Annals of Nuclear Energy 138 (2020) 107194

Contents lists available at ScienceDirect

Annals of Nuclear Energy

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

Research of quasi 2-D refined simulation and modeling for deaerator

in nuclear power plant
Meijie Gong, Minjun Peng ⇑, Haishan Zhu
Fundamental Science on Nuclear Safety and Simulation Technology Laboratory, Harbin Engineering University, Harbin 150001, China

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

Article history: Deaerator is an important component which is used widely in secondary loop system of nuclear power
Received 23 August 2019 plant. In this work, the quasi 2-d simulation model for deaerator is developed. To make the simulation
Received in revised form 18 October 2019 and modeling process more detailed and refined, some reasonable assumptions are proposed for math-
Accepted 1 November 2019
ematical modeling. There are two different zones of spraying zone and bubbling zone with more nodes.
Available online 12 November 2019
Simulating with practical physical process, the calculation process could demonstrate parameters distri-
bution. In the steady state calculation and comparison, the simulation model is in good agreement com-
Keywords:
paring with design data onto actual nuclear power plant deaerator. The dynamic processes are validated
Deaerator
Direct contact heat transfer
by JTopmeret with main parameters. Two typical different dynamic processes are used to analyze deaer-
Quasi 2-D ator operation characteristics. The refined different parameters distribution is calculated with time
Parameter distribution changing. It is real-time simulation model with such division that could guarantee the calculation speed
Dynamic process and accuracy at the same time.
Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction non-linear dynamic model (Katebi, 2007), one-node model

(Roldan-Villasana and Vazquez, 2010) are developed for the deaer-
Deaerator is an important component in nuclear power plant ator used in power plants system simulation. Oko and Wang
secondary loop system. Different from surface tube-and-shell heat (2014) adopted first principle of mass-heat balance and built the
exchanger as condenser and feed water heater exchanger, deaera- deaerator model as mixing tank. Xue and Sun (2014) built the
tor is a direct contact heat exchanger. It could heat the condensing deaerator model with special structure based on single-pressure
and feeding water to saturated state with heating steam from tur- node during different dynamic operation. Wang et al. (2016a) trea-
bine or steam generator. The condensing and feeding water are ted deaerator model as a source item for the part of system simu-
coming from high-pressure feed water heat exchanger’s drain lation. Alobaid et al. (2017) considered deaerator as hydraulic
water and low-pressure feed water heat exchanger’s feed water accumulators based on mass-energy balance under different states.
respectively. The cold and hot working medium is mixing in deaer- The lumped method for modeling deaerator in different power
ator which is direct contact heat transfer process rather than heat plant systems was utilized by Dong et al. (2018a,b). Some code
transfer through tube surface. In general, there are not many sen- and software could be used for the power plant deaerator simula-
sors installed inside the actual deaerator. So, it is difficult to obtain tion. Wan et al. (2015) considered the impact of deaerator pressure
detailed and distributional parameters of deaerator during opera- on system response in AP1000 steam supply system built by
tion. In traditional modeling methods for deaerator, the lumped MATLAB/Simulink. Chen el al. (2017a) used Dymola code with
parameter method and heat balance method are usually adopted. Modelica language to simulate deaerator in power plant system
However, the deaerator simulation model built by those methods modeling. Wang et al. (2018) used the lumped parameter method
could not reflect the internal characteristics and the dynamic pro- to build two-zone deaerator model in Simulink. Zhou and Novog
cess with detailed parameters distribution. (2018) considered the deaerator tank effect during system model-
There are some researches of different simulating and modeling ing by RELAP5 code. Zhang et al. (2018) built the deaerator three-
methods for the deaerator and direct contact heat exchanger in zones model based on the thermodynamic balance in Matlab/
different thermal systems. Two-tank model (Lu et al., 1998), Simulink. Li et al. (2019a) used TESS library in TRNSYS to build
the deaerator model for system modeling. STAR-90 simulation
platform could be used for modeling deaerator in system simula-
⇑ Corresponding author. tion by Li et al. (2019b).
E-mail address: [email protected] (M. Peng).

https://doi.org/10.1016/j.anucene.2019.107194
0306-4549/Ó 2019 Elsevier Ltd. All rights reserved.
2 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

Nomenclature

l viscosity, Pa
s
Variables p circumference ratio
A area, m2 q density of the fluid, kg/m3
d sphere diameter, m
G flow rate, kg/s Subscripts
h enthalpy, kJ/kg bb bubbling zone
Ja Jacob number c condensing process
K heat transfer coefficient, W/(m2
K) cs cross-sectional
Lv liquid level, m dea deaerator
Nu Nusselt number in node inlet
n number of bubble or droplet out node outlet
P pressure, MPa s steam phase
Pr Prandtl number sp spraying zone
Q heat transmission, kW w water phase
r latent heat of vaporization, kJ/kg x horizontal direction
Re Reynolds number y vertical direction
T temperature,
t time, s Superscripts
Dt time step, s n current moment
V volume, m3 n+1 next moment
e flow resistance coefficient
k thermal conductivity, W/(m
K)

There are some researches about the thermal and performance mass-heat balance model. Some other researches would take
analysis, design optimization for deaerator and direct contact heat deaerator as the basic unit rather than specific object with dynamic
exchanger. Rodrı́guez-Toral et al. (2000) developed equation- process in large system modeling for responding analysis. The sim-
oriented mathematical model for deaerator optimization with ulation of direct contact heat exchangers mainly includes system
current design practice. Srinivas (2009) aimed at efficiency and modeling code and analysis, CFD-based detailed local phenomenon
conducted analysis of different deaerator locations based on ther- simulation, performance analysis based on thermodynamic.
modynamics. Wang et al. (2011) purposed RBF neural network In this work, the quasi 2-D refined simulation model of deaera-
method for deaerator liquid level control. Nomura et al. (2013) tor is developed. This is the simulation based on actual physical
studied on the thermal performance of other type direct contact processes. There are two zones corresponding to the actual struc-
heat exchanger based on design. Wang et al. (2016b) used the sta- ture. Then, more nodes are divided for simulating parameters dis-
tic model and processing model for deaerator based on steam–wa- tribution. Reasonable assumptions are also proposed based on
ter mass-energy balance matrix which is used for the system actual physical process with droplets and bubbles. According to
analysis in the case of condensate throttling. Wang et al. (2017) these assumptions, the generic mathematical model is derived
carried out thermal performance analysis for combined heat and for node in each zone. It is a real-time simulation model with such
power plants with direct contact heat exchanger based on division which means it could ensure the speed and accuracy of the
steady-state energy balance. Kowalczyk et al. (2019) conducted calculation at the same time respectively. There is parameters dis-
the energy and exergy analysis of overall system with deaerator tribution during two typical dynamic processes that could use for
as one of the devices. There are some researches about the detailed characteristics analysis. The deaerator model developed in this
process in deaerator and direct contact heat exchanger by experi- work could provide support for nuclear power plant digital simula-
ment and computational fluid dynamics method. Cascella and tion, operation, design, etc. Detailed deaerator modeling and simu-
Teyssedou (2015) built the thermodynamic model to estimate lation could provide precise and detailed boundary for system
the heat transfer characteristics in direct contact heat exchanger, analysis.
which is focusing on droplets size distribution and heat transfer
process. Shukla et al. (2017) used CFD method to carry out detailed
simulation calculation for steam condensation, which is partial 2. Modeling and simulation methodology
process of direct contact heat exchanger. Baqir et al. (2018)
summarized the direct contact heat transfer model with one- 2.1. Physical process and model
dimensional flow model accompanied with experimental measure-
ments. Fu et al. (2018) implemented numerical study for direct For the actual deaerator as represented in Fig. 1(a), there are
contact heat exchanger through steady-state flow field analysis two zones. One zone is spraying zone located in the upper half of
by CFD method compared with experiment. Some other type direct deaerator and the other is bubbling zone located in the lower half.
contact heat exchanger was studied both by experiment and the 3- In spraying zone, the condensing and feeding water is coming from
D steady-state numerical model. Zeng et al. (2018) adopted such the inlet by the nozzle in the top of deaerator. The water would
means to simulate flow and heat transfer characteristics. Yang become large number of droplets. The condensing and feeding
et al. (2019) also studied on the direct contact heat transfer char- water liquid droplets are heated in spraying zone by the steam
acteristics of deaerator by experimental method. around in space. The condensing and feeding water flows from
From the review above, most researches have focused on the top to bottom. In bubbling zone, the heating steam would become
system modeling and analysis including deaerator. Some modeling steam bubbles by the bubbling tube located in the bottom of the
methods utilized the single-pressure lumped parameter model or deaerator. Bubbling tube is the tube with small holes to generate
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 3

2 3

10

8 7
6

9
(a) Deaerator structure and diagram
1-Spraying zone; 2-Condensing and feeding water inlet; 3-Spraying nozzle; 4-Liquid droplets flow; 5-
Bubbling zone; 6-Steam inlet; 7-Bubbling tube; 8-Steam bubbles flow; 9-Deaerator outlet; 10-Liquid
level.
1

7 Liquid
level
8

15

Droplet flow 1 8 15

1 in 3 out 1 in 2 out 1 in 1 out 2 in 1 out 1 in 3 out 2 in 2 out

Spraying zone inlet
Spraying z one edge node Spraying z one inne r node
Bubble flow

1 in 3 out 1 in 2 out 1 in 1 out 2 in 1 out 1 in 3 out 2 in 2 out

Bubbling zone inlet Bubbling zone edge node Bubbling zone inner node
(b) Deaerator simulation model node division
Fig. 1. Deaerator diagram and node division.
4 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

steam bubbles. The heating steam could heat the water around in zone, the droplets diameter would increase. There is the similar
bubbling zone. The steam flows from bottom to top. Some steam process for the steam bubble number in bubbling zone. With
that not condenses in this zone would flow up to spraying zone. the steam condensing process in bubbling zone, the bubble
The condensing and feeding water are flowing out from deaerator diameter would decrease. The number of droplets and bubbles
bottom. In order to correspond to the actual situation, there are would not increase or disappear, and the number moving into
also two obvious practical zones divided in the deaerator simula- different directions would also follow the conservation related
tion model with such actual physical process. The deaerator dia- to flow rate. The diameter of droplets and bubbles would affect
gram with zones and nodes division is represented in Fig. 1(b). In by the condensation process respectively, and the diameter
spraying zone there is only the inlet node which is used to generate would change according to the heat transfer process during
liquid droplets. In bubbling zone there is only the inlet node which dynamic process. This could reflect the deaerator local
is used to generate steam bubbles. There are two zones and characteristics.
15  15 nodes division which is 2-D division with flow in each 7. The thermal parameters such as pressure, enthalpy and density
zone and node. In the initial state, spraying zone is composed of are calculated inside the node. The flow rate is calculated
7 rows and bubbling zone is composed of 8 rows. between nodes.

2.2. Basic assumptions for deaerator Through these above reasonable assumptions, the deaerator
model could mathematically establish. The simulation model could
Based on the above actual physical processes, some basic better conform to the actual physical process. The actual physical
assumptions could be made for the deaerator model developed phenomena can simulate both based on time scale and spatial
in this work. These could make the simulation model more authen- scale with different parameters distribution. It could further indi-
tic. There are these reasonable model assumptions: cate the applicability of the simulation model.

1. In spraying zone, the condensing and feeding water is becoming 2.3. Mathematical modeling
into lots of tiny water droplets by the nozzle locating at the con-
densing and feeding water inlet. So, the condensing and feeding From the actual physical process and basic assumptions as
water could regard as the droplets flow. The droplet relative mentioned above, there is the detailed mathematical modeling
velocity is faster than the steam around, which means the process for nodes in each zone. For different nodes in each zone,
steam in this zone is in relatively static state. The motive single there are only the differences of flow-in and flow-out as legends
phase flow of liquid droplets in this zone could consider as represented in Fig. 1(b), but the basic form is the same and generic.
mainstream fluid. The parameters of liquid droplets flow could By the water and steam physical property with IAPWS-IF97
represent the parameters of this zone. The droplets are in sub- (Wagner et al., 2000), the physical properties of different parame-
cooling state. ters could be calculated depending on the physical state.
2. There is the similar assumption for bubbling zone. The steam is
becoming into lots of tiny steam bubbles by the bubbling tube 2.3.1. Spraying zone modeling and derivation
located at the inlet of deaerator heating steam. The steam bub- Based on the 1st assumption, the droplets flow in this zone is
ble in the bubbling zone could regard as bubbles flow. The treated as the mainstream fluid. For each node in spraying zone,
velocity of steam is faster than water in this zone. The water there is the following generic parameter expression from mathe-
in this zone is in static state relative to steam. There is the matical derivation.
motive single phase flow of steam bubbles in this zone which The basic mass conservation equation (Patankar, 1980) for the
could consider as mainstream fluid. The parameters of bubbles liquid droplets flow in spraying zone is represented in Eq. (1):
flow could represent the parameters of bubbling zone. The bub-

   
bles are in saturated state. @ qsp Asp;cs @ Gsp þ Gsp;c @ Gsp þ Gsp;c
3. In these two zones, the droplets and bubbles flow could decom- þ þ ¼0 ð1Þ
@t @x @y
pose into horizontal(x) and vertical(y) directions respectively.
The heat transfer area is calculated by the outer surface of the After the integration of the space location, the expression with-
droplets and bubbles. The flow rate of droplets and bubbles out spatial variable could obtain.
could calculate from pressure difference between nodes. By

considering the droplet and bubble flow as whole at relatively d qsp V sp
X 
macroscopic perspective holistically, the main flow directions þ Gsp  Gsp;c ¼ 0 ð2Þ
dt
are decomposed. It is used as quasi 2-D calculation expansion
Converting the differential in the Eq. (2) to difference form
and parameters distribution calculations.
could have the following expression:
4. The change of the liquid level would influence the boundary of
two zones during dynamic process. Meanwhile, the mutual

D qsp V sp
X 
transformation of two zones would lead to the change in liquid þ Gsp  Gsp;c ¼ 0 ð3Þ
level. Dt
5. The heat transfer process in spraying zone could be seen as the For difference term in Eq. (3), there is the following expression
steam condensing process outside the sphere. The heat transfer and expansion to get the variation of volume (DVsp) and density
process in bubbling zone could be seen as the steam bubble (Dqsp).
condensing in subcooling water. There are corresponding P 
empirical relationships used to calculate the heat transfer coef- nþ 1 Gsp  Gsp;c
DV sp ¼ V sp  V nsp ¼ Dt ð4Þ
ficient in both zones. In each node, the diameter of droplet or qnsp
bubble is the same.
6. The total number of droplets in spraying zone would not change Dqsp ¼ qnspþ 1  qnsp ð5Þ
without the inlet flow changing. The droplets number in each
node is calculated based on flow. The interaction between dro- Linearizing and expanding the difference term in Eq. (3) with
plets is ignored. With the steam condensing process in spraying that in Eqs. (4) and (5),
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 5


nþ1
n nþ1 n
D qsp V sp ¼ qsp V sp  qsp V sp hsp ¼ hsp þ Dhsp ð17Þ




In spraying zone, it is in subcooled status. Other parameters as
¼ V spnþ 1
 V nsp qnsp þ qnspþ 1  qnsp V nsp þ o Dqsp
DV sp ð6Þ
pressure (Psp) could be calculated by density (qsp in Eq. (9)) and
In Eq. (6) the second-order terms could ignore. Then there is the enthalpy (hsp in Eq. (17)).
expansion of Eq. (3) by that.





X  Psp ¼ f qsp ; hsp ð18Þ
V nspþ 1  V nsp qnsp þ qnspþ 1  qnsp V nsp þ Gsp  Gsp;c Dt
Base on the 7th assumption, the general relationship (Batchelor,
¼0 ð7Þ 2010; Kundu et al., 2008) between the pressure difference (DPsp)
and the flow (Gsp) between nodes is represented as following:
So, here is the density change (Dqsp) and density (qsp) expres-

sion in spraying zone derived from above equations. DPsp ¼ f esp ; G2sp ð19Þ
P 
Dt
Gsp  Gsp;c þ q
DV sp
n
Dqsp ¼ 
sp
ð8Þ Since the physical parameters of adjacent nodes would not
V nsp change much, the flow resistance coefficient could consider basi-
cally the same. Thus, the flow rate could calculate from the pres-
qnspþ 1 ¼ qnsp þ Dqsp ð9Þ sure difference proportional allocation. So, the flow rate is
considered related to the square root of the pressure difference
The basic energy conservation equation (Patankar, 1980) for the
between nodes. The flow rate between the nodes could acquire
liquid droplets flow in spraying zone is represented in Eq. (10). The
according to the sum of the pressure difference and the pressure
droplets flow is heated in spraying zone which means it would
difference in horizontal or vertical directions respectively. In spray-
absorb heat (Qsp).
ing zone, steam is condensing around the droplets. The flow


   
@ qsp hsp Asp @ Gsp hsp @ Gsp hsp @Q sp @Q sp
through one node needs to add by the condensing rate (Gsp,c).
þ þ þ þ ¼0 ð10Þ Based on the 3rd assumption, the flow rate (Gsp) for each direction
@t @x @y @x @y between nodes could get.
After integrating the spatial position, the transform without pffiffiffiffiffiffiffiffiffiffi
DPsp  
spatial position could get. Gsp;out ¼ P pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Gsp;in þ Gsp;c ð20Þ

 DP sp;x;y
d qsp V sp hsp
X   
þ Gsp hsp  Q sp ¼ 0 ð11Þ
dt 2.3.2. Bubbling zone modeling and derivation
Converting differential term to difference term in Eq. (11), Based on the 2nd assumption, the bubbles flow is treated as the

 mainstream fluid in bubbling zone. For each node in this zone,
D qsp V sp hsp
X    there is the following generic parameter expression derivation.
þ Gsp hsp  Q sp ¼ 0 ð12Þ The basic mass conservation equation form of bubbling zone is
Dt
almost the same as that in spraying zone. The density change
For the enthalpy (hsp), there is the similar form like Eq. (5) to get
(Dqbb) could derive in Eq. (23) with the same processing method.
the enthalpy variation (Dhsp):      
nþ1 n @ qbb Abb;cs @ Gbb  Gbb;c @ Gbb  Gbb;c
Dhsp ¼ hsp  hsp ð13Þ þ þ ¼0 ð21Þ
@t @x @y
Linearizing and expanding the difference term in Eq. (12) with P 
Eq. (4), Eq. (5) and Eq. (13), Gbb þ Gbb;c
DV bb ¼  Dt ð22Þ


nþ1
n qnbb
D qsp V sp hsp ¼ qsp V sp hsp  qsp V sp hsp ¼



 P 
Dt
Gbb þ Gbb;c þ qnbb
DV bb
sp V sp qsp hsp þ qsp  qsp V sp hsp þ hsp hsp V sp qsp
n n n n nþ1 n
V nþ1 Dqbb ¼ 
n nþ1 n n n
ð23Þ


 
 V nbb

þo Dqsp
DV sp þo Dqsp
Dhsp þo Dhsp
DV sp þo Dqsp
DV sp
Dhsp
The basic energy conservation equation in bubbling zone is rep-
ð14Þ resented in Eq. (24). The bubble is condensing in bubbling zone
which means it would release heat (Qbb). This is different from
In Eq. (14), the second-order and low-order terms could ignore.
spraying zone.
There is the expansion of Eq. (12) combining with Eq. (14).  


 @ ðqbb hbb Abb Þ @ ðGbb hbb Þ @ ðGbb hbb Þ @Q bb @Q bb
þ þ  þ ¼0 ð24Þ
sp  V sp qsp hsp þ qsp  qsp V sp hsp
n n n n
V nþ1 n nþ1 n
@t @x @y @x @y

 hX
 i
þ hsp  hsp V nsp qnsp þ
nþ1 n n
Gsp hsp  Q sp Dt After the same processing method as that in spraying zone,
there is enthalpy change (Dhbb) in bubbling zone.
¼0 ð15Þ P  n 
 n 
Gbb hbb þ Q nbb Dt þ DV bb qnbb hbb þ Dqbb V nbb hbb
n

Transforming the Eq. (15) with Eq. (4) and Eq. (5), the enthalpy Dhbb ¼ 
V bb qnbb
n
change (Dhsp) of droplets in spraying zone and enthalpy expression
(hsp) could derive: ð25Þ
hP
 i
 In bubbling zone, it is in saturated state. Thermal parameters as
Gsp hsp  Q nsp Dt þ DV sp qnsp hsp þ Dqsp V nsp hsp
n n n
pressure (Pbb) could obtain by only one calculated parameter.
Dhsp ¼  ð16Þ
V nsp qnsp There is the pressure (Pbb) expression for each node in bubbling
zone from enthalpy change (Dhbb) calculation.
6 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

@P  2
  30:25
Pnþ1 n
bb ¼ P bb þ P ¼ Pnbb Dhbb ð26Þ k3sp;w qsp;w qsp;w  qsp;s g hsp;s  hsp;w
@h K sp ¼ 0:8154   5 ð30Þ
lsp;w d T sp;s  T sp;w
For the flow in bubbling zone, there is the same processing
method as that in spraying zone. Different from spraying zone,
Based on the 6th assumption, the number of droplets (nsp) in
the bubble is condensing which means the steam becoming into
one node could calculate from the flow and droplet diameter in
water. The flow rate (Gbb) through one node needs to subtract from
upstream node in Eq. (31).
the condensing rate (Gbb,c) with pressure difference (DPbb) distribu- P
tion calculation. Gsp;in
nsp ¼ ð31Þ

 pq 3
sp;in Dtdsp;in
DPbb ¼ f e 2
bb ; Gbb ð27Þ 6

The heat transfer area (Abb) is calculated by the number and

pffiffiffiffiffiffiffiffiffiffi outer surface of droplets.
DPbb  
Gbb;out ¼ P pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Gbb;in  Gbb;c ð28Þ
DPbb;x;y Abb ¼ nsp pdsp;in
2
ð32Þ

The condensing flow rate (Gsp) is calculated from following

2.3.3. Liquid level variation model expression.
The deaerator liquid level (Lv) is calculated according to the
flow rate through deaerator as represented in Eq. (29). Q sp;c
Gsp;c ¼ ð33Þ
r sp
RGdea
Lv nþ1 ¼ Lv n þ Dt ð29Þ
qAcs The droplets diameter (dsp) in one node could calculate by the
flow through this node.
Based on the 4th assumption for the simulation model, the sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P ffi
change of liquid level would cause the change of boundary 3 Dt
Gsp;in þ Gsp;c
dsp ¼ 6 ð34Þ
between spraying zone and bubbling zone during dynamic process. qsp nsp p
As represented in Fig. 2, there are the liquid level variation and
zones transform process. Under steady state, the liquid level is In bubbling zone, the bubbles are also considered as sphere.
located at the bottom row of spraying zone. During calculation pro- Based on the 5th assumption the heat transfer coefficient (Kbb)
cess with boundary parameters changing, if the change of liquid would utilize steam bubble condensing in water empirical rela-
level is maintained in a row of spraying zone, no boundary move- tionship (Kim and Park, 2011):
ment and zones transform would occur. If the liquid level rises to 0:4564 0:2043
Nubb ¼ 0:2575Re0:7
bb Pr bb Jabb ð35Þ
the upper limit of the spraying zone bottom row and above, then
the bottom row of spraying zone (7th row) would transform into Nubb kbb
bubbling zone. On the contrary, if the liquid level falls to the bub- K bb ¼ ð36Þ
dbb
bling zone top row upper limit and below, then the top row of bub-
bling zone (8th row) would transform into spraying zone. If the In bubbling zone there is similar calculation process. The differ-
liquid level continues to rise or fall, more rows of spraying zone ence is that the bubble diameter (dbb) is reduced due to the con-
and bubbling zone would transform to each other respectively. densation of steam into water during steam bubble condensing
With the liquid level changing, the boundary would change and process. In Eq. (40) the condensing flow rate (Gbb,c) is subtracted
zones would transform to each other. The calculation in other rows which is calculated in Eq. (39).
and nodes untransformed would stay constant. P
Gbb;in
nbb ¼ ð37Þ
pq 3
2.3.4. Heat transfer model used in deaerator 6 bb;in Dtdbb;in

In spraying zone, the droplets are considered as sphere. Based

Abb ¼ nbb pdbb;in
2
on the 5th assumption, the steam is condensing outside the sphere. ð38Þ
So, the heat transfer coefficient (Ksp) expression would use Nusselt
sphere external surface condensation heat transfer coefficient Q bb;c
Gbb;c ¼ ð39Þ
empirical relationship (Holman, 2010). r bb

6 Spraying zone
Liquid level rises to the upper
limit of spraying zone, a row
of spraying zone is
transformed into a row of
Spraying zone bubbling zone
bottom row 7

Liquid level

iquid level falls to the upper

mit of bubbling zone, a row
Bubbling zone
of bubbling zone is 8
top row
transformed into a row of
Bubbling zone
spraying zone

Fig. 2. Liquid level variation and zone transform diagram.

 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 7

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 
3 Dt
Gbb;in  Gbb;c and liquid level (Lv) are the important indicators for measuring
dbb ¼ 6 ð40Þ deaerator operating state. The calculation of deaerator outlet flow
qbb nbb p
(Gdea,out) could illustrate the conservation of inlet and outlet in
steady state. There is the actual time comparing with simulating
2.3.5. Simulation object time which could demonstrate the simulation model calculation
In this work, the deaerator in Daya Bay 900 MW nuclear power speed. The calculation takes less time than actual time. From the
plant (Guangdong Nuclear Power Training Center, 2005) secondary comparison, the simulation model built in this work could reach
loop system is chosen as the simulation object. The main parame- real-time calculation speed and some more fast than real-time.
ters of structure and boundary are listed in Table 1.
The deaerator condensing and feeding water inlet is the sum of 3.2. Dynamic process simulation and analysis
feed water and drain water, and the total inlet enthalpy is the
weighted average enthalpy of the two fluid, which is calculated To validate the simulation model built in this work during
as one fluid. dynamic process, the benchmark deaerator simulation model is
built in JTopmeret (GSE systems Inc., 2006a,b). It is used for com-
2.4. Calculation diagram parison of macroscopic thermal parameters with time changing
and dynamic process. JTopmeret is a software which is used widely
First, the zones division and nodes division are performed for for different thermal system modeling. Gong et al. (2019) used
deaerator simulation model as represented in Fig 1(b). The bound- JTopmeret to validate the simulation model built for condenser
ary of two zones is also determined in the initial state. The spraying during different dynamic processes. Chen et al. (2017b), Zhao
zone and bubbling zone inlet calculation are starting simultane- et al. (2018, 2019) built the deaerator simulation model by JTop-
ously, and the parameter calculation is sequentially advanced to meret for total system modeling and analysis. There is the local
the surrounding nodes with flow calculation. For the deaerator deaerator simulation model built in JTopmeret similar to these
simulation model with the 15  15 nodes divided, and the time researches for overall system modeling. Therefore, it could use as
step is set as 0.01 s, which could accomplish that the working fluid the validation criteria for the simulation model built in this work
would pass through one node completely within one time step. during dynamic process. Fig. 4 and Table 3 could describe the sim-
Such node division and time step could meet CFL (Courant- ulation model built in JTopmeret.
Friedrichs-Lewy) condition (Courant et al., 1967) as following to
ensure the calculation stability:
3.2.1. Condensing and feeding water increasing
G
Dt At the 5th second, the condensing and feeding water (Gsp,in)
61 ð41Þ
DV suddenly increases from 1548.7 kg/s to 1800 kg/s while other
boundary conditions and parameters remain unchanged. The
At the same time, such node division and time step can also
parameters trend and distribution change during this working con-
meet the demand for real-time computation speed. The simulation
dition could observe and obtain. This could illustrate the deaerator
model is programmed in C++ language. The simulating and model-
characteristics under such conditions. When there is the case of
ing diagram are represented in Fig. 3.
condensing water and feeding water adjustment, it is a normal
and typical working condition for deaerator in secondary loop
3. Simulation and analysis system.
In Fig. 5(a) there is the deaerator pressure (Pdea) changing over
3.1. Steady calculation and comparison time. The condensing and feeding water increase suddenly, which
means more cold fluid enters into the deaerator spraying zone.
There is the comparison of main thermal parameters and design More steam would condense which would lead to the reduction
values for steady state calculation. The steady state is that the sim- of pressure. There is the same pressure trend comparing with JTop-
ulation model is running continuously and without any boundary meret. The pressure relative error during this dynamic process is
conditions or parameters changing. The comparison and calcula- represented in Fig. 5(b). The spraying zone and bubbling zone pres-
tion differences between the calculated results and the design data sure are represented in Fig. 5(c). At 18.41 s, the liquid level would
are represented in Table 2. From Table 2 it could see that the main reach the 7th row upper limit. This would cause a row of spraying
simulating parameters with design data are all in the acceptable zone transforming into bubbling zone completely. The top 6 rows
range little than 1%. The comparison could prove the credibility of spraying zone and bottom 8 rows of bubbling zone are the nodes
for steady state simulation. The pressure (Pdea), temperature (Tdea) untransformed. In spraying zone, there is a temporary pressure
increase with the spraying zone reducing from 7 rows to 6 rows.
In bubbling zone, there is also a short pressure reduction with bub-
Table 1
Structure and boundary parameters of deaerator. bling zone expanding from 8 rows to 9 rows. The pressure of two
zones are both showing the downward trend. The pressure change
Parameters Value Unit
trend of spraying zone is more obvious than that of bubbling zone.
Structure parameters Total height 4.3 m In Fig. 6(a) there is the deaerator liquid level (Lv) changing over
Total width 4.3 m
time. As the condensing and feeding into the deaerator increases,
Total length 50 m
Bubbling tube hole diameter 0.008 m the flow rate of the deaerator inlet and outlet would not reach bal-
Nozzle diameter 0.001 m ance during a short period of time. This would lead to an increase
Boundary parameters Feed water inlet pressure 0.802 MPa in liquid level. The rise of liquid level would cause a row of spray-
Feed water inlet temperature 143.94 ing zone transforming into bubbling zone. There is the same liquid
Feed water inlet flow rate 1211.61 kg/s
Drain water inlet pressure 0.915 MPa
level trend comparing with JTopmeret. The relative error of liquid
Drain water inlet temperature 176.08 level during this dynamic process is represented in Fig. 6(b).
Drain water inlet flow rate 337.09 kg/s From the pressure and liquid level calculation and comparison,
Heating steam temperature 169.54 the model built in this work could validate for the condensing and
Heating steam flow rate 64.7 kg/s
feeding water increase working condition. The relative error during
8 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

Start of the
simulation
Structure
Node division
parameter
Simulation Parameter
input initialization
Dynamic Boundary
process parameter

Subcooling status Saturation status

Heat transfer Heat transfer

calculation Inlet Inner calculation
Inner Inlet
Flow rate
Flow rate Spraying zone Bubbling zone calculation
calculation
Thermal Thermal Time step
parameter parameter
calculation Level variation Edge calculation forward( t )
Edge

Level
calculation

Parameter changing with

time(Pressure, enthalpy, level, etc.)

Parameter
output

Stop condition

End of the
simulation

Fig. 3. Calculation diagram.

Table 2 such dynamic process is represented in Table 4 which could prove

Calculation under steady state and comparison between simulation and design value.
the calculation accuracy.
Parameters Simulation Design Relative The deaerator spraying zone enthalpy (hsp) trend is represented
results value error in Fig. 7. The increase of condensing and feeding water would
Deaerator pressure MPa 0.75194 0.7515 0.058% result in the increase of total heat transfer amount. For the unit
Deaerator liquid level m 2.5 2.52 0.8% of droplets, the enthalpy is reduced due to flow rate increasing
Deaerator temperature 167.863 167.84 0.013%
more relative to heat transfer process. At the same time, there is
Deaerator outlet flow rate Kg/s 1613.4 1613.4 0%
Actual time s 600 s also the same situation with the enthalpy change at 18.41 s and
Simulating time consuming s 547.962 s the enthalpy would slightly increase during zones transform
process.
In Fig. 8 there are the spraying zone pressure (Psp) distribution
of three moments chosen during this working condition. In Fig. 8
(a) the pressure distribution is the largest at the inlet. In other posi-
tions it is reduced in turn with flow. At the zones transform
moment, the spraying zone becomes smaller. In Fig. 8(b) and (c)
the spraying zone pressure distribution generally shows the down-
ward trend. It can be seen in Fig. 8(d) that the maximum pressure
change in the spraying zone occurs at the inlet and in the middle of
deaerator spraying zone. The pressure change at edge is small. The
pressure distribution change could correspond to Fig. 5(c).
In Fig. 9 there are the spraying zone enthalpy (hsp) distribution
of three moments chosen during this working condition. In Fig. 9
(a) the enthalpy distribution is the smallest at the inlet of spraying
Fig. 4. Deaerator simulation model built in JTopmeret. zone. The enthalpy in peripheral node is larger with flow. The
enthalpy distribution is in contrast to the pressure distribution in
Table 3
Fig. (8). During the condensing and feeding water increase, Fig. 9
The instructions of simulation model built in JTopmeret. (b) and (c) represent the enthalpy distribution downward trend.
It can be seen in Fig. 9(d) that the maximum enthalpy change in
Type Number Calculation
the spraying zone occurs at and near the inlet. The enthalpy change
Node #1 Deaerator at the bottom and border is small. The overall enthalpy distribution
heat transfer calculation
change could correspond to Fig. 7.
liquid level calculation
Flow boundary #1 Heating steam inlet In the flow vector diagram, the flow rate (Gsp) is expressed
#2 Condensing and feeding water inlet by different color. The flow direction is the synthesis in the
Pressure boundary #1 Deaerator outlet horizontal(x) direction and the vertical(y) direction. It could indi-
Link #1 Condensing and feeding water flow
cate fluid movement at different positions inside the deaerator
#2 Heating steam flow
#3 Deaerator outlet flow
spraying zone. In Fig. 10 there is spraying zone flow vector
schematics. The main direction of droplet flow is from inlet to
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 9

(a) Pressure comparing with JTopmeret (b) Pressure relative error

(c) Pressure in each zone

Fig. 5. Deaerator pressure variation during condensing and feeding water increase.

(a) Liquid level comparing with JTopmeret (b) Liquid level relative error
Fig. 6. Deaerator liquid level variation during condensing and feeding water increase.

the oblique below as represented in Fig. 10(a). The flow rate distri-
bution is overall uptrend as condensing and feeding water increas-
Table 4 ing. At the inlet and middle of deaerator spraying zone, the flow
Maximum error and dynamic process duration during condensing and feeding water rate increases more basically consistent with pressure distribution
increase.
change. In Fig. 10(c) and (d) the flow rate changes more in vertical
Parameter Maximum error in dynamic process direction. More flow rate is concentrated in the middle as repre-
Pressure 0.172% sented in Fig. 10(e) and (f) during this working condition.
Liquid level 1.309% In Fig. 11 there are the droplets diameters (dsp) distribution dur-
Dynamic process duration in this work 56.75 s ing this working condition. The droplet diameter is the smallest at
Dynamic process duration in 52.5 s
inlet. It would increase with the flow and condensation process as
JTopmeret
represented in Fig. 11(a). The droplets diameters in each position
10 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

would decrease as represented in Fig. 11(b) after transient. Due to

the decrease of heat transfer amount for per unit droplet, the con-
densing water amount is reducing which would result in the
decrease of droplet diameter.
Fig. 12 could represent the bubbling zone pressure (Pbb) distri-
bution during this work condition. Pressure changes in this zone
are not obvious compared with that in spraying zone. It changes
slightly larger at the boundary border with the spraying zone. It
is basically the same everywhere. The amount of pressure reduc-
tion from top to bottom decreases gradually for overall downward
trend in this zone. The pressure distribution change of bubbling
zone could correspond to Fig. 5(c).
Through the above simulation and calculation during condens-
ing and feeding water increasing, the deaerator operating charac-
teristics with different parameters distribution could be acquired.
The comparison with JTopmeret can confirm and verify the
Fig. 7. Deaerator spraying zone enthalpy variation during condensing and feeding established simulation model. Under such condition, the parame-
water increase. ters variation in spraying zone is more pronounced than that in
bubbling zone. Parameter distribution changes follow similar law

(a) Before increase (b) Zone transform moment

(c) After increase (d) Pressure reduction

Fig. 8. Spraying zone pressure distribution during condensing and feeding water increase.

(a) Before increase (b) Zone transform moment

(c) After increase (d) Enthalpy reduction

Fig. 9. Spraying zone enthalpy distribution during condensing and feeding water increase.
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 11

(a) Overall before increase (b) Overall before increase

(c) near inlet before increase (d) near inlet after increase

(e) Middle before increase (e) Middle after increase

Fig. 10. Spraying zone flow vector during condensing and feeding water increase.

(a) Before increase (b) After increase

Fig. 11. Spraying zone droplets diameters distribution during condensing and feeding water increase.

(a) Before increase (b) Zone transform moment

(c) After increase (d) Pressure reduction

Fig. 12. Bubbling zone pressure distribution during condensing and feeding water increase.
12 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

(a) Pressure comparing with JTopmeret (b) Pressure relative error

(c) Pressure in each zone

Fig. 13. Deaerator pressure variation during heating steam increase.

with flow. In summary, the increase of condensing and feeding 3.2.2. Heating steam increasing
water means an increase in the cold source, and the thermal At the 5th second, the heating steam (Gbb,in) would suddenly
parameters would reduce. At the same time, a large amount of increase from 64.7 kg/s to 80 kg/s, while the other boundary
water entering into the deaerator would cause the rising of liquid conditions and parameters remained unchanged. Under such cir-
level. cumstances, the parameters trends and distribution could simu-

(a) Liquid level comparing with JTopmeret (b) Liquid level relative error
Fig. 14. Deaerator liquid level variation during heating steam increase.
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 13

Table 5 trend of bubbling zone is more obvious than that of spraying zone
Maximum error and dynamic process duration during heating steam increase. during this working condition.
Parameter Maximum error in dynamic process In Fig. 14(a) there is deaerator liquid level (Lv) changing over
Pressure 0.896% time. As the heating steam into the deaerator increases, the flow
Liquid level 0.224% rate between deaerator inlet and outlet would not reach balance
Dynamic process duration in this work 14.25 s during a short period of time. The condensing rate in deaerator
Dynamic process duration in 16.5 s would gradually increase. Then the deaerator would reach mass-
JTopmeret
heat balance again which would lead to an increase in liquid level.
The change of heating steam has little effect on the liquid level.
During this working condition the liquid level does not change
much. So, there is not the movement of boundary and transform
of zones. There is the same liquid level trend comparing with JTop-
meret. The relative error of liquid level during this dynamic process
is represented in Fig. 14(b).
From pressure and liquid level calculation, the model built in
this work could validate for the heating steam increase working
condition comparing with JTopmeret. These two parameters trend
change are basically the same. The relative error is also within the
acceptable range as represented in Table 5.
The deaerator spraying zone enthalpy (hsp) trend is represented
in Fig. 15. When the heating steam increase suddenly, bubbling
zone could not condense them at the beginning. More steam would
enter into spraying zone with more heat release, which would
result in an increase of enthalpy in spraying zone.
In Fig. 16 there are the spraying zone pressure (Psp) distribution
during this working condition. The pressure distribution could pre-
sent overall upward trend which is related to Fig. 13(c). It can be
Fig. 15. Deaerator spraying zone enthalpy variation during heating steam increase.
seen that the trend of change is reduced from bottom to top in
Fig. 16(c). Pressure changes in most nodes are consistent. The pres-
sure at edge and border nodes would change more than at other
late. This typical condition may occur in the case of steam regula- positions.
tion in secondary loop. In Fig. 17 there are spraying zone enthalpy (hsp) distribution
Fig. 13(a) could show the deaerator pressure (Pdea) changing during this working condition. According to Fig. 15, the enthalpy
with time. As the heating steam increase suddenly, more hot fluid distribution is also increasing on the whole. The enthalpy trend
enters into the deaerator bubbling zone and more heat release in is also reduced from bottom to top as represented in Fig. 17(c).
bubbling zone. The deaerator cooling capacity is not enough to The enthalpy change in middle is slightly larger than in surround-
condense so much more steam. This would lead to raising of deaer- ing. The enthalpy change in border is also slightly larger than the
ator pressure. There is the same pressure tendency comparing with near. The pressure and enthalpy of spraying zone are following
JTopmeret. The relative error of pressure during this dynamic pro- basic consistent trends and similar law.
cess is represented in Fig. 13(b). The spraying zone and bubbling Fig. 18 could represent bubbling zone pressure (Pbb) distribution
zone pressure are represented in Fig. 13(c). The pressure of two during this work condition. The pressure distribution is the highest
zones are both showing the upward trend. The pressure change at the inlet position. In other positions, it is sequentially decreased

(a) Before increase (b) After increase

(c) Pressure increment

Fig. 16. Spraying zone pressure distribution during heating steam increase.
14 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

(a) Before increase (b) After increase

(c) Enthalpy increment

Fig. 17. Spraying zone enthalpy distribution during heating steam increase.

(a) Before increase (b) After increase

(c) Pressure increment

Fig. 18. Bubbling zone pressure distribution during heating steam increase.

with flow. Pressure changes more at and near the inlet. The inter- with flow and condensing process as represented in Fig. 20(a).
nal trend of change is basically the same with small difference as With the increase of heating steam, the total amount of heat
represented in Fig. 18(c). The pressure change in border position transfer in deaerator increases. But for the unit of bubble, the
is larger than in most of the middle. The pressure change of bub- amount of heat transfer would decrease because of insufficient
bling zone is related to Fig. 13(c). cooling capacity with condensing rate decrease. Less bubbles
In Fig. 19 there are bubbling zone flow (Gbb) vector schematics. would become water. This would result in the increase of bubble
The main direction of steam bubbles flow is from the inlet to the diameter. The overall bubble diameter distribution demonstrates
upper side of the slope as represented in Fig. 19(a). During this an increasing trend as represented in Fig. 20(b).
working condition, the flow rate distribution is overall uptrend. From comparing with JTopmeret, the simulation model built in
Near the inlet and the bottom edge in Fig. 19(c) and (d), the flow this work could validate during heating steam increasing. In sum-
rate increases more which is basically in line with the trend of mary, the increase of heating steam flow is equivalent to an
pressure change. In other positions as represented in Fig. 19(e) increase in hot source, and the deaerator is not sufficient to con-
and (f), the flow rate has also increased. From Fig. 19 more steam dense the steam. This would result in the increase of thermal
would flow from the middle to the border during this dynamic parameters in deaerator. The parameters variation in bubbling
process. zone is more pronounced than that in spraying zone. The change
In Fig. 20 there are steam bubble diameters (dbb) distribution of parameters distribution also corresponds to the flow in each
during this working condition. At the inlet, the bubble diameter zone. The liquid level is not changing obvious during this working
is the largest. The diameter at other positions gradually decreases condition.
 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194 15

(a) Overall before increase (b) Overall after increase

(c) Near inlet before increase (d) Near inlet after increase

(e) Top edge before increase (f) Top edge after increase
Fig. 19. Bubbling zone flow vector during heating steam increase.

(a) Before increase (b) After increase

Fig. 20. Bubbling zone bubble diameter distribution during heating steam increase.

4. Conclusions distribution and detailed calculation with time changing could

simulate. The parameters such as thermal parameters and flow
In this work, the quasi 2-D refined simulation model is devel- vector at different positions are calculated in two zones with
oped for the deaerator in nuclear power plant which is direct con- more nodes division.
tact heat exchanger. Reasonable assumptions based on actual 2. The model built based on proposed assumptions and
physical processes are proposed. The parameters expressions with mathematical modeling could confirm the accuracy. There is
mathematical modeling are derived for each zone with nodes. the verification with the design data under steady state and
With this modeling method, the calculation speed and precision comparison with JTopmeret under different dynamic processes.
could be guaranteed. This work has focused on the simulation The validation could also prove that the assumptions and
of time scale and spatial dimension expansion both at the same modeling are reasonable.
time. 3. Calculation with time changing and different parameters distri-
The main contribution of this work could be summarized as bution could reflect the deaerator operating characteristics dur-
follows: ing different dynamic processes. The positions of the internal
maximum parameters and the positions of parameters maxi-
1. Quasi 2-D dynamic simulating and modeling for deaerator is mum change in each zone during the dynamic processes could
feasible. With such processing method, the parameters reveal and obtain by simulation.
16 M. Gong et al. / Annals of Nuclear Energy 138 (2020) 107194

4. By regulating the flow rate of the condensing and feeding water, Katebi, R., 2007. Modelling, simulation and control of large power plants. IFAC
Proceedings Volumes 40 (8), 3–14.
the parameter change of spraying zone is more obvious than
Kim, S., Park, G., 2011. Interfacial heat transfer of condensing bubble in subcooled
that of bubbling zone, and the liquid level change is more obvi- boiling flow at low pressure. Int. J. Heat Mass Transfer 54 (13–14), 2962–2974.
ous. By regulating the flow rate of heating steam, the overall Kowalczyk, T., Badur, J., Bryk, M., 2019. Energy and exergy analysis of hydrogen
parameters changes of the deaerator are more obvious, but production combined with electric energy generation in a nuclear cogeneration
cycle. Energy Convers. Manage. 198, 111805.
the liquid level does not change significantly. Kundu, P., Cohen, I., Ayyaswamy, P., Hu, H., 2008. Fluid mechanics. Elsevier/
Academic Press, Amsterdam, pp. 283–287.
In the future work, more operational working conditions with Li, X., Wang, Z., Yang, M., Yuan, G., 2019a. Modeling and simulation of a novel
combined heat and power system with absorption heat pump based on solar
experiments and actual equipment operation would use for verify- thermal power tower plant. Energy 186, 115842.
ing the simulation model. For the improvement of developed sim- Li, X., Xu, E., Ma, L., Song, S., Xu, L., 2019b. Modeling and dynamic simulation of a
ulation model, more detailed processes such as chemical processes, steam generation system for a parabolic trough solar power plant. Renewable
Energy 132, 998–1017.
mass transfer processes and bubbles-droplets distribution in Lu, C., Bell, R., Rees, N., 1998. Scheduling control of a deaerator plant. Control Eng.
deaerator would be considered. The deaerator model developed Pract. 6 (12), 1541–1548.
in this work could serve as a basic simulation model in nuclear Nomura, T., Tsubota, M., Sagara, A., Okinaka, N., Akiyama, T., 2013. Performance
analysis of heat storage of direct-contact heat exchanger with phase-change
power plant secondary loop system simulating, modeling and anal- material. Appl. Therm. Eng. 58 (1–2), 108–113.
ysis. It is applicable to provide reference for similar direct contact Oko, E., Wang, M., 2014. Dynamic modelling, validation and analysis of coal-fired
heat exchanger in other thermal system. subcritical power plant. Fuel 135, 292–300.
Patankar, S.V., 1980. Numerical heat transfer and fluid flow. Hemisphere Publishing,
Washington, D.C, pp. 11–23.
Declaration of Competing Interest Rodrı́guez-Toral, M., Morton, W., Mitchell, D., 2000. Using new packages for
modelling, equation oriented simulation and optimization of a cogeneration
The authors declare that they have no known competing finan- plant. Comput. Chem. Eng. 24 (12), 2667–2685.
Roldan-Villasana, E. J., Vazquez, A. K., 2010. Model of the Feed Water System
cial interests or personal relationships that could have appeared Including a Generic Model of the Deaerator for a Full Scope Combined Cycle
to influence the work reported in this paper. Power Plant Simulator. In: Fourth Uksim European Symposium on Computer
Modeling & Simulation. IEEE Computer Society.
Shukla, S., Samad, A., Ghosh, S., 2017. CFD simulation of steam condensation in a
Appendix A. Supplementary data subcooled water pool. Thermal Science and Engineering Progress 2, 80–86.
Srinivas, T., 2009. Study of a deaerator location in triple-pressure reheat combined
power cycle. Energy 34 (9), 1364–1371.
Supplementary data to this article can be found online at
Wan, J., Song, H., Yan, S., Sun, J., Zhao, F., 2015. Development of a simulation
https://doi.org/10.1016/j.anucene.2019.107194. platform for dynamic simulation and control studies of AP1000 nuclear steam
supply system. Ann. Nucl. Energy 85, 704–716.
Wang, B., Liu, C., Wu, X., Liu, L., 2011. RBF Neural Network Based Adaptive Tracking
References
Control for a Class of Nonlinear Plant Using Stochastic U-Model. Key Eng. Mater.
474–476, 1209–1214.
Alobaid, F., Mertens, N., Starkloff, R., Lanz, T., Heinze, C., Epple, B., 2017. Progress in Wang, D., Zhou, Y., Zhou, H., 2016a. A mathematical model suitable for simulation of
dynamic simulation of thermal power plants. Prog. Energy Combust. Sci. 59, 79– fast cut back of coal-fired boiler-turbine plant. Appl. Therm. Eng. 108, 546–554.
162. Wang, J., Fu, L., Zhao, X., Zhao, J., 2017. Mathematical simulation of the flue-gas
Baqir, A., Mahood, H., Sayer, A., 2018. Temperature distribution measurements and recovery system for coal-fired boilers based on the direct-contact heat
modelling of a liquid-liquid-vapor spray column direct contact heat exchanger. exchanger. Procedia Eng. 205, 321–328.
Appl. Therm. Eng. 139, 542–551. Wang, P., Fu, Y., Wei, X., Zhao, F., 2018. Simulation study of frequency control
Batchelor, G.K., 2010. An introduction to fluid dynamics. Cambridge University characteristics of a generation III + nuclear power plant. Ann. Nucl. Energy 115,
Press, pp. 210–216. 502–522.
Cascella, F., Teyssedou, A., 2015. Modeling a Direct Contact Heat Exchanger used in Wang, W., Liu, J., Zeng, D., Niu, Y., Cui, C., 2016b. Modeling for condensate throttling
a supercritical water loop. Appl. Therm. Eng. 79, 132–139. and its application on the flexible load control of power plants. Appl. Therm.
Chen, C., Zhou, Z., Bollas, G., 2017a. Dynamic modeling, simulation and optimization Eng. 95, 303–310.
of a subcritical steam power plant. Part I: Plant model and regulatory control. Wagner, W., Cooper, J.R., Dittmann, A., Kijima, J., Kretzschmar, H.J., Kruse, A., et al.,
Energy Convers. Manage. 145, 324–334. 2000. IAPWS Industrial Formulation 1997 for the Thermodynamic Properties of
Chen, W., Li, B., Zhang, S., Liu, M., Liu, J., 2017b. Simulation investigation on the Water and Steam. J. Eng. Gas Turbines Power 122 (1), 150–184.
design and operation strategy of a 660 MW coal-fired power plant coupled with Xue, R. J., Sun, J. L., 2014. Modeling and Simulation of Deaerator in Nuclear Power
a steam ejector to ensure NOx reduction ability. Appl. Therm. Eng. 124, 1103– Plant. In: International Conference on Nuclear Engineering. American Society of
1111. Mechanical Engineers.
Courant, R., Friedrichs, K., Lewy, H., 1967. On the partial difference equations of Yang, S., Seo, J., Hassan, Y., 2019. Thermal hydraulic characteristics of unstable
mathematical physics. IBM J. Res. Dev. 11 (2), 215–234. bubbling of direct contact condensation of steam in subcooled water. Int. J. Heat
Dong, Z., Pan, Y., 2018. A lumped-parameter dynamical model of a nuclear heating Mass Transfer 138, 580–596.
reactor cogeneration plant. Energy 145, 638–656. Zeng, Z., Sadeghpour, A., Ju, Y., 2018. Thermohydraulic characteristics of a multi-
Dong, Z., Pan, Y., Zhang, Z., Dong, Y., Huang, X., 2018. Dynamical modeling and string direct-contact heat exchanger. Int. J. Heat Mass Transfer 126, 536–544.
simulation of the six-modular high temperature gas-cooled reactor plant HTR- Zhang, Y., Wang, J., Yang, S., Gao, W., 2018. An all-condition simulation model of the
PM600. Energy 155, 971–991. steam turbine system for a 600 MW generation unit. J. Energy Inst. 91 (2), 279–
Fu, H., Ma, L., Wang, H., 2018. Experimental and numerical studies of residence time 288.
in SK direct contact heat exchanger for heat pump. Chem. Eng. Res. Des. 135, Zhao, Y., Liu, M., Wang, C., Wang, Z., Chong, D., Yan, J., 2019. Exergy analysis of the
94–102. regulating measures of operational flexibility in supercritical coal-fired power
Gong, M., Peng, M., Zhu, H., 2019. Research of parameter distributing simulation and plants during transient processes. Appl. Energy 253, 113487.
modeling for the condenser in nuclear power plant. Ann. Nucl. Energy 133, 313– Zhao, Y., Wang, C., Liu, M., Chong, D., Yan, J., 2018. Improving operational flexibility
326. by regulating extraction steam of high-pressure heaters on a 660 MW
GSE systems Inc, 2006a. JTopmeret USER Guide. GSE systems Inc, USA. supercritical coal-fired power plant: A dynamic simulation. Appl. Energy 212,
GSE systems Inc, 2006b. JTopmeret Model Maintenance Guide. GSE systems Inc, 1295–1309.
USA. Zhou, F., Novog, D., 2018. Mechanistic modelling of station blackout accidents for a
Guangdong Nuclear Power Training Center, 2005. 900 MW PWR nuclear power generic 900 MW CANDU plant using the modified RELAP/SCDAPSIM/MOD3.6
plant system and equipment, second volume. Atomic Press, Beijing, China., code. Nucl. Eng. Des. 335, 71–93.
395–405 (In Chinese)
Holman, J.P., 2010. Heat transfer. 10th ed. McGraw-Hill. pp. 489- 491.