Computational Fluid Dynamics Simulations to Evaluate the Performance Improvement for Air-cooler Equipped with a Water Spray System

This paper introduces a series of Thermo-fluid dynamics simulations, realized by the software ANSYS [1]. Computational Fluid Dynamics can be a valuable tool for estimating the potential and the performance of evaporative cooling by water spray system.

In water spray cooling, water is sprayed into the inlet air in order to reduce the inlet air temperature by evaporative cooling. Cooler inlet air means better heat exchange for the air-cooler. In this way increases the overall system efficiency and helps the plant to recover some of the performance reduction caused by hot ambient temperatures. Evaporative cooling technologies are a smart way of air cooling in hot weather and in temperate climates [2].

 An efficacious water spray design has to avoid non-uniform cooling distribution and incomplete evaporation of droplets. These issues can be avoided, and an optimum design can be achieved, only if the spray cooling mechanisms under these conditions is well understood. Spray cooling systems performance have been a subject of research for many years and many experimental studies and numerical models have been carried out [3-6].

Spray cooling performance is influenced by air velocity, nozzle arrangement, cone angle, injection rate, droplet velocity, injection direction and droplet size [7-8].

Montazeri, Blocken and Hensen presented a systematic evaluation of the Lagrangiane-Eulerian approach for evaporative cooling provided by the use of a water spray system with a hollow-cone nozzle configuration. The evaluation was based on grid-sensitivity analysis and validated using wind-tunnel measurements [9]. Their study showed that CFD simulation of evaporation by using the Lagrangiane-Eulerian (3D steady RANS) approach, in spite of its limitations, can accurately predict the evaporation process with an acceptable accuracy. The local deviations from the wind-tunnel measurements were within 10 % for dry bulb temperature, 5 % for wet bulb temperature and 7 % for the specific enthalpy. The average deviations for all three variables were less than 3 % in absolute values.

Another experimental investigation on water spray for cooling tower application was conducted by Abdullah et al. [10]. The authors made use of open-circuit wind tunnel to simulate NDDCT built at the University of Queensland (QU). The phase dropper particle analyser was employed to characterize water spray. The study showed that low air velocity or small droplet size distribution are beneficial for cooling performance. The reason is that both droplet size and air velocity determine spray coverage that directly influences the spray cooling efficiency. Then the authors made a parametric analysis on the effect of different spray characteristics parameters, like droplet size distribution, injection velocities, spray cone angles and hollow-cone droplet size distribution patterns [11]. Considering an optimised nozzle and comparing this with a datum nozzle, it results around 15 % improvement in the spray cooling efficiency.

Sun et al. conducted a study to evaluate the influence of injection direction on the spray cooler performance in a natural draft dry cooler tower [12]. The optimal injection for a hollow cone nozzle had been identified based on CFD study. This study shows that injection direction has a great influence on the evaporation process of the injected water droplets. For a single nozzle with the water mass flow rate of 5 g/s, the largest temperature drop is 1.27 °C, corresponding to the radiator temperature of 38.73 °C. Moreover, the increment of injection angle can enlarge the water-cooled area of radiators and the optimum injection angle varies with the height of nozzle location.

M. Sadafi and K. Hooman conducted a study with the aim to improve the knowledge base associated with the use of saline water in spray cooling applications [13]. Comparing saline water with pure water, it is obtained that saline water can improve cooling efficiency by 8 % close to the nozzles. Moreover, full evaporation is achieved earlier compared to the pure water case. This accelerated evaporation process gives engineers the possibility to reduce the evaporation distance of up to 30 % from the nozzle exit.

In this section a comparison between different air inlet velocities, nozzle arrangements, droplet size, and spray cone angle is conducted, in order to evaluate the better configuration, while other parameters are fixed in order to avoid to affect the results.

 In a preliminary analysis is important to evaluate Reynolds number, that results equal to 1.25E⁶ and indicates a turbulent flow. In our case, the fluid involved is the air that we can consider incompressible because the Mach number is minor than 0.3, hence, pressure-based solver was chosen and the k-ε model was applied, in his realizable variant.

In order to simulate the system, it is essential to know a series of input data, for instance the inlet air temperature and mass fraction of water. Data of these simulations are taken from a real case and are shown in Table 1.

Final air humidity is the desired humidity that we want to achieve at the end of evaporative cooling process. A better situation occurs if saturated air is obtained, but in real applications this is impossible to be achieved. The causes are mixing time insufficient, exchange area not uniform and thus the air humidity doesn’t reach 100 %. If a further quantity of water is injected, it doesn’t evaporate but remains in the form of droplets. This could be a problem for fan blades corrosion and for the increased consumption of water. Therefore, manufacturers suggest 85 % as limit, on the basis of experimental results. This work refers to the air cooler shown in Figure 1, where the heat exchanger, fan, plenum and spray nozzles are highlighted. The simulations focus on the fan, spray nozzle and plenum section of the system.

1.png

Figure 1. Reference geometry of the air cooler

For computational reason, geometry will be slightly different with respect the real one. Computational domain is a truncated cone and its dimensions are shown in Figure 2.

Authors decided to extend the computational domain for the simulations to 15 m, in order to avoid the influence of the boundary condition on the solution, in the zone of interest.

Mesh, for the computational domain, was created with the software GAMBIT of Fluent (Figure 3). The mesh created for the computational domain is composed by 49.630 cells. It is clear that calculation is quite computationally demanding. As you can see, no nozzles and fans appear in the computational domain, but these components were taken into account later.

2.png

Figure 2. Computational domain

3.png

Figure 3. Mesh for the computational domain

For the simulation, other data have to be set, and such data are the results of the psychrometric calculation carried out (Table 1) and reference geometry.

In order to simulate the motion produced by the fan, an angular component for the inlet velocity was set, evaluated from the fan technical sheet (Table 4).

Table 4. Fan technical sheet

4.png

Figure 4. Full Cone Spray Nozzle (Spraytech)

For data concerning spray nozzle system, the first step is to choose number and type of nozzles. The calculation made refers to the model chosen in Figure 4, that is a full cone nozzle with an angle of 60°. In this work, different types of nozzle arrangements have been taken into account.

3.1 Boundary conditions

The boundary conditions used are those of the wall, speed at inlet and outlet and pressure. The axial and angular velocity, temperature (312 K) and water mass fraction of the incoming air are set at the inlet.

At the outlet were set only air temperature (302.22 K) and air specific humidity (0.0217 kg_w/kg_a). The goal is to investigate in which section the desired outlet conditions are achieved.

3.2 Discrete phase model 

In addition to solving transport equations for the continuous phase, FLUENT allows to simulate a discrete second phase in a Lagrangian frame of reference. This second phase consists of spherical particles (which may be taken to represent droplets or bubbles) dispersed in the continuous phase. FLUENT computes the trajectories of these discrete phase entities, as well as heat and mass transfer to/from them. The coupling between the phases and its impact on both the discrete phase trajectories and the continuous phase flow can be included.

Authors decided to use Discrete Phase Model to simulate the water injection on the continuous phase, that is the air. In particular, droplet evaporation has to be evaluated. It can include a discrete phase in the FLUENT model by defining the initial position, velocity, size, and temperature of individual particles. These initial conditions, along with the inputs defining the physical properties of the discrete phase, are used to initiate trajectory and heat/mass transfer calculations.

The trajectory and heat/mass transfer calculations are based on the force balance on the particle and on the convective/radiative heat and mass transfer from the particle, using the local continuous phase conditions as the particle moves through the flow.

3.3 List of cases 

In this paper several cases have been studied, in order to compare them and find the best conditions. In particular different types of nozzle arrangement are simulated, as shown in Figure 5.

5.png

Figure 5. Different types of nozzle arrangement

In addition, simulations were conducted for different air velocities, spray cone angles and droplet size, in order to assess their influence on the evaporative process (Table 5).

Table 5. Values of the parameters used for the simulations

This work introduces a study conducted by Thermo-Fluid Dynamic simulation, in order to evaluate the performance of evaporative cooling, provided by a water spray system. Droplet evaporation and the resulting air cooling were studied, as a function of some parameters, like air velocity, nozzle arrangement, spray cone angle and droplet size. A cooling efficiency factor was introduced to evaluate the entity of cooling process.

The main achievements of this work are as follows:

(1) Considering the nozzle arrangement reported in Figure 5, the best configuration is the number 4, because of its impact on the spray coverage area. Cooling efficiency value range is from 71 % to 65 % for the worst arrangement, the number 1.  For the other configurations, values are very closer, 69 % for arrangements 2 and 3 and 68 % for arrangement 5. The final air temperature varies from 304.2 K for arrangement 1, to 303.4 K for arrangement 4. In the better case the temperature drop is equal to 8.6 K.

(2) Changing air velocity from 2.0 m/s to 10.0 m/s, cooling efficiency decreases from 81 % to 45 %. It means that air velocity has a significant impact on cooling performance, because of its impact on residence time. In fact, final temperature increases from 302.3 K to 306.5 K that is a wide range. Air humidity reaches 84 % for air velocity of 2.0 m/s, the greater value obtained.

(3) The angle of the spray cone has varied from 30° to 60° and the results indicate that this parameter has a negligible impact on cooling efficiency. In fact, the cooling efficiency is almost constant.

(4) For the evaluation of impact of droplet size, last one was changed in a range from 10 μm to 100 μm. Cooling efficiency increases if droplet size decreases, for the influence on residence time. Cooling efficiency varies from 69% to 46% and final temperature from 303.7 K to 306.4 K. This impact is significant on cooling performance.

(5) The results of this study confirm that such technology enhances the performance of a plant, built in arid areas, operating on existing devices, without high cost. In fact, we arrive to a temperature drop of 9.7 °C, and such result improve the heat transfer coefficient of the air-cooler, when ambient air temperature is very high.

(6) The reported study shows that humidity in the better case is equal to 84%, according to manufactures suggestions, so we can say that the results are quite accurate. To validate the results obtained, the authors intend to proceed with experimental measures.