Pre-cooling the inlet air of a dry cooling tower by means of a spray can improve the tower performance during periods of high temperature. To study the spray effect on the thermal performance of natural draft dry cooling towers (NDDCTs), in this study 3-D numerical simulations of such a process have been conducted using Fluent 16.2 (a two-way coupled Eulerian-Lagrangian approach). The considered NDDCT is 120 m high and only half system is simulated due to its structural symmetry. Three different spray strategies have been investigated at a typical crosswind speed of 4 m/s, which is the most frequent wind speed. The results have shown that: (1) The three implemented spray strategies can improve the thermal performance of the studied NDDCT with a varying degree of success. In one case, the heat rejection rate can be increased by 35.2%, and the tower outlet water temperature can be decreased by 2.1°C when compared with the no spray case; (2) To improve the thermal performance of the NDDCT using a small amount of water, the design of the spray pre-cooling system must include more nozzles on the windward and fewer or even no nozzles on the leeward sides of the NDDCT.
A cooling tower is the key component of thermal power plants, its performance affects the operating efficiency of served power plants [
To improve the thermal performance of a NDDCT during periods of high temperature, many scholars have proposed various methods such as dry-wet hybrid cooling and evaporative pre-cooling of the inlet air of the tower. He [
Studies about spray pre-cooling of a NDDCT have been carried out by many researchers. Sadafi et al. [
The spray pre-cooling system could improve the thermal performance of a NDDCT during periods of high temperature, but crosswind will change airflow field of the tower. Du Preez et al. [
Literature review found that crosswind will change airflow field inside and outside the tower, and therefore, the nozzle design of the spray pre-cooling system under crosswind will be different from windless conditions. It is necessary to investigate the nozzle design of spray pre-cooling system under typical crosswind, especially under the most frequent wind speed. To this end, a 3-D model of a spray pre-cooled NDDCT was implemented using Fluent 16.2. The airflow field and temperature field of the NDDCT under typical crosswind with and without nozzle spray were compared. Further, the effects of three spray strategies on the thermal performance of the NDDCT under typical crosswind were extensively explored. This study could provide guidance for the engineering design of a spray pre-cooled NDDCT.
The innovations of the current study are: (1) A 120 m high and 1/2 NDDCT model is developed to investigate spray effect on the thermal performance of the NDDCT pre-cooled by nozzle spray under typical crosswind of 4 m/s; (2) The thermal performance of the NDDCT pre-cooled with three spray strategies under typical crosswind is extensively studied, including droplet movement, air temperature distribution, the heat rejection performance of the tower, air ventilation performance of the tower and the outlet water temperature of the tower; (3) A good spray strategy for improving the performance of the studied NDDCT is obtained.
The geometric model of the studied NDDCT is shown in
Main parameters and units | Value | Main parameters and units | Value |
---|---|---|---|
Tower height/m | 120 | Radiator arrangement | A frame |
Radiator height/m | 13.7 | Radiator type | Finned tube |
Radiator surface diameter/m | 83 | Number of tube rows | 4 |
Outlet diameter/m | 58 | Finned tube length | 15.0 |
Water flowrate/(kg⋅s−1) | 4390 | Effective windward area of the radiator/m2 | 4625.3 |
Inlet water temperature/°C | 61.5 |
Boundary | Boundary conditions | Boundary | Boundary conditions |
---|---|---|---|
Windward surface | Pressure inlet (No wind) | Bottom surface | Ground |
Velocity inlet (Wind) | Center surface | Symmetry | |
Leeward surface | Pressure inlet (No wind) | Inlet | Interior |
Pressure outlet (Wind) | Outlet | Interior | |
Top surface | Pressure outlet (No wind) | Radiator surface | Radiator |
Velocity inlet (Wind) | Shell | Wall |
In the spray pre-cooled NDDCT, the spray pre-cooling system was arranged at the air inlet zone of the NDDCT. Water was injected from the nozzles as fine droplets, the droplets then exchanged heat and mass with air. The spray cooling model in the Fluent software was used to simulate the spray pre-cooling process.
A structured grid was used to do mesh generation. The operating conditions used for the grid refinement study were: ambient air temperature of 40°C, inlet water temperature of 61.5°C, and circulating water at a flowrate of 4390 kg/s. Three grid numbers were studied and the results are listed in
Grid number | Air speed at the radiator exit (m/s) | The percentage variations of air speed at the radiator exit (%) | Air temperature at the radiator exit (K) | The percentage variations of air temperature at the radiator exit (%) | Air temperature at the tower outlet (K) | The percentage variations of air temperature at the tower outlet (%) |
---|---|---|---|---|---|---|
2768687 | 2.54 | - | 305.44 | - | 319.84 | - |
3313640 | 2.58 | 1.575 | 305.40 | −0.013 | 320.15 | 0.097 |
3793851 | 2.60 | 0.775 | 305.39 | −0.003 | 320.20 | 0.016 |
The Eulerian-Lagrangian method was used to describe the motion of the two-phase flow in the spray evaporative cooling process. In the Eulerian-Lagrangian method, air was the continuous phase and was described by the Eulerian method while water droplet was the discrete phase and was described by the Lagrangian method. Different methods can be adopted to describe the interaction between continuous and discrete phases in turbulent flow [
In the Eulerian-Lagrangian method, the air was the continuous phase which was treated as a steady, incompressible ideal gas, and of turbulent flow. Dry air was simplified to be composed of 77% nitrogen and 23% oxygen, and wet air was simplified to be composed of nitrogen, oxygen, and water vapor. Air flow field was described by the Reynolds-averaged Navier-Stokes conservation equations (RANS) combined with the standard
where
The source terms
where
In the process of spray evaporative cooling, particle size, temperature, speed, and trajectory of the spray droplets were different. Due to a huge number of droplets, “droplet group” was introduced as the research object to decrease computational time. Each droplet group represented a certain water flowrate. In addition, it was assumed that all the droplets in each droplet group had the same characteristics and followed the same evaporation laws, thus the discrete phase was solved for all the droplet groups.
A hollow cone nozzle produces finer droplets when compared with a full cone nozzle. Furthermore, the evaporation process is accelerated because the contact area between air and water droplets is enlarged. In view of its excellent performance for cooling the air, a LNN1.5 commercial hollow cone nozzle was used during simulation. The parameters of the LNN1.5 nozzle are listed in
Parameter | Velocity (m/s) | Nozzle cone angle (°) | ||
---|---|---|---|---|
Value | 22 | 39 | 55 | 85 |
The water droplets were treated as droplet parcels during the simulation because of the huge number of water droplets in the spray cooling system. The “parcel” represented a fraction of the total water mass flowrate. Each parcel included water droplets with identical properties (i.e., diameter, velocity, trajectory, and temperature). Zhang et al. [
where
where
In the Lagrangian framework, the droplet motion follows Newton’s second law, and the trajectory of a droplet can be obtained by integrating the droplet motion equation. To simplify calculation, assuming that droplet had its own velocity and was spherical, the equation of the motion of a single droplet is:
where
The radiator was used to describe the radiator of the NDDCT. The radiator was treated as a surface with resistance coefficient and heat transfer coefficient. The empirical correlations for the resistance coefficient and heat transfer coefficient of the radiator are [
where
As shown in
where
A same NDDCT model with the reference [
Spray pre-cooling the inlet air is proposed to improve the thermal performance of a NDDCT during periods of high temperature, this paper intends to conduct research under typical high ambient temperatures. Therefore, air temperature of 40°C, relative humidity of 40% and crosswind speed of 4 m/s (which was the most frequent wind speed) were selected for simulation [
Spray water flowrate required to fully saturate the airflow of the half tower was calculated as 16.0 kg/s. Therefore, nine branch lines were evenly arranged in the half tower as shown in
Working conditions | Spray water flowrate/kg/s | Heat rejection rate of the half tower/MW | Air ventilation of the half tower/kg/s | The outlet water temperature/°C |
---|---|---|---|---|
No spray | 0.0 | 54.2 | 3259.2 | 55.6 |
Spray strategy 1 | 16.0 | 70.3 | 3013.1 | 53.8 |
Spray strategy 2 is a case that all spray nozzles on the branch lines of 0°, 157.5° and 180° are closed while spray nozzles on the branch lines of 112.5° and 135° are partly closed. Specifically, the spray nozzles on the branch lines of 112.5° and 135° at the radiuses of 21.5∼41.5 m are closed while the nozzles at the radiuses of 0∼21.5 m are open. For such a case, the spray water flowrate is reduced from 16 to 10 kg/s.
In spray strategy 3, all spray nozzles on the branch lines of 0°, 157.5° and 180°, and spray nozzles on the branch lines of 112.5° and 135° at the radiuses of 21.5∼41.5 m are moved to the branch lines of 33.75°, 56.25° and 78.75°. For such a case, the spray water flowrate is still 16 kg/s.
To further evaluate the improvement in the performance of the NDDCT per unit spray water flowrate, the heat rejection increment per unit spray water flowrate is introduced as
Working conditions | Spray water flowrate/kg/s | Heat rejection rate of the half tower/MW | Air ventilation of the half tower/kg/s | The outlet water temperature/°C | Heat rejection rate per unit spray water flowrate/MJ/kgwater |
---|---|---|---|---|---|
No spray | 0.0 | 54.2 | 3259.2 | 55.6 | - |
Spay strategy 1 | 16.0 | 70.3 | 3013.1 | 53.8 | 1.01 |
Spay strategy 2 | 10.0 | 66.7 | 3041.9 | 54.2 | 1.25 |
Spay strategy 3 | 16.0 | 73.3 | 2972.3 | 53.5 | 1.19 |
In terms of the heat rejection increment per unit spray water flowrate, spray strategy 2 has highest value, followed by spray strategy 3 and lowest for spray strategy 1. On the one hand, the spray water flowrate for spray strategy 1 is the same as spray strategy 3. Besides, spray strategy 3 has higher heat rejection rate when compared with spray strategy 2. This is because some spray droplets are blown out of the tower due to crosswind at spray strategy 1. The spray droplets blown out of the tower will not participate in heat exchange and will not improve the thermal performance of the NDDCT. This will result in invalid evaporation and waste of water resources. Therefore, the heat rejection rate per unit spray water flowrate by spray strategy 3 is higher than that of spray strategy 1. On the other hand, spray strategy 2 closes some nozzles and the spray water flowrate is reduced by 37.5% when compared with spray strategies 1 and 3. Spray strategy 3 sprays more water when compared with spray strategy 2, and makes the low-temperature area on the radiator surface more concentrated. As a result, spray strategy 3 has a lower outlet water temperature and higher heat rejection rate when compared with spray strategy 2. Therefore, increasing the spray water flowrate can improve the heat rejection rate of the NDDCT, but it is not recommended to increase the spray water flowrate indefinitely.
In this paper, a 3-D model of a spray pre-cooled NDDCT with a tower height of 120 m was implemented. The thermal performance of the NDDCT pre-cooled with three spray strategies under typical crosswind was extensively studied. The following conclusions can be drawn: All the studied spray strategies improve the thermal performance of the 120 m high NDDCT. The thermal performance values of the NDDCT pre-cooled with spray strategies 1, 2 and 3 are increased by 29.7%, 23.1% and 35.2% (i.e., the heat rejection rates of the NDDCT are increased from 54.2 to 70.3, 66.7 and 73.3 MW), respectively when compared with no spray case. The values of the outlet water temperature of the NDDCT are decreased by 1.8°C, 1.4°C and 2.1°C (i.e., the values of the outlet water temperature are decreased from 55.6°C to 53.8°C, 54.2°C and 53.5°C), respectively. Spray strategy 3 has the highest heat rejection performance of the studied NDDCT when compared with spray strategies 1 and 2. If one pursues the high improvement of tower performance without consideration of the water consumption, spray strategy 3 is recommended; if one pursues the improvement of tower performance with conditions of limited water availability/high water price, then spray strategy 2 is recommended as its heat rejection increment per unit water flowrate is high. Spray strategy 3 has the characteristics of arranging more nozzles on the windward side and fewer or even no nozzles on the leeward side of the NDDCT.
Droplet resistance coefficient
Specific heat capacity, J/(kg⋅K)
Diameter, m
Diffusion coefficient, m2/s
Internal energy, J/kg
Energy of droplets, J/kg
Force, N
Gravitational acceleration, m/s2
Heat and mass transfer coefficients
Specific enthalpy of
Latent heat of water evaporation, J/kg
Diffusion flux, kg/(m2⋅s)
Resistance coefficient
Air thermal conductivity, W/(m⋅K)
Mass, kg
Heat transfer, W
Energy source term, W/m3
Droplet surface area, m2
Mass source term, kg/(m3⋅s)
Momentum source term, kg/(m2⋅s2)
Temperature, K
Velocity, m/s
Calculation unit volume, m3
Mass fraction of substance
Height, m
Average strain tensor, 1/s
Viscosity, kg/(m⋅s)
Density, kg/m3
Viscous dissipation, W/m3
The authors would like to thank the reviewers and editors for their useful suggestions for the improvement in the quality of our manuscript.
This work was supported by the Shandong Provincial Science and Technology SMEs Innovation Capacity Improvement Project (2022TSGC2018), the Shandong Natural Science Foundation (Grant No. ZR2022ME008), the Shenzhen Science and Technology Program (KCXFZ20201221173409026), the “Young Scholars Program of Shandong University” (YSPSDU, No. 2018WLJH73), the Open Project of State Key Laboratory of Clean Energy Utilization, Zhejiang University (Program Number ZJUCEU2020011) and the Shandong Natural Science Foundation (Grant No. ZR2021ME118).
The authors declare that they have no conflicts of interest to report regarding the present study.