Nuclear Engineering and Design 326 (2018) 290–300

Contents lists available at ScienceDirect

Nuclear Engineering and Design

journal homepage: www.elsevier.com/locate/nucengdes

An improved correlation on the onset of heat transfer deterioration in T

supercritical water
⁎
Ma Donglianga,b,c, Zhou Taoa,b,c, , Li Binga,b,c, Muhammad Ali Shahzada,b,c, Huang Yanpingd
a
School of Nuclear Science and Engineering, North China Electric Power University, Beijing 102026, China
b
Institute of Nuclear Thermal-hydraulic Safety and Standardization, North China Electric Power University, Beijing 102206, China
c
Beijing Key Laboratory of Passive Safety Technology for Nuclear Energy, North China Electric Power University, Beijing 102206, China
d
Science and Technology on Reactor System Design Technology Laboratory, Nuclear Power Institute of China, Chengdu 610041, China

A R T I C L E I N F O A B S T R A C T

Keywords: On the basis of the supercritical water experimental data, the distribution of onset heat flux is analyzed for the
Supercritical water heat transfer deterioration. This study evaluates the performance of empirical correlations for the prediction of
Heat transfer deterioration heat flux, causing the onset of heat transfer deterioration in vertical tubes of supercritical water. In light of the
Error analysis calculation results, it is shown that the onset of heat transfer deterioration effect is better predicted using the
Multiple regression analysis method
Mokry, Li and Schatte criterion. On the basis of experimental data, a new correlation is developed for predicting
the heat flux, which causes the onset of heat transfer deterioration in supercritical water. The new calculation
model has considered the influence of mass flux, pipe diameter, pressure and inlet temperature. The newly
developed correlation has an average relative error of 0.020837 and an average absolute deviation of 0.1627.
The improved criterion is more conducive for the design of supercritical water reactors in the future.

1. Introduction presenting an improved k-ε/kt-εt model. Near heat transfer deteriora-

tion, the wall temperature is well-predicted by the new model, com-
Supercritical-Water-cooled Reactor is internationally recognized as pared with the traditional low-Reynolds models and traditional k-ε
one of the potential IV generation nuclear reactors. Due to its high turbulence models. The results show that the buoyancy effect redis-
thermal efficiency and lower economic costs, the supercritical water tributes shear stress and radial flow velocity, leading to heat transfer
reactor is strongly competitive with other reactor concepts in the fu- deterioration and recovery. The new model can predict heat transfer
ture. Based on its characteristics, the supercritical fluid heat transfer (in deterioration more accurately. Sharabi et al. (2008) carried out an
tubes and channels) has been a heavily-studied topic in the past five analysis on supercritical CO2 within triangle and square channels. Due
decades. A vast number of experiments have been carried out for the to the buoyancy effect, the heat transfer deterioration phenomenon is
heat transfer in supercritical water. When supercritical water heat observed in supercritical CO2. Through simulation the three-dimen-
transfer deterioration (HTD) phenomenon occurs, the pipe wall tem- sional features of the fluid under strong buoyancy in different channel
perature rises very quickly. This will lead to the emergence of wall shapes is discussed. It facilitates a better understanding of the me-
temperature peak value. The heat transfer coefficient between the pipe chanism of heat transfer deterioration phenomenon. Wen and Gu
wall and fluid becomes abnormally low around the location of this peak (2011) found that the change of dynamic viscosity and specific heat (at
value. This is a serious threat to the safe and stable operation of the constant pressure) can suppress the heat transfer deterioration. How-
supercritical water reactor. Therefore the initial heat flux value, causing ever, a decrease in thermal conductivity can promote the heat transfer
the onset of HTD for the design of the supercritical water reactor, has a deterioration phenomenon. Eter et al. (2016) carried out the super-
crucial significance. critical CO2 heat transfer deterioration experimental research in a
Through the analysis of the experimental study of supercritical three-rod bundle equipped with wire-wrap and grid spacers. When the
water and supercritical CO2, Pioro and Duffey (2005) showed that the mass flux is lower (200 kg/m2s −700 kg/m2s), the heat flux exceeds a
deteriorated heat transfer usually appears at higher heat fluxes and certain degree and heat transfer deterioration occurs. Gu et al. (2015)
lower mass fluxes. This phenomenon can be suppressed by increasing performed experimental research on supercritical water inside circular
the turbulence level with flow obstructions. Zhang et al. (2012) tubes. It was believed that when the DNB ratio (ratio of actual to

⁎
Corresponding author at: School of Nuclear Science and Engineering, North China Electric Power University, Beijing 102026, China.
E-mail addresses: [email protected] (M. Dongliang), [email protected] (Z. Tao).

https://doi.org/10.1016/j.nucengdes.2017.11.026
Received 11 July 2017; Received in revised form 6 November 2017; Accepted 13 November 2017
0029-5493/ © 2017 Elsevier B.V. All rights reserved.
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

calculated heat transfer co-efficient) is lower than 0.3, the heat transfer transition, it is affected by buoyancy as well as fluid acceleration at the
deterioration may occur at any random point. Because it was observed same time. Therefore, the occurrence of heat transfer deterioration and
that when the ratio is 0.3, there is a continuous change. It is believed its development mechanism in the supercritical water, is regarded as an
that the ratio of 0.3 does not represent any physical phenomena with a extremely complex phenomenon. Amongst the academia, there is still a
specific behavior. Thus, the change of heat transfer coefficient is mainly lack of unified conclusion on this subject. This article presents a study
based on the acceleration number πA and buoyancy number πB. Hence on the heat transfer deterioration calculation formula, based on an in-
the very definition of heat transfer deterioration is still controversial itial power value of the supercritical water, which is conducive to the
and having nounified conclusion. Podila and Rao (2015) simulated the design and development of future supercritical water reactors.
heat transfer deterioration phenomenon in supercritical water reactor
bundle channel tube. The study shows that when the mass flux is lower, 2. Calculation model
an increase in the inlet temperature or system pressure can effectively
suppress the occurrence of heat transfer deterioration phenomenon. Liu 2.1. Definition of heat transfer deterioration
et al., 2013a,b) found that when the mass flux is high, the effect of
acceleration plays a vital role in heat transfer deterioration. And at low When heat transfer deterioration phenomenon occurs then the heat
mass flux, the buoyancy effect plays an important role for the heat transfer coefficient decreases suddenly up to a very low value compared
transfer deterioration. When the pipe diameter is larger, it will result in to the normal heat transfer. And the wall temperature will suddenly rise
a higher wall temperature, hence the heat transfer is more likely to very quickly. So far in academia, the definition of deterioration of heat
deteriorate. At high mass flux, heat equivalent diameter is comparable transfer of supercritical water is not specified in a widely accepted
with the hydraulic diameter, also playing an important role in heat form. For comparing the formula about predicting deterioration heat
transfer deterioration. Yildiz and Groeneveld (2014) analyzed and transfer in supercritical water, a quantitative definition in terms of the
summarized the effect of pipe diameter size on the heat transfer dete- heat transfer coefficient is necessary. The general definition of heat
rioration in a supercritical medium (CO2, water, R-22 and R-12). Stu- transfer deterioration has been proposed by Koshizuka et al. (1995). It
dies show that increasing the pipe diameter will reduce the initial heat is used frequently in research paper is the following:
flux value, causing the heat transfer deterioration phenomenon. When
α < 0.3·αDB (1)
the mass flow rate, heat flux and the axial length are constant, the fluid
acceleration effect (due to fluid expansion) will be more obvious in In this definition, αDB is the heat transfer coefficient under normal
pipes of smaller diameter. Whereby for pipes of larger diameter, it will circumstances, as calculated by the Dittus-Boelter correlation. The D-B
have a more pronounced buoyancy effect. The thickness of the low correlation is shown below:
density layer (near the wall) is increased in such pipes, hence resulting
λb ·Nub λ
in more critical heat transfer deterioration. In solar central heating αDB = = ⎛ b ⎞·0.023·Re0.8
⎜ ⎟
b ·Pr
0.4
b
systems, Zhang et al. (2016) studied the heat transfer deterioration in di ⎝ di ⎠ (2)
vertical pipelines under unilateral non-uniform condition. Based on the
G·d i
non-uniform heating characteristics of solar energy, a suitable heat Re b =
ηb (3)
transfer deterioration criteria was proposed, along with the calculation
scheme for that condition. Based on earlier experimental data for su- ηb ·c p,b
percritical media, Jackson (2013) proposed an approach for describing Prb =
λb (4)
the combined effects on heat transfer of buoyancy and acceleration. It
can help to interpret experimental data on heat transfer at supercritical where λb is the bulk thermal conductivity (W/m K), d i is the pipe
pressure. Thereby, it becomes possible to use the supercritical flow diameter (mm), G is the mass flux (kg/m2s), ηb is the bulk dynamic
experimental data for guiding the thermal-hydraulic system design. At viscosity (Pa.s) and c p,b is the bulk specific heat at constant pressure (J/
the same time, the mechanism of heat transfer deterioration is also kg K).
analyzed. Jager et al. (2011) analyzed the calculation results of TRACE
code for the heat transfer coefficient of supercritical water, comparing 2.2. Existing criterion for the onset of heat transfer deterioration
them with the results of different correction formulae. It is suggested
that Bishop, Sandberg and Tong empirical formulae can be used as There are nine different power criterion, proposed for the calcula-
recommended formulae for nuclear reactor safety assessment and de- tion of onset of heat transfer deterioration. These criterion are sum-
sign. It is also suggested to evaluate and validate the experimental data marized in Table 1:
of heat transfer deterioration, which occurs in high power and lower From the existing criterion, as listed in Table 1, it can be seen that
mass fluxes. Zahlan et al. (2015) and Loewenberg et al. (2008) analyzed most of the criterion consider only the mass flux, pipe diameter, or
the heat transfer characteristics of supercritical water in the form of a pressure alone. But there is a lack of criterion which consider the mass
look-up table. Although this method has a higher accuracy, it is effec- flux, pipe diameter and pressure altogether. Although the No. 9 cri-
tive under the range of limited experimental parameters. Jingjing et al. terion has considered the influence of flow pipe diameter, mass flux and
(2014) used the sensitivity analysis method on some parameters of pressure, but because the specific heat at constant pressure and coef-
artificial neural network, getting the influence of different parameters ficient of thermal expansion the pseudo-critical point change with the
on the sensitivity of critical heat flux (CHF). Huang et al. (2016) eval- change of pressure, these two parameters are a function of pressure,
uated five different criterion for the prediction of heat transfer dete- generally not directly obtained, so they are need to be calculated first.
rioration. It was pointed out that heat transfer deterioration is a very So the use of this formula process increases the complexity. It is im-
complicated phenomenon. This phenomenon has a close relationship portant to consider the influence of multiple parameters and that the
with heat flux, mass flux, medium, pipe diameter, geometry, system criterion to be as simple as possible.
pressure and so on.
In the above discussion, various studies have analyzed the heat 2.3. Model for regression analysis of heat transfer deterioration
transfer deterioration phenomenon from all kinds of aspects, such as the
experimental research, numerical simulation analysis, calculation When calculating the initial heat flux (causing heat transfer dete-
model optimization, code calculation analysis, mechanism analysis and rioration), four parameters considered; the mass flux, pipe diameter,
the usage of look-up tables. Since the physical parameters of super- pressure and the inlet temperature, respectively. The calculation form is
critical water change continuously during the process of pseudo-critical shown in Eq. (5). This problem can be solved by multiple regression

291
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Table 1
Existing criterion for the onset of heat transfer deterioration in supercritical water.

No. Source Mathematical form DHT definition

1 Vikhrev et al. (1967) qDHT = 0.4·G Not stated

2 Styrikovich et al. (1967) qDHT = 0.58·G Not stated
3 Yin et al. (2006) qDHT = 1/2.16·G Not stated
4 Yamagata et al. (1972) qDHT = 0.2·G1.2 Not stated
5 Kim et al. (2005) qDHT = 0.0002·G 2 Not stated
6 Mokrya et al. (2011) qDHT = −58.97 + 0.745·G Not stated
7 Cheng et al. (2009) c p,pc α < 0.3·αDB
qDHT = G·1.354·10−3·⎛ ⎜
⎞
⎟
βpc
⎝ ⎠
8 Li et al. (2014) 1.21 α < 0.3·αDB
qDHT = di·⎡0.36·
⎣
( )−1.1⎤⎦
G
di

9 Schatte et al. (2016) c p,pc

2.065 α < 0.3·αDB
qDHT = 1.942·10−6·G 0.795·(30−di)0.339·⎛ ⎜
⎞
⎟
βpc
⎝ ⎠

analysis method, which is introduced by Backhaus et al., 2008. A si- deviation and the standard deviation etc. The standard deviation is
milar method is also used by Schatte et al. (2016) for developing a calculated to predict the degree of data deviation from the arithmetic
correlation of the onset of heat transfer deterioration. average. The smaller the standard deviation is, the better the existing
and the new criterion of heat transfer deterioration are. The root mean
qDHT = C1·f1 (G )·f2 (d i )·f3 (p)·f4 (Tin ) (5)
square error is used to measure the deviation size between the true
The practical formula for computing the initial heat flux of heat value and the predicted value. The formulae of relative deviation, the
transfer deterioration is shown in Eq. (6). average relative deviation, mean absolute relative deviation, the root
e e mean square deviation and the standard deviation formulae are shown
qDHT = C1 G eG·(di )edi ·(p)eP ·(Tin )eTin = C1·X1eG ·X2 di ·X3eP ·X4 Tin (6) in the following equations, respectively.
The task for Eq. (6) is to determine the values of [C1,eG,e di,ep,e Tin]. qDHT,cal−qDHT,exp
These parameters should keep the sum of squared errors (SSE), between Relative deviation RDi =
qDHT,exp (12)
the predicted result and the experimental data, to a minimum. The
natural logarithm is applied on both sides of Eq. (6), transforming it i=n
into a linear regression problem, as shown in Eq. (7). ∑ RDi
i=1
ln(qDHT ) = ln(C1) + eGln(X1) + e di ln(X2 ) + eP ln(X3) + e Tin ln(X 4 ) Average relative deviation MRD =
(7) n (13)
Eq. (7) can be solved by optimization of parameters, using the linear sumii==1n |RDi |
regression method. The computation formula is shown in Eq. (8). Mean absolute relative deviation MARD =
n (14)
B = (X ′X )−1X ′Y (8) i=n

X, Y and B are matrices and vectors, as defined below: ∑ RDi2

i=1
Root mean square deviation RMSRD =
⎡1 ln(X1 (1)) ln(X2 (1)) ln(X3 (1)) ln(X 4 (1)) ⎤ n (15)
⎢1 ln(X1 (2)) ln(X2 (2)) ln(X3 (2)) ln(X 4 (2)) ⎥
X=⎢ ⎥ i=n
⎢⋮ ⋮ ⋮ ⋮ ⋮ ⎥ ∑ (RDi−MRD)2
⎢ ⎥
⎣1 ln(X1 (N )) ln(X2 (N )) ln(X3 (N )) ln(X 4 (N )) ⎦ (9) Standard deviation SD = i=1
n−1 (16)

⎡ ln(qDHT,exp (1)) ⎤
⎢ ln(q ⎥
DHT,exp (2)) ⎥ 3. Data processing
Y=⎢
⎢⋮ ⎥
⎢ ln(q (N )) ⎥ 3.1. Data source
⎣ DHT,exp ⎦ (10)
The source of data is representative for the supercritical water ex-
⎡ ln(C1,opt ) ⎤
⎢ eG,opt ⎥ periments, available in open literature from 1963 to 2015. The ex-
B = ⎢ e di,opt ⎥ perimental parameters of the investigated measurements are shown in
⎢ ⎥ Table 2.
⎢ ep,opt ⎥
⎢ e Tin,opt ⎥ It can be seen from Table 2 that the supercritical water experimental
⎣ ⎦ (11)
data includes 5130 data points. The subject of the experimental data is
After optimization of the parameters [C1,eG,e di,ep,e Tin], the criterion a uniformly heated vertical pipe. The IAPWS-IF97 formula is used to
of heat transfer deterioration is determined in the power calculation calculate the physical parameters of supercritical water. The range of
formula of supercritical water using the regression model. all data is summarized as follows: bulk water specific enthalpy is
451.30–3162.85 kJ/kg, mass flux is 203–3000 kg/m2s, heat flux is
2.4. Formulas for the error analysis 166–2960 kW/m2, pressure is 22.5–31 MPa, the pipe diameter is
0.7–38.1 mm and inlet temperature is 103.49–372.28 °C.
The error analysis is performed on the results of existing criterion
and as well as the new criterion, showing its influence on the heat 3.2. Data statistics
transfer deterioration. This includes the relative deviation, average re-
lative deviation, mean absolute relative deviation, root mean square According to the definition of heat transfer deterioration (as given

292
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Table 2
Supercritical water experiment statistical data.

Author Specific enthalpy Hb Heat transfer Mass flux G Heat flux q Pressure p Pipe diameter Inlet temperature Point
[kJ/kg] coefficient h [kg/m2s] [kW/m2] [MPa] D Tin [°C] number
[kW/m2K] [mm]

Ackerman (1970) 1012.96–2551.31 5.72–11.70 1220.00 1260.00 22.75 9.40 233.91 15.00
Glushchenko and Gandzyuk 512.97–1977.11 10.35–44.39 2200.00 1150–2960 23.50 1.00 118.24–126.03 63.00
(1972)
Griem (1996) 1521.87–2410.25 7.26–26.66 500–1000 300–400 25.00 14.00 334.26–359.50 71.00
Gu et al. (2015) 1343.44–2451.43 3.25–23.86 600–1000 700–1000 23–25 7.60–10 302.38–315.68 359.00
Wang et al. (2015) 889.80–3135.87 3.42–55.76 400–1000 200–400 23–28 2.00 206.37–305.89 266.00
Herkenrath et al. (1967) 1512.39–2755.58 6.60–20.21 1000.00 500–750 24–25 10.0–20.0 332.39–368.21 69.00
Wang et al. (2009) 1253.78–2557.11 2.05–22.48 900.00 300.00–500.00 26.00 26.00 284.81–311.16 37.00
Pan et al. (2011) 1428.87–3043.12 1.41–20.82 400–1200.0 284.00–663.00 22.5–30 17.00 319.14–334.82 305.00
Kirillov et al. (2005) 1551.88–2204.54 7.17–14.26 496.00 239.00 24.03 10.00 338.75 42.00
Lee and Haller (1974) 1165.69–2108.35 4.38–22.30 542–1627 252–1101 24.10 38.10 266.66–313.13 105.00
Ornatskiy et al. (1972) 451.30–2639.13 10.22–98.08 3000.00 1180–2960 23.50 0.70 103.49–104.32 104.00
Jie et al. (2011) 1349.30–3162.85 2.46–157.55 1009–1626 216.00–649.00 22.5–30 17.00 303.18–345.95 266.00
Mokrya et al. (2010) 1570.37–3142.97 1.25–48.84 203.00–1499 166.00–884.00 23.9–24.2 10.00 341.65–353.17 1323.00
Shitsman (1963) 1381.38–2934.82 1.72–24.90 430.00 210–386 23.30 10.00 309.28–334.14 106.00
Swenson et al. (1965) 659.64–3053.54 13.07–63.63 2150.00 786–1730 23–31 9.42 151.90–337.60 110.00
Vikhrev et al. (1967) 624.77–2679.82 6.47–39.36 1400.00 700–1160 26.50 20.40 144.34–147.81 72.00
Wang Fei et al. (2013) 1322.94–2479.42 1.54–29.91 450–1500 450–1250 23–26 10.00 298.23–329.02 679.00
Feng et al. (2005) 1194.61–2781.18 10.21–171.08 600–1200 200.00–600.00 23–30 12.00 272.71–372.28 353.00
Yamagata et al. (1972) 1089.93–2996.40 7.91–78.22 1114–1260 233–930 22.6–29.4 7.5–10 250.63–338.34 507.00
Yoshida and Mori (2000) 1233.30–2707.81 14.43–38.33 1180.00 698.00 24.50 10.00 280.60 40.00
Zhu et al. (2009) 1239.79–2852.95 1.04–41.87 600–1200 200–400 26–30 26.00 281.98–336.59 78.00
Zhu et al. (2007) 1621.69–2677.27 4.48–18.92 600.00 200.00 26–30 26.00 351.25–352.00 76.00

Total 451.30–3162.85 1.04–171.08 203.00–3000 166.00–2960 22.50–31 0.7–38.1 103.49–372.28 5130

Fig. 2. Experimental point distribution for deteriorated heat transfer.

Fig. 1. Complete experimental point distribution.

3.3. Data selection

in Eq. (1)), the experimental data points of supercritical water can be
classified into normal-, enhanced heat transfer and heat transfer dete-
The statistical analysis of experimental parameters, shown in
rioration groups. The distribution of all experimental points is shown in
Table 1, are at different pressures, pipe diameters, mass fluxes and heat
Fig. 1. The distribution for heat transfer deterioration is shown in Fig. 2.
flux conditions. As the bulk fluid temperature and specific enthalpy
The statistical distribution of heat transfer coefficient can be seen
gradually rise, the points of heat transfer deterioration are obtained
from Fig. 1. The horizontal axis Tf/Tpc represents the ratio of bulk fluid
under different temperatures and specific enthalpies. When analyzing
temperature and pseudo-critical point temperature. The vertical axis H/
the experimental data (as long as the deterioration has occurred on one
HDB represents the ratio of experimental heat transfer coefficient and
of the temperatures or specific enthalpies), it is believed that the heat
the heat transfer coefficient calculated through D-B formula. When the
transfer deterioration phenomenon will occur under the same condi-
ratio is less than 0.3, the occurrence of heat transfer deterioration is
tions of pressure, pipe diameter, mass flux and heat flux. Thereby, when
indicated. The total number of experimental points are 5130, in which
the pressure, pipe diameter, mass flux and heat flux are in the same
the data points of heat transfer deterioration count 592. These points
condition, and only the specific enthalpy or bulk fluid temperature
account for 11.54% of all experimental data points.
changes, it is assumed that this belongs to a particular type of heat
The condition of heat transfer deterioration is visible from the ex-
transfer condition. Through this parameter simplification method ap-
perimental point distribution, as shown in Fig. 2. When the phenom-
plied on the heat transfer deterioration points, the 592 heat transfer
enon of heat transfer deterioration occurs, the ratio of the bulk fluid
deterioration points are classified into 62 groups of working conditions.
temperature and the pseudo-critical point temperature is 0.8–1.2, with
The specific parameters of these 62 groups are shown in Table 3.
a major concentration in the range of 1.0–1.05.

293
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Table 3
The specific parameters of 62 groups of working conditions.

No. Pipe diameter i Mass flux G Heat flux q Pressure p Inlet temperature Data Source
[mm] [kg/m2/s] [kW/m2] [MPa] T.in [°C]

1 9.40 1220.00 1260.00 22.75 233.91 Ackerman (1970)

2 20.00 1000.00 750.00 25.00 368.21 Herkenrath et al. (1967)
3 10.00 430.00 300.00 23.30 329.17 Shitsman (1963)
4 10.00 430.00 337.00 23.30 329.88 Shitsman (1963)
5 10.00 430.00 384.00 23.30 309.28 Shitsman (1963)
6 10.00 430.00 386.00 23.30 330.76 Shitsman (1963)
7 20.40 1400.00 1160.00 26.50 147.81 Vikhrev
8 10.00 1180.00 930.00 22.60 323.25 Yamagata et al. (1972)
9 7.60 1000.00 700.00 23.00 315.68 Gu et al. (2015)
10 7.60 1000.00 700.00 25.00 313.49 Gu et al. (2015)
11 10.00 1000.00 700.00 23.00 306.51 Gu et al. (2015)
12 10.00 1000.00 700.00 25.00 304.93 Gu et al. (2015)
13 10.00 600.00 700.00 23.00 305.52 Gu et al. (2015)
14 10.00 600.00 700.00 25.00 305.59 Gu et al. (2015)
15 10.00 1000.00 1000.00 23.00 309.55 Gu et al. (2015)
16 10.00 1000.00 1000.00 25.00 302.38 Gu et al. (2015)
17 26.00 900.00 400.00 26.00 311.16 Wang et al. (2009)
18 26.00 900.00 500.00 26.00 284.81 Wang et al. (2009)
19 17.00 400.00 300.00 22.50 319.38 Jie et al. (2011)
20 17.00 430.00 284.00 22.50 327.32 Jie et al. (2011)
21 17.00 430.00 473.00 22.50 332.89 Jie et al. (2011)
22 17.00 430.00 663.00 22.50 326.80 Jie et al. (2011)
23 17.00 600.00 284.00 22.50 323.14 Jie et al. (2011)
24 17.00 770.00 284.00 22.50 330.89 Jie et al. (2011)
25 17.00 770.00 284.00 28.00 324.93 Jie et al. (2011)
26 17.00 770.00 284.00 30.00 319.14 Jie et al. (2011)
27 17.00 820.00 284.00 28.00 326.00 Jie et al. (2011)
28 17.00 820.00 473.00 28.00 334.82 Jie et al. (2011)
29 17.00 820.00 530.00 28.00 334.76 Jie et al. (2011)
30 17.00 820.00 663.00 28.00 326.50 Jie et al. (2011)
31 10.00 203.00 203.00 24.10 347.93 Mokrya et al. (2010)
32 10.00 206.00 166.00 24.00 343.76 Mokrya et al. (2010)
33 10.00 206.00 184.00 24.10 351.40 Mokrya et al. (2010)
34 10.00 207.00 202.00 24.00 351.36 Mokrya et al. (2010)
35 10.00 499.00 289.00 24.10 341.65 Mokrya et al. (2010)
36 10.00 500.00 335.00 24.20 352.81 Mokrya et al. (2010)
37 10.00 994.00 686.00 24.00 341.78 Mokrya et al. (2010)
38 10.00 1000.00 826.00 24.00 343.95 Mokrya et al. (2010)
39 10.00 1002.00 681.00 23.90 350.68 Mokrya et al. (2010)
40 10.00 1487.00 883.00 24.10 342.99 Mokrya et al. (2010)
41 10.00 1491.00 739.00 24.00 353.17 Mokrya et al. (2010)
42 10.00 1495.00 884.00 24.10 352.90 Mokrya et al. (2010)
43 26.00 600.00 300.00 26.00 334.48 Zhu et al. (2009)
44 26.00 600.00 300.00 30.00 336.59 Zhu et al. (2009)
45 26.00 600.00 400.00 30.00 291.64 Zhu et al. (2009)
46 17.00 1009.00 541.00 22.50 324.08 Pan et al. (2011)
47 10.00 450.00 450.00 23.00 307.68 Fei et al. (2013)
48 10.00 450.00 450.00 25.00 300.00 Fei et al. (2013)
49 10.00 458.80 455.50 23.00 307.49 Fei et al. (2013)
50 10.00 458.80 491.80 23.00 301.18 Fei et al. (2013)
51 10.00 458.80 536.60 23.00 303.15 Fei et al. (2013)
52 10.00 791.20 586.10 23.00 302.99 Fei et al. (2013)
53 10.00 791.20 696.30 23.00 305.72 Fei et al. (2013)
54 10.00 791.20 805.60 23.00 308.09 Fei et al. (2013)
55 10.00 791.20 893.20 23.00 308.71 Fei et al. (2013)
56 10.00 1190.00 931.00 23.00 300.61 Fei et al. (2013)
57 10.00 1190.00 1186.90 23.00 299.11 Fei et al. (2013)
58 10.00 1190.00 1228.70 23.00 298.23 Fei et al. (2013)
59 10.00 1500.00 1250.00 23.00 329.02 Fei et al. (2013)
60 10.00 1500.00 1250.00 25.00 329.02 Fei et al. (2013)
61 10.00 1500.00 1250.00 26.00 323.50 Fei et al. (2013)
62 12.00 1200.00 600.00 23.00 363.35 Feng et al. (2005)

3.4. Data reduction deterioration occurs) under a certain set of pipe diameters, pressures,
mass fluxes and inlet temperature. When the heat flux is beyond the
As shown in Table 3, the parameters of working conditions indicate onset power of heat transfer deterioration under that particular set of
the occurrence of the heat transfer deterioration phenomenon. When conditions, the case needs to be ignored and removed from the de-
the heating power increases and exceeds a certain value, the dete- terministic range. Hence for the supercritical water, the 62 groups of
rioration of heat transfer will be much more serious. A similar con- working conditions (of heat transfer deterioration) are eventually re-
clusion can also be made for decreasing mass flux values. Therefore, the duced to 40. The ultimately onset power of heat transfer deterioration is
model only needs to determine the minimum power (when heat transfer given as shown in Table 4.

294
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Table 4
Onset power of heat transfer deterioration in supercritical water.

No. Pipe diameter D Mass flux G Heat flux q Pressure p Inlet temperature Data Source
[mm] [kg/m2/s] [kW/m2] [MPa] T.in [°C]

1 9.4 1220 1260 22.754 233.91 Ackerman (1970)

2 20 1000 750 25 368.21 Herkenrath et al. (1967)
3 10 430 300 23.3 329.17 Shitsman (1963)
4 20.4 1400 1160 26.5 147.81 Vikhrev
5 10 1180 930 22.6 323.25 Yamagata et al. (1972)
6 7.6 1000 700 23 315.68 Gu et al. (2015)
7 7.6 1000 700 25 313.49 Gu et al. (2015)
8 10 1000 700 23 306.51 Gu et al. (2015)
9 10 1000 700 25 304.93 Gu et al. (2015)
10 26 900 400 26 311.16 Wang et al. (2009)
11 17 400 300 22.5 319.38 Pan et al. (2011)
12 17 600 284 22.5 323.14 Pan et al. (2011)
13 17 770 284 22.5 330.89 Pan et al. (2011)
14 17 770 284 28 324.93 Pan et al. (2011)
15 17 770 284 30 319.14 Pan et al. (2011)
16 17 820 284 28 326.00 Pan et al. (2011)
17 10 203 203 24.1 347.93 Mokrya et al. (2010)
18 10 206 166 24 343.76 Mokrya et al. (2010)
19 10 206 184 24.1 351.40 Mokrya et al. (2010)
20 10 207 202 24 351.36 Mokrya et al. (2010)
21 10 499 289 24.1 341.65 Mokrya et al. (2010)
22 10 500 335 24.2 352.81 Mokrya et al. (2010)
23 10 994 686 24 341.78 Mokrya et al. (2010)
24 10 1000 826 24 343.95 Mokrya et al. (2010)
25 10 1002 681 23.9 350.68 Mokrya et al. (2010)
26 10 1487 883 24.1 342.99 Mokrya et al. (2010)
27 10 1491 739 24 353.17 Mokrya et al. (2010)
28 10 1495 884 24.1 352.90 Mokrya et al. (2010)
29 26 600 300 26 334.48 Zhu et al. (2009)
30 26 600 300 30 336.59 Zhu et al. (2009)
31 17 1009 541 22.5 324.08 Jie et al. (2011)
32 10 450 450 23 307.68 Fei et al. (2013)
33 10 450 450 25 300.00 Fei et al. (2013)
34 10 458.8 455.5 23 307.49 Fei et al. (2013)
35 10 791.2 586.1 23 302.99 Fei et al. (2013)
36 10 1190 931 23 300.61 Fei et al. (2013)
37 10 1500 1250 23 329.02 Fei et al. (2013)
38 10 1500 1250 25 329.02 Fei et al. (2013)
39 10 1500 1250 26 323.50 Fei et al. (2013)
40 12 1200 600 23 363.35 Feng et al. (2005)

4. Calculation results and analysis is formed in the pipe. It would let the low density layer becomes thicker
near the wall. And this is much induce deterioration in heat transfer.
4.1. Correlation coefficient matrix The correlation coefficient of pressure and inlet temperature are also
negative, it indicates that when the pressure or inlet temperature is
Using partial least squares analysis method, the correlation coeffi- lower, the onset of heat transfer deterioration is larger. This is because
cient of experimental data in Table 4 is computed. The five variables when the pressure is smaller, the heat capacity is larger at the pseudo-
simple correlation coefficient matrix is obtained. It is shown in Table 5. critical point. So, it becomes easier for the fluid to absorb more heat.
It can be seen from the correlation coefficient matrix in Table 5, Therefore, more heat flux is required to trigger the heat transfer dete-
mass flux and the onset of heat transfer deterioration is positive cor- rioration phenomenon. When the inlet temperature is low, the over-
relation. It shows that the onset of heat transfer deterioration increases cooling of the inlet fluid is increased, so more heat is needed to make
with the increase of mass flux. The pipe diameter and the onset of heat the heat transfer deterioration phenomenon occur.
transfer deterioration is negative correlation. This is because when the
pipe diameter is smaller, the acceleration effect (due to the fluid ex-
pansion) is more apparent. Hence it is more conducive to enhances heat 4.2. Evaluation of existing criterion for onset of heat transfer deterioration
transfer. Similarly when the diameter is larger, a larger buoyancy force
The calculation results of the heat transfer deterioration criterion

Table 5
Correlation coefficient matrix.

Heat flux q Pipe diameter D Mass flux G Pressure p Inlet temperature

[kW/m2] [mm] [kg/m2/s] [MPa] T.in [°C]

Heat flux q [kW/m2] 1.00000000 −0.29018062 0.87473434 −0.16733964 −0.38146547

Pipe diameter D [mm] −0.29018062 1.00000000 −0.10131265 0.53725127 −0.13201445
Mass flux G [kg/m2/s] 0.87473434 −0.10131265 1.00000000 −0.01286586 −0.20984118
Pressure p [MPa] −0.16733964 0.53725127 −0.01286586 1.00000000 −0.07263195
Inlet temperature T.in [°C] −0.38146547 −0.13201445 −0.20984118 −0.07263195 1.00000000

295
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Fig. 3. Comparison of Vikhrev criterion with experimental data. Fig. 6. Comparison of Yamagata criterion with experimental data.

Fig. 4. Comparison of Styrikovich criterion with experimental data. Fig. 7. Comparison of Kim criterion with experimental data.

Fig. 5. Comparison of Yin criterion with experimental data. Fig. 8. Comparison of Mokry criterion with experimental data.

(shown in Table 1) are analyzed and compared with the experimental As it can be seen from the comparison presented from Figs. 3–11,
data. The analysis results can be seen from Figs. 3–11. The horizontal there is significant difference between the onset heat flux values of the
axis represents the experimental heat flux value of the onset of the heat calculations using various criterion and the experimental data. The
transfer deterioration. The vertical axis represents the calculated value Cheng and Yamagata criterion calculations yield larger power values
for each criterion of the onset of the heat transfer deterioration re- than the experimental heat transfer deterioration heat flux. Vikhrev,
spectively. Kim and Yin criterion generally produce smaller power values than the

296
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

also smaller for these 3 criterion. Thus, the Mokry, Li and Schatte cri-
terion seem to be the recommended for calculating the onset heat flux
of heat transfer deterioration in supercritical water.

4.3. Derivation of a new heat transfer deterioration criterion

The regression model (section 2.3) has been used for derivation a
new criterion to determine the onset heat flux value for heat transfer
deterioration in supercritical water. After optimization of the para-
meters [C1,eG,e di,ep,e Tin] using regression analysis, the optimization
constants are obtained, as shown in Eq. (17).

[C1,eG,ed ,ep,eT ]opt = [8255.2117,0.8325,−0.4958,−0.7486,−0.8125]

i in (17)

Based on the parameters given in Eq. (17), the new regression for-
mula for predicting the onset heat flux value of heat transfer dete-
rioration, is shown in Eq. (18):

Fig. 9. Comparison of Cheng criterion with experimental data. qDHT = 8255.2117·G 0.8325·d i−0.4958·p−0.7486 ·Tin−0.8125 (18)

where qDHT is the initial heat flux of heat transfer deterioration (kW/
m2), G is the mass flux (kg/m2s), d i is pipe diameter (mm), p is system
pressure (MPa) and Tin is system inlet temperature (°C).
The onset heat flux of heat transfer deterioration is calculated by Eq.
(18), and the calculation results are compared with experimental data,
as shown in Fig. 12. As it can be seen, the calculation results are basi-
cally evenly well distributed on both sides of the experimental value of
heat transfer deterioration. The deviation degree of calculation results
with the experimental data is relatively small. The margin of error is
basically within ± 15%.

4.4. Error analysis

Using the error analysis and calculation formula in section 2.4, the
standard deviation, average deviation, root mean square deviation and
average absolute deviation are calculated respectively. For each in-
vestigated criterion on the onset of heat transfer deterioration, the
comparisons of errors are shown in Table 6.
For comparison of average error and mean absolute deviation, bar
Fig. 10. Comparison of Li criterion with experimental data.
charts are plotted for each formula, as shown in Figs. 13 and 14, re-
spectively. Similarly, standard deviation and RMS deviation are com-
pared in Figs. 15 and 16, respectively.
Fig. 14 indicates the comparison of mean absolute relative deviation
for each criterion. Fig. 15 indicates the comparison of standard devia-
tion for each criterion. Among the existing nine criterion, the mean
relative error of Mokry calculation criterion is the smallest. It is also
relatively small for the Li and Schatte criterion. At the same time, Kim’s

Fig. 11. Comparison of Schatte criterion with experimental data.

experimental data. The Styrikovich criterion results in a well-dis-

tributed heat flux across the experimental data, with a large degree of
deviation from the experimental data. The results from Mokry, Li and
Schatte criterion are more accurate compared to the previously men-
tioned criterion. The deviation of values (from the experimental data) is
Fig. 12. Comparison new criterion with experimental data.

297
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

Table 6
Error comparisons of each investigated criterion.

Name of criterion Standard Mean Root Mean Mean

Deviation Relative Square Absolute
SD Deviation Deviation Relative
MRD RMSRD Deviation
MARD

Vikhrev 0.1999 −0.3869 0.1972 0.4044

Styrikovich 0.2898 −0.111 0.2859 0.2655
Yin 0.2313 −0.2904 0.2282 0.3384
Yamagata 0.4143 0.1749 0.4086 0.3199
Kim 0.1453 −0.7227 0.1433 0.7227
Mokry 0.3683 0.0085 0.3633 0.2754
Cheng 0.4504 0.2321 0.4443 0.3223
Li 0.3355 0.0079 0.331 0.2514
Schatte 0.3483 0.0233 0.3435 0.2172
New correlation 0.2117 0.020837 0.21004 0.1627

Fig. 15. Comparison of standard deviation for each criterion.

Fig. 13. Comparison of mean relative deviation for each criterion.

Fig. 16. Comparison of RMS deviation for each criterion.

criterion has the smallest standard deviation. Mokry’s criterion has a

relative high standard deviation. But Kim’s criterion has the largest
mean absolute relative deviation. The mean relative error of the new
fitting formula is 0.020837, and the mean absolute deviation is 0.1627,
which is the best among all of the other criterion. Comparing the mean
relative error of the new criterion to the existing criterion, it can be seen
that there are only two criterion smaller than the new: they are the
Mokry and Li criterion. Thus, the new one is still better than the Styr-
ikovich and Schatte criterion. The new criterion has the lowest mean
absolute deviation value compared to the existing criterion. Hence, the
calculation results of the newly fit regression formula is closer to the
investigated experimental data.
For supercritical water, on the other hand, there is still no unified
definition for declaring the deterioration of heat transfer. One of the
reasons for the discrepancy in different criterion may be the application
of different definition heat transfer deterioration in each case. In the
present study, the geometry was regarded to be a vertical, smooth
uniformly heated pipe. The roughness effect of the pipe was not con-
sidered. This is also one of the possible reasons for the deviation of
Fig. 14. Comparison of mean absolute relative deviation for each criterion.
calculated results. There is also a need to improve the prediction re-
gression model for the future research work. At the same time, the angle

298
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

of pipe inclination, its shape and the heat flux distribution are all of supercritical fluidsin circular tubes. Ann. Nucl. Energy 36, 1120–1128.
possible parameters, which may have an influence on the onset heat Eter, Ahmad, Groeneveld, Dé, Tavoularis, Stavros, 2016. An experimental investigation of
supercritical heat transfer in a three-rod bundle equipped with wire-wrap and grid
flux of heat transfer deterioration in supercritical water. The con- spacers and cooled by carbon dioxide. Nucl. Eng. Des. 303, 173–191.
sideration of all of such parameters is the subject of future studies. Fei, Wang, et al., 2013. Experimental research on heat transfer performance of super-
critical water in vertical tube. Atomic Energy Sci. Technol. 47 (6), 933–939.
Feng, Xu, et al., 2005. Experimental investigation to the heat transfer characteristics of
5. Conclusion water in vertical pipes under supercritical pressure. J. Xi′an Jiao Tong Univ. 39 (5),
468–471.
Based on the experimental data for heat transfer deterioration in Glushchenko, L.F., Gandzyuk, O.F., 1972. Temperature conditions at the wall of an an-
nular channel with internal heating at supercritical pressures. High Temp. 10 (4),
supercritical water, the distribution of experimental data are statisti- 734–738.
cally analyzed. On the basis of the criterion of heat transfer deteriora- Griem, H., 1996. A new procedure for the prediction of forced convection heat transfer at
tion and experimental data, the accuracy of various kinds of criterion is near and supercritical pressure. Heat Mass Transfer 31, 301–305.
Gu, H.Y., Zhao, M., Cheng, X., 2015. Experimental studies on heat transfer to supercritical
quantitatively analyzed. The application range of existing criterion is
water in circulartubes at high heat fluxes. Exp. Thermal Fluid Sci. 65, 22–32.
evaluated. The new empirical criterion has been formed on the basis of Herkenrath, H., Moerkenstein, P., Jung, U., Weckman, F., 1967. Waermeuebergangan
experimental data. The existing criterion are verified and analyzed with Wasserbeierzwungener Stroemungim Druckbereich von 140 bis250bar. EURATOM,
the experimental data. Berlin. EUR 3658 d.
Huang, Dan, Wu, Zan, Sunden, Bengt, Li, Wei, 2016. A brief review on convection heat
transfer of fluids at supercriticalpressures in tubes and the recent progress. Appl.
(1) Mokry, Li and Schatte’ s criterion have a smaller mean relative error Energy 162, 494–505.
and standard deviation, as compared with the experimental heat Jackson, J.D., 2013. Fluid flow and convective heat transfer to fluids at supercritical
pressure. Nucl. Eng. Des. 264, 24–40.
transfer deterioration data. In spite of Kim’s criterion has the Jäger, Wadim, Espinoza, Victor Hugo Sánchez, Hurtado, Antonio, 2011. Review and
smallest standard deviation, but it’s mean absolute relative devia- proposal for heat transfer predictions at supercritical waterconditions using existing
tion is the largest. So Mokry, Li and Schatte’s correlations are re- correlations and experiments. Nucl. Eng. Des. 241, 2184–2203.
Jie, Pan, et al., 2011. Experimental investigation on heat transfer characteristics of water
commended for calculating the onset heat flux of heat transfer de- in vertical upward tube under supercritical pressure. Nucl. Power Eng. 32 (1), 75–80.
terioration in supercritical water. Jingjing, Li, Tao, Zhou, Zhongyun, Ju, Qijun, Huo, Zejun, Xiao, 2014. Sensitivity analysis
(2) In supercritical water, the onset heat flux of heat transfer dete- of CHF parameters under flow instabilityby using a neural network method. Ann.
Nucl. Energy 71 (71), 211–216.
rioration will increase with the increase of mass flux, and it will Kim, J.K., Jeon, H.K., Yoo, J.Y., Lee, J.S., 2005. Experimental study on heat transferch-
become smaller when the pipe diameter or system pressure be- aracteristics of turbulent supercritical flow in vertical circular/non-circulartubes. In:
comes larger. For a detailed analysis, the changing trends of all of Proc of the 11th Nureth-11, Avignon, France.
Kirillov, P., Pometko, R., Smirnov, A., Grabezhnaia, V., Pioro, I., Duffey, R., Khartabil, H.,
these parameters (mass flux, pipe diameter, pressure and inlet
2005. Experimental study on heat transfer to supercritical water flowing in 1- and 4-
temperature) should be considered. This is a direction for estab- m-long vertical tubes. In: Proc. GLOBAL’05, Tsukuba, Japan, Oct. 9–13, p. 518.
lishing a more accurate and effective criterion, which can predict Koshizuka, S., Takano, N., Oka, Y., 1995. Numerical analysis of deterioration phenomena
the onset heat flux of heat transfer deterioration in supercritical in heat transfer to supercritical water. Int. J. Heat Mass Transfer 38 (16), 3077–3084.
Lee, R.A., Haller, K.H., 1974. Supercritical water heat transfer developments and appli-
water in order to enhance the calculation accuracy. cations. In: Proceedings of the 5th International Heat Transfer Conference, vol. IV,
(3) On the whole, the new calculation model (for predicting heat Tokyo, Japan, September 3–7, pp. 335–339, Paper B7.7.
transfer deterioration in supercritical water) has considered the Li, Z., Zhang, D., Wu, Y., Lu, J., Liu, Q., 2014. A new criterion for predicting deterior-
ationof heat transfer to supercritical water in smooth tubes. In: CSEE, pp. 6304–6309.
influence of mass flux, pipe diameter, pressure and inlet tempera- Liu, Lei, Xiao, Zejun, Yan, Xiao, Zeng, Xiaokang, Huang, Yanping, 2013b. Numerical si-
ture. It has a relatively simple structure for the proposed criterion. mulation of heat transfer deterioration phenomenonto supercritical water in annular
Compared with the existing criterion of heat transfer deterioration, channel. Ann. Nucl. Energy 53, 170–181.
Liu, Lei, Xiao, Zejun, Yan, Xiao, Zeng, Xiaokang, Huang, Yanping, 2013a. Heat transfer
the calculation results of the new criterion have a smaller mean deterioration to supercritical water in circular tube and annular channel. Nucl. Eng.
relative deviation and mean absolute deviation. But there is still a Des. 255, 97–104.
requirement of more experimental data, in order to assess and im- Loewenberg, Matthias F., Laurien, Eckart, Class, Andreas, Schulenberg, Thomas, 2008.
Supercritical water heat transfer in vertical tubes: a look-up table. Progr. Nucl.
prove this formula in the future.
Energy 50, 532–538.
Mokry, Sarah, Pioro, Igor, Kirillov, Pavel, Gospodinov, Yevgeniy, 2010. Supercritical-
Author introduction water heat transfer in a vertical bare tube. Nucl. Eng. Des. 240, 568–576.
Mokry, Sarah, Pioro, Igor, Farah, Amjad, King, Krysten, Gupta, Sahil, Peiman, Wargha,
Kirillov, Pavel, 2011. Development of supercritical water heat-transfer correlation for
Ma Dongliang (1982-), Ph. D. Student, North China Electric Power vertical baretubes. Nucl. Eng. Des. 241, 1126–1136.
University, nuclear science and engineering, nuclear thermal hydraulic Ornatskiy, A.P., Glushchenko, L.F., Gandzyuk, O.F., 1972. An experimental study of heat
and safety. transfer in externally-heated annuli at supercritical pressures. Heat Transfer-Soviet
Res. 4 (6), 25–29.
Zhou Tao (1965-), vice president of the International Education Pan, Jie, Yang, Dong, Dong, Zichun, Zhu, Tan, Bi, Qincheng, 2011. Experimental in-
Institute of North China Electric Power University, North China Electric vestigation on heat transfer characteristics of low mass fluxrifled tube with upward
Power University, doctoral tutor, Professor of nuclear science and en- flow. Int. J. Heat Mass Transfer. 54, 2952–2961.
Pioro, Igor L., Duffey, Romney B., 2005. Experimental heat transfer in supercritical wa-
gineering, nuclear thermal hydraulic and safety. terflowing inside channels (survey). Nucl. Eng. Des. 235, 2407–2430.
Podila, K., Rao, Y.F., 2015. CFD analysis of flow and heat transfer in Canadian super-
Acknowledgements critical waterreactor bundle. Ann. Nucl. Energy 75, 1–10.
Schatte, Gerrit A., Kohlhepp, Andreas, Wieland, Christoph, Spliethoff, Hartmut, 2016.
Development of a new empirical correlation for the predictionof the onset of the
This research is financially supported by Nuclear Reactor Thermal deterioration of heat transfer to supercriticalwater in vertical tubes. Int. J. Heat Mass
Hydraulic Technology Key Laboratory fund (20130901), Key Transfer 102, 133–141.
Sharabi, M., Ambrosini, W., He, S., Jackson, J.D., 2008. Prediction of turbulent con-
Laboratory of Nuclear Reactor System Design Technology
vective heat transfer to a fluid at supercritical pressure in square and triangular
(2015BJ0151). channels. Ann. Nucl. Energy 35, 993–1005.
Shitsman, M.E., 1963. Impairment of the heat transmission at supercritical pressures.
Teplofiz. Vysokikh Temp. (High Temperatures) 1 (2), 237–244 267–275.
References
Styrikovich, M.A., Margulova, T.K., Miropol’skii, Z.L., 1967. Problems in the development
of designs of supercritical boilers. Therm. Eng. 14 (6), 5–9.
Ackerman, J.W., 1970. Pseudoboiling heat transfer to supercritical pressure waterin Swenson, H.S., Carver, J.R., Kakarala, C.R., 1965. Heat transfer to supercritical waterin
smooth and ribbed tubes. Heat Transfer Trans. ASME 92 (3), 490–498 (P69-WA/ smoothbore tubes. J. Heat Transfer Series C 87, 477–484.
HT-2). Vikhrev, Yu.V., Barulin, Yu.D., Kon’kov, A.S., 1967. A study of heattransfer in vertical
Backhaus, K., Plinke, W., Erichson, B., et al., 2008. Multivariate Analysemethoden 7, tubes at supercritical pressures. Therm. Eng. 14 (9), 116–119.
Auflage. Springer-Lehrbuch. Wang, Han, Bi, Qincheng, Yang, Zhendong, Wang, Linchuan, 2015. Experimental and
Cheng, X., Yang, Y.H., Huang, S.F., 2009. A simplified method for heat transfer prediction numerical investigation of heat transfer from a narrowannulus to supercritical

299
M. Dongliang et al. Nuclear Engineering and Design 326 (2018) 290–300

pressure water. Ann. Nucl. Energy 80, 416–428. vertical tube with concentrated incident solar heat flux on one side. Int. J. Heat Mass
Wang, Jianguo, Li, Huixiong, Guo, Bin, Shuiqing, Yu, Zhang, Yuqian, Chen, Tingkuan, Transfer. 95, 944–952.
2009. Investigation of forced convection heat transfer of supercritical pressurewater Zhang, Ge, Zhang, Hao, Gu, Hanyang, Yang, Yanhua, Cheng, Xu, 2012. Experimental and
in a vertically upward internally ribbed tube. Nucl. Eng. Des. 239, 1956–1964. numerical investigation of turbulent convective heat transfer deterioration of su-
Wen, Q.L., Gu, H.Y., 2011. Numerical investigation of acceleration effect on heat transfer percritical water in vertical tube. Nucl. Eng. Des. 248, 226–237.
deteriorationphenomenon in supercritical water. Progr. Nucl. Energy 53, 480–486. Zhu, Xiaojing, Bi, Qincheng, Yang, Dong, Chen, Tingkuan, 2009. An investigation on heat
Yamagata, K., Nishikawa, K., Hasegawa, S., Fuji, T., Yoshida, S., 1972. Forced con- transfer characteristics of different pressuresteam-water in vertical upward tube.
vectiveheat transfer to supercritical water flowing in tubes. Int. J. Heat Mass Transfer Nucl. Eng. Des. 239, 381–388.
15, 2575–2593. Zhu, X.J., Bi, Q.C., Chen, T.K., 2007. An investigation on heat transfer characteristics of
Yildiz, S., Groeneveld, D.C., 2014. Diameter effect on supercritical heat transfer. Int. Steam-water at different pressure in vertical upward tube. In: 3rd International
Commun. Heat Mass Transfer 54, 27–32. Symposium on SCWR-Design and Technology, Shanghai, China, March 2007, 12–15.
Yin, F., Chen, T.K., Li, H.X., 2006. An investigation on heat transfer to supercritical
waterin inclined upward smooth tubes. Heat Transfer Eng. 27, 44–52.
Yoshida, S., Mori, H., 2000. Heat transfer to supercritical pressure fluids flowing in tubes. Further reading
In: Proc. First Int. Symposium on SCWR Design and Technology. University of Tokyo,
pp. 72–78. Duffey, Romney B., Pioro, Igor L., 2005. Experimental heat transfer of supercritical
Zahlan, H., Tavoularis, S., Groeneveld, D.C., 2015. A look-up table for trans-critical heat carbon dioxideflowing inside channels (survey). Nucl. Eng. Des. 235, 913–924.
transfer in water-cooled tubes. Nucl. Eng. Des. 285, 109–125.
Zhang, G., Li, Y., Dai, Y.J., Wang, R.Z., 2016. Heat transfer to supercritical water in a

300