Cryogenics 70 (2015) 76–84

Contents lists available at ScienceDirect

Cryogenics
journal homepage: www.elsevier.com/locate/cryogenics

The measurement of thermodynamic performance in cryogenic

two-phase turbo-expander
Lu Niu, Yu Hou ⇑, Wan Sun, Shuangtao Chen
State Key Laboratory of Multiphase Flow in Power Engineering, Xi’an Jiaotong University, Xi’an 710049, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Liquid fraction measurement in cryogenic two-phase flow is a complex issue, especially for an industrial
Received 22 February 2015 cryogenic system. In this paper, a simple thermal method is proposed for measuring the liquid fraction in
Received in revised form 30 April 2015 cryogenic two-phase turbo-expander by an electric heating unit in experimental study. The liquid frac-
Accepted 10 June 2015
tion of the cryogenic two-phase flow is determined through the heat balance built at the outlet of the
Available online 18 June 2015
turbo-expander (inlet of heating unit) and the outlet of the heating unit. Liquid fractions from 1.16% to
5.02% are obtained from five two-phase expansion cases. Under the same turbo-expander inlet pressure
Keywords:
and rotating speed, five superheated expansion cases are tested to evaluate the wetness loss in two-phase
Cryogenics
Radial-inflow turbo-expander
expansion. The results show that the proposed method is successful in measuring the liquid fraction of
Two-phase expansion cryogenic two-phase expansion for turbo-expander in an industrial air separation plant. The experimen-
Liquid fraction measurement tal isentropic efficiency ratio and the tested Baumann factor decrease with the increasing mean wetness.
Baumann rule Based on prediction of Baumann rule, the cryogenic turbo-expander with low liquid fraction in two-phase
Efficiency analysis expansion cases suffers from more severe wetness loss than that with the higher liquid fraction.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction pressure and temperature of working flow are independent param-

eters of each other, i.e. the thermodynamic state parameters such
Turbo-expander is well known as one of the most significant as specific enthalpy, specific entropy of gaseous fluid are deter-
devices to generate cooling capacity through isentropic expansion mined by both pressure and temperature. However, in two-phase
in high rotating speed. It is characterized by very high speed, light region, the pressure and temperature are one-to-one correspon-
weight and high efficiency and is widely used in almost every mod- dence and no longer independent parameters of each other. The
ern air separation plants. More cooling capacity and higher lique- working flow thermodynamic state such as specific enthalpy,
fied fraction can be provided to the host cryogenic system when specific entropy and mass liquid fraction cannot be determined
a turbo-expander operates into two-phase region, which could from its pressure and temperature. Therefore, one more indepen-
lead to the improvement of system efficiency through lower the dent parameter besides pressure or temperature is needed for liq-
system power consumption. Increasingly more designers have uid fraction determination of two-phase flow.
strived to utilize two-phase expansion turbo-expander in cryo- Since heating method is successfully applied in operating tur-
genic system. In 1990, a miniature wet turbo-expander for a bine by Moore in British Central Electric Research Laboratory
helium liquefier has been developed by Sixsmith et al. [1]. In the (CERL) in 1976, condensation method, throttling method etc. were
same year, a performance tests on a cryogenic turbo-expander developed for wetness measurement [3]. For cryogenic
operated in two-phase region for a large scale helium cryogenic radial-inflow turbo-expander, Ardashev and Plachendovskii [4]
system are conducted by Kato [2]. conducted the research on cryogenic turbo-expander operated in
Liquid fraction (wetness) measurement is crucial in perfor- two-phase regime, liquid fraction in the outlet of turbo-expander
mance analysis of a two-phase expansion turbine. Therefore, as reached 8–14% measured by an electric heater in 1984, but the liq-
rapid gaseous flow experiences phase transition into two-phase uid fraction measurement equipment and principle was not
region, determination of flow thermodynamic state becomes a described particularly. Optical method is another option for wet-
focused point of vital concernment. In superheated region, the ness measurement. Based on the Mie scattering from liquid dro-
plets, diameter and distribution droplets can be measured, which
has been proved by many successful experiments. In 2009, Cai
⇑ Corresponding author. and Niu [5] experimentally studied the properties of the wet steam
E-mail address: [email protected] (Y. Hou).

http://dx.doi.org/10.1016/j.cryogenics.2015.06.002
0011-2275/Ó 2015 Elsevier Ltd. All rights reserved.
 L. Niu et al. / Cryogenics 70 (2015) 76–84 77

Nomenclature

T temperature (K) g isentropic efficiency

p pressure (MPa) r error
h Enthalpy (J kg1)
s entropy (J kg1 K1) Subscripts
U voltage (V) in turbo-expander inlet
I electric current (A) out turbo-expander outlet
D diameter (mm) he electric heater outlet
P electric power (W) t two-phase expansion test
Q heat (J kg1) s superheated expansion test
m mass flow rate (kg s1) h heating dissipation test
y liquid fraction

flow in the low pressure 300 MW direct air-cooling steam turbine main categories, namely the thermodynamic loss, the mass flow
by using an optical-pneumatic probe. In 2014, Schatz and Eberle loss and the mechanical loss, according to the evaluation principle
[6] researched both steam quality and the droplet size spectrum of wetness loss first developed by Gyarmathy and classification of
which measured by using a miniature combined optical/pneumatic wetness losses introduced by Moore, Guo [17], Laali [18] and
probe in last stages of low pressure steam turbines. Kawagishi [19].
The occurrence of super-saturation and spontaneous condensa- In mathematical analysis for radial-inflow turbo-expander per-
tion in a rapid gas flow expanding into two-phase region has been formance in wetness condition, Aungier [20] represented an
researched for many years. Since the classical nucleation theory empirical treatment for liquid phase by assuming the liquid dro-
(CNT) was formulated in the early 20th century by Volmer and plets are ineffective in transferring work because it cannot follow
Weber et al., experimental studies of nucleation have primarily the vapor phase streamlines. Another theoretical method is pro-
focused on water. In cryogenic field, following the study on nitro- posed by Obata to research the wetness effect on two-phase
gen condensation in a hypersonic wind tunnel that conducted by turbo-expander performance based on a large helium refrigeration
Faro in 1952 [7], numerous successful experiments on the limits system [21].
of super-saturation (Wilson Line) and nucleation rate with phase As reference to ASME PTC6-2004 for steam turbine, wetness can
transition of argon and nitrogen by using supersonic nozzle [8], be determined from pressure and enthalpy, and heat balance is an
shock tube [9,10] and nucleation pulse chamber [11,12] were applicable method to measure the wetness. In wet
carried out. turbo-machinery performance tests, heating method is always
As spontaneous condensation takes place in the flow path in a firstly applied in wetness measurement. Moreover, in these cryo-
turbo-expander, liquid phase presence could lead to two main genic gas expansion and spontaneous condensation experiments
problems. The first problem is mechanical erosion caused by the [8–12], there are no external disturbances such as secondary flow
long-term high velocity impact of liquid particles on or corner flow as well as boundary layer which will occur in
turbo-expander blades. The impeller blades can be roughened, pit- turbo-expanders. So, taken the complicated cryogenic condition
ted or even mutilated. So, it is normal to set an acceptable wetness into consideration, the advantages of simple measuring principle
limit in both wet steam turbines and cryogenic turbo-expander and high reliability of heating method is preferred to other meth-
design. Secondly, it was found that the isentropic efficiency of ods in cryogenic turbo-expander performance tests. In this article,
steam turbines or turbo-expander operated in two-phase region a simple thermal method based on heat balance method is pro-
was considerably lower than that in superheated region. In wet posed for measuring the liquid fraction in cryogenic two-phase
steam turbine, Baumann rule was established as early as 1910 turbo-expander. An experimental test part is designed and inte-
for evaluating the wetness loss in wet steam turbine, which indi- grated into the pipeline after the two-phase turbo-expander in
cated that 1% of mean wetness was likely to cause about 1% reduc- an industrial air separation plant. With an electric heating unit,
tion of dry isentropic efficiency [13]. Corresponding to the original the two-phase flow out of the turbo-expander is heated into super-
Baumann factor aB = 1, a considerable variation of tested Baumann heated state. And then based on thermal equilibrium relationship
factor (0.4–2.0) was observed based on a large number of experi- and take the heat loss along the test part into account, the liquid
ments in reaction and impulse type turbines [14]. However, this fraction of the cryogenic two-phase flow can be determined
conveniently applicable theory can provide little fundamental through the enthalpy and pressure at the outlet of the
knowledge of wetness loss origin. For a better understanding of turbo-expander and the outlet of the heating unit. Liquid fraction
condensing flow in steam turbine, Gyarmathy [15] illustrated the from 1.16% to 5.02% is obtained from five two-phase expansion test
detailed relaxation process, specifically, spontaneous condensation cases 1t–5t (arranged according to the increasing liquid fraction).
takes place if the rapid steam expands to a certain level of In order to evaluate the wetness loss, five superheated expansion
sub-cooling, and the latent heat is released from droplets to processes, 1s–5s, are tested under identical turbo-expander inlet
sub-cooling vapor and brings the steam back to equilibrium condi- pressure and identical rotating speed corresponding to test cases
tion. In 1972, Miller and Schofield [16] published the tested 1t–5t, respectively. Compared with Baumann rule which indicated
Baumann factors for different mean wetness conditions in a wet that 1% growth in liquid fraction leads to 1% efficiency reduction
steam turbine through experimental tests. They found that the for the two-phase expansion process in a steam turbine, the isen-
tested Baumann factors at low average wetness region are larger tropic efficiency ratio and tested Baumann factors in comparative
than value of order 1 that expected by Baumann, which indicates tests indicates that cryogenic turbo-expander with low liquid frac-
that super-saturation effect produces a significant part of total loss tion in two-phase expansion tests suffers from wetness loss more
in wet expansion. In order to assess the wetness loss more accu- severely than that the two-phase expansion with the higher liquid
rately, the wetness losses in steam turbine are divided into three fraction.
78 L. Niu et al. / Cryogenics 70 (2015) 76–84

2. Experimental apparatus Table 1

Main parameters of the tested turbo-expander.

2.1. Test part Dimensional parameters Thermodynamic designing parameters

Nozzle inlet DN 74 mm Inlet temperature 118 K
This cryogenic two-phase turbo-expander test part is built and Impeller inlet D1 56 mm Inlet pressure 0.410.41 MPa (A)
integrated into the turbo-expander unit in a small scale industrial Impeller outlet D2 36.2 mm Outlet pressure 0.1330.41 MPa (A)
Mass flow-rate 0.195 kg/s
air separation plant which has a capacity of producing 30 L/h liquid
Reaction ratio 0.5
nitrogen. Meanwhile, the maximum pressure of the output liquid
nitrogen can reach 0.5 MPa. Schematic diagram and photo of the
air separation system are shown in Fig. 1(a) and (c). There are
The main structure and thermodynamic designing parameters
two turbo-expanders in the host air separation plant, TE-I and
of TE-II cryogenic turbo-expander are listed in Table 1. And Fig. 2
TE-II respectively. TE-II is the tested two-phase turbo-expander
shows the cross-section view and the drawing along the flow pas-
whose affiliated pipes were reformed as the test part for
sage of TE-II.
testing the two-phase expansion performance of cryogenic
An electric heater is used to superheat the two-phase outlet
turbo-expander. The schematic diagram of the test part is shown
flow from the turbo-expander by adjusting the input AC voltage.
in Fig. 1(b). It is composed of a plate-fin heat exchanger, a cryo-
Photo of the heating unit is shown in Fig. 3(a). As shown in
genic turbo-expander, an electric heater assembly, a volumetric
Fig. 3(c), the electric heater consists of two electric heating bars
flow-meter and the sensors for temperature, pressure and rotating
which have straight fins in radial direction to enhance the heater
speed measurement. In the experiment, the connecting points of
transfer. In order to reduce the radial heat conduction, the diame-
TE-II pipeline in host system are A to A0 and B to B0 as shown in
ter of the fins in heating bar is designed 2 mm lower than the host
Fig. 1(a) and (b).
tube inner diameter, Fig. 3(c). Taken the heat conduction from
Feed stream in turbo-expander test part is extracted from the
electric heater to adjacent units into consideration, the insulating
top of the fractionating column of the host system which flows
spacers and the screw bolts with low thermal conductivity were
through the main heat exchanger in the host system firstly and
applied. The total available heating power of the heating unit is
then entering into the test part. And the mole composition of the
3500 W, and 1750 W for each heating bar.
gaseous feed stream in turbo-expander is 79% nitrogen, 20.08%
oxygen and 0.92% argon. In the test part, the feed stream is split
2.2. Principle of liquid fraction determination
into two flows. The first feed stream flows directly into TE-II.
And the second feed stream goes through the heat exchanger
In the experiment, system heat balance is built by electric hea-
and then joins the first flow at the inlet of TE-II. After exchanging
ter. As shown in Fig. 3(a), the measurement points of temperature
heat with the outlet flow of TE-II, the second feed stream is used
and pressure are located at the inlet and outlet of TE-II, and the
for regulating the thermal state at the inlet of TE-II. As TE-II oper-
outlet of the electric heater.
ates into two-phase condition, its outlet flow is heated into super-
When TE-II operates in superheated region, all temperature and
heated state after flowing through the heater assembly. Then the
pressure are both applicable, TE-II outlet enthalpy (hout) is deter-
superheated air flow goes through the heat exchanger to cool
mined from its outlet temperature (Tout) and pressure (pout).
down the second feed stream and returns to the host system.
Valves 1 and 2 are used to control the flow-rate of the feed streams. hout ¼ hðT out ; pout Þ ð1Þ

Fig. 1. Cryogenic turbo-expander test rig and host air separation system: (a) schematic diagram of the host system; (b) schematic diagram of the two-phase turbo-expander
test part and (c) photo of the host system.
 L. Niu et al. / Cryogenics 70 (2015) 76–84 79

Fig. 2. Turbo-expander cross-sectional view and flow path.

Fig. 3. Heating unit: (a) photo of heating unit in test part pipeline; (b) photo of electric heater bar; (c) cross-sectional view of the electric heater assembly.

When TE-II works in two-phase region, heat from the electric Based on heat balance, enthalpy of the TE-II outlet flow (hout) is
heater transforms the two-phase TE-II outlet flow into superheated determined from hhe and Q.
state. The enthalpy of the superheated flow after the electric
hout ¼ hhe  Q ð4Þ
heater (hhe) can be obtained by its temperature (The) and pressure
(phe). As a result, mass liquid fraction (y) at the outlet of TE-II is deter-
mined from the outlet enthalpy of (hout) and the outlet pressure
hhe ¼ hðT he ; phe Þ ð2Þ
(pout).
The heat loaded on TE-II outlet two-phase flow (Q) is the
Y ¼ yðhout ; pout Þ ð5Þ
function of heating power (U  I) and mass flow-rate (m). U is the
electric voltage and I is the heating electric current. For TE-II isentropic efficiency, Eqs. (1) and (4) are used for TE-II out-
let enthalpy calculation in superheated and two-phase operating
Q ¼ ðU  IÞ=m ð3Þ
conditions respectively, Eq. (6) is applied in calculating efficiency.
80 L. Niu et al. / Cryogenics 70 (2015) 76–84

g ¼ ðhin  hout Þ=ðhin  hout:s Þ ð6Þ 2.4. Data processing and uncertainty analysis

where the entropy and isentropic enthalpy in TE-II outlet is deter-

The uncertainty analysis discussed contains the errors of mea-
mined by the following equations:
surement and calculation. Measure parameters include tempera-
sout ¼ sin ¼ sðT in ; pin Þ ð7Þ ture, pressure, flow-rate, voltage and current. All the platinum
resistance thermometers used in the test have been calibrated by
hout:s ¼ hðpout ; sin Þ ð8Þ Chinese Academy of Sciences, and the maximum uncertainty is
±0.1 K in the temperature range from 55 K to 300 K. SMC
In the experiment, the static enthalpy is applied in calculation. PSE530-R06 pressure sensors are adopted in the test rig has a full
scale uncertainty of less than ±1% in the range of 0.1–1 MPa. The
2.3. Heat dissipation evaluation full scale uncertainty of LGBF-0050021022 standard orifice-plate
volume flow-meter is ±0.5% in the range of 30–650 m3/h. The
According to the wetness measuring probe based on thermody- external voltage and current are supplied by the AC power whose
namic method, the heat dissipation or heat loss is inevitable in relative uncertainties are both ±1%. Specific enthalpy is obtained
heating assembly [22]. The probe introduced by Morre has a vac- from property data in NIST REFPROP. Base on the following error
uum jacket around heating section, given the heating insulation propagation formula [23]:
performance of pearlite and large loaded heating power in the vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u n  2
experiment, the measurement of heat dissipation percentage is uX @f
necessary for accurate liquid fraction calculation. ry ¼t r2xi ð10Þ
i¼1
@xi
The main experiment is composed of turbo-expander thermo-
dynamic performance tests in two-phase condition and super- The uncertainty of heat (Q), mass liquid fraction (y) and isentropic
heated condition. Two-phase expansion tests are carried out efficiency (g) are ±1.3%, ±1.8% and ±2.05%, respectively.
firstly: turbo-expander is operated into two-phase region by
self-cooling, when the turbo-expander operates steadily in
two-phase condition for about 30 min, turning on electric heater 3. Results and discussion
and heating the turbo-expander outlet flow into superheated state,
then keeping steady operation for about 30 min. In two-phase test 3.1. Main experimental parameters and results
data processing, mean temperature, pressure, volume flow-rate
and electric power during the 30 min steady heating operation per- 5 two-phase expansion tests and 5 comparative superheated
iod are applied in data processing. Then superheated tests are con- expansion tests are conducted. Mean temperature, mean pressure
ducted, the identical inlet pressure and rotation speed in and mean volumetric flow-rate in steady operation period are
two-phase tests are used, consequently, the inlet temperature is applied for data processing. The results of the 10 test cases are
increased adequately and the outlet pressure is decreased to main- listed in Table 2. According to Eq. (9), the test parameters and
tain the same rotation speed. Similarly, mean temperature, pres- results of 5 heat dissipations tests are shown in Table 3.
sure and volume flow-rate in a steady operation period are
applied for data processing. In term of heat dissipation in electric 3.2. Two-phase expansion tests
heater, superheated expansion is performed to evaluate the test
system heat dissipation. Identical turbo-expander outlet pressure There are four main stages in a two-phase expansion test. Take
and identical heating power in previous two-phase expansion No. 4t test in Fig. 4 as an example. Stage I is the steady superheated
tests are applied, and adequate increasing temperature in operating stage. Stages II and III are the temperature drop stages in
turbo-expander inlet is required. The heating dissipation percent- which TE-II operates into stable two-phase condition. Stage IV is a
age is the function of turbo-expander outlet enthalpy (hout), heater superheating stage. TE-II operates at the inlet pressure of
outlet enthalpy (hhe) and heat (Q) as shown in Eq. (9). 0.372 MPa and outlet pressure of 0.132 MPa in Stage I, and the
temperature difference between the inlet and the outlet of TE-II
P  m1  ðhhe  hout Þ
rQ ¼ ð9Þ is about 28 K. In Stage II (about 600 s to 3600 s in Fig. 4), the dra-
P  m1 matic temperature drop appears for both inlet and outlet of TE-II at
The heat dissipation percentage is taken into account firstly in the beginning by turning up Valve 2 and turning down Valve 1. The
heating power calculation before the determination of temperature difference between the inlet and the outlet of TE-II is
turbo-expander outlet liquid fraction. The heat dissipation test about 22 K. In fact, this primary temperature drop in Stage II is
parameters and the heat leakage percentage will show in part 3 aiming for slowing down the changes of TE-II inlet and outlet
in this article. parameters to ensure the safe operation. The two-phase operation

Table 2
Test parameters in two-phase expansion and superheated expansion.

No. Tin (K) Tout (K) pin (MPa) pout (MPa) m (kg s1) pin/pout n (rpm)
1s 112.98 85.66 0.361 0.122 0.161 2.96 52,837
1t 105.81 / 0.361 0.129 0.167 2.80 52,604
2s 114.63 85.83 0.386 0.124 0.173 3.10 52,439
2t 104.53 / 0.386 0.136 0.179 2.83 52,143
3s 115.36 86.04 0.398 0.126 0.179 3.16 51,900
3t 105.06 / 0.398 0.140 0.185 2.84 51,683
4s 112.38 86.16 0.358 0.126 0.159 2.84 50,529
4t 98.57 / 0.359 0.136 0.166 2.63 50,362
5s 115.81 85.05 0.426 0.129 0.191 3.30 55,500
5t 110.3 / 0.427 0.143 0.199 2.98 55,102

Subscript ‘‘s’’ and ‘‘t’’ represents superheated expansion test and two-phase expansion test.
 L. Niu et al. / Cryogenics 70 (2015) 76–84 81

Table 3
Heat dissipation in electric heater corresponding to two-phase test heating power.

No. Tout (K) pout (MPa) hout (J kg1) m (kg s1) P (W) The (K) phe (MPa) hhe (J kg1) rQ (%)
1h 85.62 0.129 82621.9 0.164 1570.1 94.22 0.127 91827.1 3.85
2h 85.76 0.136 82600.1 0.175 1936.8 95.72 0.134 93272.4 3.57
3h 86.07 0.14 82,837 0.181 2575 98.98 0.138 96635.3 3.01
4h 86.15 0.136 83021.8 0.168 2331.5 98.74 0.134 96447.3 3.26
5h 85.02 0.143 81620.8 0.192 2723.3 97.87 0.14 95,426 2.67

Subscript ‘‘in’’, ‘‘out’’ and ‘‘he’’ represents turbo-inlet, turbo-outlet and heater outlet.

Fig. 5. Expansion lines of two-phase and superheated tests of No. 4 test in

enthalpy–entropy diagram. Solid lines represent the real expansion and dash lines
Fig. 4. Temperature variation of turbo-expander and electric heater from super-
represent the ideal isentropic expansion.
heated to two-phase region. Tin is turbo-inlet temperature, Tout1 is turbo-outlet
temperature in superheated region, Tout2 is turbo-outlet temperature in two-phase
region and The is electric heater outlet temperature.
can be observed in the h–s diagram. According to the
50%-reaction of TE-II, the larger percentage of the two-phase
of TE-II are realized in Stage III (3600–7500 s) by turning off Valve expansion line under the saturated line indicates that most of
1 and opening Valve 2 completely. In Stage III, the decrease of TE-II the enthalpy drop is accomplished in the impeller. Therefore, it is
inlet temperature is much larger than the decrease of TE-II outlet probable that the liquid phase has already formed in the nozzle
temperature due to the appearance of liquid phase. The huge latent of TE-II.
heat of the liquid fraction prevents the continuous drop of the tem- For the two-phase expansion tests 1t, 2t and 3t with lower liq-
perature at the outlet of TE-II. Temperature drop lasts about 40 min uid fraction, the ending thermodynamic states at stage III is
in this stage, and after about 30 min steady operation in two-phase adjusted by turning up Valve 1 and turning down Valve 2 partly
expansion, turning on the heating unit and the whole process rather than turning on or off them completely as operated in No.
comes into Stage IV (7500–10,000 s). The outlet temperature of 4t test. As a consequence, the test part operation pressure is
the electric heater arises dramatically while TE-II outlet tempera- increased in No. 5t test to realize the higher pressure ratio and
ture remains stable for about 30 min from 8500 s to 10000 s. The obtain the higher liquid fraction two-phase expansion. Liquid frac-
stable TE-II outlet temperature indicates that heating process in tion measurement results in two-phase expansion tests are shown
the system does not impact the operating conditions of TE-II. in Table 4.
The two-phase and superheated expansion lines of No. 4t test As shown in enthalpy–entropy diagram in Fig. 6, with the
are shown in h–s diagram in Fig. 5. In the diffuser, the gas flow increase of wetness, the inlet temperature and initial specific
experiences the process of pressure increase and velocity decrease. enthalpy are declined from test 1t to 4t. Although the inlet specific
The flow thermodynamic state at the impeller outlet approaches enthalpy of No. 5t test is higher than No. 4t test, the highest
vapor saturated line more closely than that after diffuser. In order enthalpy drop is observed in test 5t due to its largest expansion
to remove the sub-cooling effect in superheated expansion, 1.5– ratio. Majority of expansion process takes place in superheated
2.5 K superheat degree in the real superheated expansion line zone No. 1t and No. 2t tests, by contrast, in No. 4t and No. 5t tests,

Table 4
Liquid fraction measurement results in two-phase expansion tests.

No. Tin (K) pin (MPa) pout (MPa) m (kg s1) The (K) phe (MPa) P (W) hout (J kg1) y (%)
1t 105.81 0.361 0.129 0.167 90.1 0.127 1569 77,955 1.16
2t 104.53 0.386 0.136 0.179 90.11 0.134 1937.1 76340.8 2.12
3t 102.57 0.398 0.140 0.185 90.96 0.137 2574.7 74253.9 3.17
4t 98.57 0.359 0.136 0.166 89.4 0.135 2330 72534.1 3.88
5t 100.3 0.427 0.143 0.199 87.18 0.137 2723.7 70459.9 5.02

Subscript ‘‘he’’ represents heater, ‘‘The’’ is obtained when electric heater is on.
82 L. Niu et al. / Cryogenics 70 (2015) 76–84

clearance loss, wheel trailing edge loss and disk friction loss are
also significant factors that impact the isentropic efficiency [24].
Including all of these energy losses in turbo-expander, overall isen-
tropic efficiency in cryogenic turbo-expander is given as Eq. (6). In
two-phase expansion tests, the calculation of isentropic efficiency
is the same as superheated expansion which depends on beginning
and final status in expansion process. Due to occurrence of
sub-cooling, nucleation and droplet growth that gas expands from
superheated region into two-phase region, wetness loss is added
into the energy loss in turbo-expander.
In wet steam turbine, the two-phase expansion thermodynamic
performance in low pressure stages has been researched since
early 20th century, and as Baumann illustrated firstly in Ref. [13],
he predicates that 1% mean wetness will bring 1% efficiency
decrease in wet steam turbine. In spite of different working flow
in steam turbine and cryogenic turbo-expander, the
non-equilibrium spontaneous process is the same in two-phase
Fig. 6. Expansion lines in two-phase tests. expansion. By applying Baumann’s theory to cryogenic
turbo-expander, the predicted Baumann value of efficiency ratio
and Baumann factor are used as theoretical results to compare
a large portion of expansion occurs under vapor saturated line with our experimental results in this paper. The formula for
which leads to relatively high wetness. calculation of mean wetness and tested Baumann factors in this
experiment are shown as follow:
3.3. Isentropic efficiency analysis yin þ yout
ymean ¼ ð11Þ
2
The performance of radial-inflow cryogenic turbo-expander
generally involves numbers of parts of energy loss as gas flows ð1  gwet =gdry Þ
through the separate components, e.g. loss in nozzle, vaneless aB ¼ ð12Þ
ymean
space, wheel and diffuser. In addition, incidence loss, rotor

Fig. 7. Expansion lines of two-phase tests and their comparative superheated test in enthalpy–entropy diagram (a) tests 1t and 1s, (b) tests 2t and 2s, (c) tests 3t and 3s and
(d) tests 5t and 5s.
 L. Niu et al. / Cryogenics 70 (2015) 76–84 83

Table 5
Comparative tests with different liquid fraction.

No. pin/pout Dh (J kg1) y (%) g gt / g s Tested Baumann factor

1s 2.96 25188.2 / 0.8772 /
1t 2.80 22144.4 1.16 0.8618 0.9825 3.01
2s 3.10 26407.6 / 0.8776 /
2t 2.83 21884.1 2.12 0.8607 0.9807 1.82
3s 3.16 26945.8 / 0.8761 /
3t 2.84 21513.4 3.17 0.8571 0.9782 1.37
4s 2.84 24139.2 / 0.873 /
4t 2.63 19390.6 3.88 0.851 0.9747 1.3
5s 3.30 28179.8 / 0.8866 /
5t 2.98 22110.6 5.02 0.8605 0.9706 1.17

The expansion lines of other 4 comparative tests are shown

from Fig. 7(a)–(d). In Fig. 7, the solid lines represent the real expan-
sion process and the dash lines represent the isentropic expansion
process.
As shown in Table 5, the superheated expansion efficiency is
stabilized around 87.5% in test 1s–4s, while the efficiency of No.
5s test is extremely high due to its relatively higher inlet
temperature and pressure, as well as larger expansion ratio, and
its expansion conditions are very close to design parameter (listed
in Table 1). In two-phase expansion tests, the efficiency is
decreased with the increase of liquid fraction, unexpectedly, due
to the large inlet pressure and expansion ratio in teat 5t, its largest
liquid fraction dose not corresponds to the lowest efficiency, while
the lowest efficiency is found in No. 4t test.
Although the largest liquid fraction does not represent the low-
est two-phase efficiency, when the two-phase expansion efficiency
is compared with the superheated expansion efficiency, the effi-
ciency ratio gt/gs is decreased with the increase of liquid fraction
from test 1t to 5t, which indicates the two-phase expansion
efficiency deviates from its corresponding superheated expansion
Fig. 9. Tested Baumann factors with different mean liquid fraction.
efficiency more severely with the liquid fraction increase.
The comparison of the experimental efficiency ratio gt/gs with
the predicted Baumann value is shown in Fig. 8. In this figure, Baumann prediction value is declined. When the mean liquid frac-
the first point is assumed as a completely superheated expansion, tion is larger than 1.5% (in test 3, 4, 5), the experimental results
so the efficiency ratio is 1. In No. 1 test, the superheated efficiency approach the prediction more closely, and the difference between
is 87.72%, based on Baumann rule, the predicted two-phase effi- experimental results and the predicted Baumann value begins to
ciency and the experimental value are 87.14% and 86.18% respec- stay steadily. As shown in Fig. 9, the tested Baumann factor has
tively, as a consequence, the efficiency ratio in test 1 deviates the same tendency as the variation of efficiency ratio. The tested
from the Baumann prediction severely. With the mean liquid frac- Baumann factor reaches 3.01 when the mean wetness is 0.58% in
tion increase, the deviation of experimental efficiency ratio from No. 1 test. And the tested Baumann factor decreases dramatically
in No. 2, but then declines slightly as mean wetness is larger than
1.5% in test 3, 4, 5. In No. 5 test, the factor value of 1.17 is very close
to Baumann value of 1.
As applied Baumann rule to evaluate the thermodynamic
performance in cryogenic turbo-expander, the variation of the
experimental efficiency ratio and tested Baumann factor indicate
that the turbo-expander in two-phase expansion condition suffers
higher wetness loss than the value predicated by Baumann rule.
Moreover, we found that the wetness loss in lower mean wetness
condition is much larger than that in relatively higher wetness, and
the prediction is considerably accurate when the mean liquid frac-
tion is larger than 1.5% in our experiment.

4. Conclusion

Liquid fraction in turbo-expander outlet in this experiment is

successfully measured by heat balance method. The heat dissipa-
tion of electric power is less than 3.85%. 5 two-phase expansion
tests with different wetness from 1.16% to 5.02% are carried out
Fig. 8. Experimental efficiency ratio compared with the value of Baumann and comparatively analyzed with 5 superheated expansion tests
prediction. which share the identical inlet pressure and rotational speed. By
84 L. Niu et al. / Cryogenics 70 (2015) 76–84

using Baumann rule, the wetness loss in two-phase expansion is [7] Faro I, Small TR, Hill FK. The supersaturation of nitrogen in a hypersonic wind
tunnel. J Appl Phys 1952;24(1):40–3.
evaluated. Although the lowest isentropic efficiency is not
[8] Sinha Somnath, Laksmono Hartawan, Wyslouzil Barbara E. A cryogenic
observed in the two-phase expansion condition with highest liquid supersonic nozzle apparatus to study homogeneous nucleation of Ar and
fraction, the experimental efficiency ratio and tested Baumann other simple molecules. Rev Sci Instrum 2008;79(11).
factor fall with the increase of mean wetness. Moreover, higher [9] Zahoransky RA, Höschele, Steinwandel. Homogeneous nucleation of argon in
an unsteady hypersonic flow field. J Chem Phys 1999;110(17):8842–3.
wetness loss in two-phase expansion with low liquid fraction is [10] Zahoransky RA, Höschele, Steinwandel. Formation of argon clusters by
found, and the efficiency ratio and the experimental Baumann homogeneous nucleation in supersonic shock tube flow. J Chem Phys
factor prone to expectation value as the mean liquid fraction larger 1995;103(20):9038.
[11] Iland Kristina, Wedekind Jan, Wölk Judith, Strey Reinhard. Homogeneous
than 1.5%. nucleation of nitrogen. J Chem Phys 2009;130(11).
[12] Iland Kristina, Wölk Judith, Strey Reinhard, Kashchiev Dimo. Argon nucleation
in a cryogenic nucleation pulse chamber. J Chem Phys 2007;127(15).
Acknowledgements [13] Baumann K. Some recent developments in large steam turbine practice. Inst
Electr Eng 1921;59:565–623.
This project was supported by the National Nature Science [14] Gas dynamics of wet steam and energy losses in wet-steam turbines. In:
Moore MJ, Sieverding CH, editors. Two-phase steam flow in turbines and
Foundation of China (51476122), the Specialized Research Fund separators. Washington DC: Hemisphere. p. 59–126.
for the Doctoral Program of Higher Education (20130201110038), [15] Gyarmathy G. Grundlagen einer Theorie der Nassdampfturbine. PhD thesis.
and the Fundamental Research Funds for the Central Universities Juris, Zurich: ETH Zurich; 1962.
[16] Miller EH, Schofield P. The performance of large steam turbines with water
(2012JDGZ03). reactors. ASME Winter Annu Meet, New York 1972.
[17] Guo T, Sumner WJ, Hofer DC. Development of highly efficient nuclear HP
steam turbines using physics based moisture loss models. Proc ASME Turbo
References Expo, Montreal, Canada 2007.
[18] Laali AR. A new approach for assessment of the wetness losses in steam
[1] Sixsmith H, Hasenbin R, Valenjuela JA. A miniature wet turboexpander. Adv turbines. PhD thesis. Electricite de France; 1991.
Cryog Eng 1990;35:989–95. [19] Kawagishi H, Onoda A, Shibukawa N, et al. Development of moisture loss
[2] Kato T, Yamaura H, Kawno K, Hiyama T, Tada E, Kakayama Y, et al. Large scale models in steam turbines. Heat Transf – Asian Res 2013;42:661–4.
turboexpander development and its performance test result. Adv Cryog Eng [20] Aungier RH. Aerodynamic performance analysis of radial-inflow turbines.
1990;35:1005–12. Turbine aerodynamics: axial-flow and radial-inflow turbine design and
[3] Wang Xinjun, Liao Gaoliang, Zhu Ding, Yao Jinling, Bai Xiaowei. Study on the analysis. New York: ASME PRESS; 2006. p. 216.
sampling quality of wetness measurement probe for thermodynamic methods. [21] Obata M, Saji N, Asakura H, Yoshinaga S, Ishizawa T. Theoretical study of the
Proc ASME Turbo Expo 2012;6:375–84. performance of a wet turbine for helium refrigerators. Adv Cryog Eng
[4] Ardashev VI, Plachendovskii DI. Examination of two-phase regimes of 2000;45:1431–8.
cryogenic turbine expansion engines. Chem Pet Eng 1984;20(3–4):182–5. [22] Instrumentation for wet steam. In: Moore MJ, Sieverding CH, editors. Two-
[5] Cai X, Niu F, Li J, Su M, Ning T, Song Y, et al. Investigation of wet steam flow in a phase steam flow in turbines and separators. Washington DC: Hemisphere. p.
300 MW direct air-cooling steam turbine. Part 1: measurement principles, 191–258.
probe, and wetness. Proc Inst Mech Eng, Part A: J Pow Energy 2010;224(1): [23] Moffat Robert. Describing the uncertainties in experimental results. Exp
129–37. Thermal Fluid Sci 1998;1(1):3–17.
[6] Schatz Markus. Experimental study of steam wetness in a model steam turbine [24] Ghosh Subrata K, Sahoo RK, Sarangi Sunil K. Mathematical analysis for off-
rig: presentation of results and comparison with computational fluid dynamics design performance of cryogenic turbo-expander. J Fluids Eng, Trans ASME
data. Proc Inst Mech Eng, Part A: J Pow Energy 2014;228(2):129–42. 2011;133(3).