟ℎ𝑒𝑎𝑡
𝜃 − 𝑖𝑠 𝑑𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑠𝑢𝑝𝑒𝑟ℎ𝑒𝑎𝑡 (𝐶𝑒𝑙𝑐𝑖𝑢𝑠) 𝑎𝑡 𝑒𝑥𝑝𝑎𝑛𝑑𝑒𝑟 𝑖𝑛𝑙𝑒𝑡
Equation 4 can be used to estimate the reduction in efficiency with respect to superheating degree for
the same electrical power generated. In this case power output was held constant by maintaining a
constant expander speed to hold power output around 930W and heat source carefully adjusted to obtain
different superheat values. It was noted that for current system an increase in 1oC of super heat will
reduce 0.011% in electric efficiency if evaporator superheat is around 10oC. As the relation is nonlinear
curve it suggests reduction in electric efficiency of 0.025% with increase of 1oC after super heat at heat
exchanger exit is above 20oC. The relation is only valid for current system and is a basis for the future
research work to be carried out to establish relationship for efficiency and superheating degree for
various systems of various power range.

Conclusion

In this work, design, equipment selection and fabrication of a small scale organic Rankine cycle test
bed system, using steam as heat source was performed. The fabricated test rig was operated and
following conclusion were made:

 1kW scale organic Rankine cycle system can be designed and fabricated with major
components available from market.
 Design requirement of components and availability causes mismatching in component selection,
whereas losses (isentropic, frictional, electrical etc.) involved in components should be
carefully analyzed to achieve required performance.
 A 1kW scale ORC system is capable of producing net electrical output if components are
precisely selected
 The system was able to produce 1kW electric output with efficiency value of 5.2%.
 The best efficiency points are not at the highest output power.
 Using a steam condensing evaporator for small flow rates of steam and working fluid superheat
control at constant pressure is difficult procedure.
 Efficiency depreciation with increasing superheat level can go as high as 0.02% per degree
Celsius for this system.
 An attempt to establish a practical approach for relating efficiency depreciation with higher
superheat is initiated. The future work will incorporate various types of systems at various
power range will be studied for efficiency drop due to superheat to highlight how much research
attention and work will be required for improvement of control systems and heat exchangers to
obtain minimum super heat levels.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 9

NOMENCLATURE
ORC Organic Rankine Cycle
𝑊̇𝑠ℎ𝑎𝑓𝑡 Expander shaft power output

𝑚̇𝑤𝑓 Working fluid mass flow rate (R245fa)

ℎ𝑒𝑜 Working fluid enthalpy at evaporator outlet
ℎ𝑒𝑖 Working fluid enthalpy at evaporator inlet
𝜂𝑖𝑠𝑒𝑛 Expander Isentropic efficiency
𝜂𝑒𝑙𝑒𝑐 Thermal to electric conversion efficiency
𝑄̇𝑒,𝑤𝑓 Heat transfer to the working fluid

REFERENCES

Cycle Tempo. TU Delft (Delft University of Technology), http://www.asimptote.nl/

He, C. et al., 2012. The optimal evaporation temperature and working fluids for subcritical organic
Rankine cycle. Energy, 38(1), pp.136–143.

Hung, T.C., Shai, T.Y. & Wang, S.K., 1997. A review of organic rankine cycles (ORCs) for the
recovery of low-grade waste heat. Energy, 22(7), pp.661–667.

Lecompte, S. et al., 2015. Review of organic Rankine cycle (ORC) architectures for waste heat
recovery. Renewable and Sustainable Energy Reviews, 47, pp.448–461.

Minea, V., 2014. Power generation with ORC machines using low-grade waste heat or renewable
energy. Applied Thermal Engineering, 69(1-2), pp.143–154.

Peris, B. et al., 2015. Experimental characterization of an ORC (organic Rankine cycle) for power and
CHP (combined heat and power) applications from low grade heat sources. Energy, 82, pp.269–
276.

Quoilin, S. et al., 2011. Thermo-economic optimization of waste heat recovery Organic Rankine
Cycles. Applied Thermal Engineering, 31(14-15), pp.2885–2893.

Song, P. et al., 2014. A review of scroll expanders for organic Rankine cycle systems. Applied
Thermal Engineering, 75, pp.54–64.

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium
 Paper ID: 164, Page 10

Wang, E.H. et al., 2011. Study of working fluid selection of organic Rankine cycle (ORC) for engine
waste heat recovery. Energy, 36(5), pp.3406–3418.

Wenzhi, G. et al., 2013. Performance evaluation and experiment system for waste heat recovery of
diesel engine. Energy, 55, pp.226–235.

Zhang, Y.-Q. et al., 2014. Development and experimental study on organic Rankine cycle system with
single-screw expander for waste heat recovery from exhaust of diesel engine. Energy, 77,
pp.499–508.

ACKNOWLEDGEMENT
This work was conducted under the framework of Research and Development program of the Korea
Institute of Energy Research and University of Science & Technology (b4-5524).

3rd International Seminar on ORC Power Systems, October 12-14, 2015, Brussels, Belgium