Technical measures for improving the cold island effect of air source heat pump unit array

Abstract

In recent years, in response to the call for energy conservation and emission reduction, the air source heat pump systems have been widely applied in regional centralized heating engineering.However, due to the existence of cold island effect, the heat pump unit operation in winter has the problems of low efficiency, high operation cost and insufficient production.This paper takes the heat source station combined with gas source heat pump and electric boiler as an example, and its heat pump unit array was simulated and calculated by analyzing relevant test data and using the numerical calculation method. Some technical measures were put forward to improve the cold island effect, and then the advantages and application scope of these measures are also analyzed.

Keywords

Air source heat pump regional heat supply cold island effect numerical simulation technical measures

1. Introduction

After the rapid economic development period, China needs to pay more attention to the environmental conservation issue. In the course of the “thirteenth Five-Year Plan”, air quality has attracted high attention in the nation, thus how to coexist with the nature harmoniously becomes a key consideration [1]. In recent years, as the environmental protection measures rise, air source heat pumps with extremely high performance coefficients have been widely used in the heating systems.Air source heat pump by compressor, evaporator, condenser and throttling device four parts, using the principle of reverse card cycle, through the refrigerant circulation system, absorb heat from the outdoor air, the refrigerant boiling point, and then through the compressor compression heating into the tank, the heat released into the water circulation system, then throttling pressure cooling back to the outdoor heat exchanger, go to the next cycle.

But through actual measurement, it is discovered that in some heat supply projects taking air source heat pumps as the heat source, the air source heat pumps operate unstably in the low-temperature winter environment, undergoing a great impact from environmental factors, therefore, the installed capacity of air source heat pump heating system is larger in actual design and installation [2]. In the practical engineering, the air source heat pump units are always installed and operated in an array. Limited by site, the intensively arranged heat pump units are liable to cause the cold air to accumulate and then sediment, which will further form a local low-temperature area in middle of the unit array, resulting in a cold island effect. This gives rise to a higher engineering investment, lower operating efficiency and higher running cost. Thereby, the influence of cold island effect on the units’ operating process should not be ignored, and this reminds us of the importance of a reasonable unit arrangement.

Till today, overseas and domestic scholars have made extensive research to enhance the performance of air source heat pump in subzero refrigeration. Shen and Song [3] from North China Electric Power University stated that the poor performance of air source heat pumps under low temperature was mainly resulted from their static-state design which can lead to a severe mismatching of restricting elements under a low-temperature work condition. Tian et al. [4] from Tsinghua University proposed a double-stage compression variable-frequency air source heat pump system for cold regions. To meet different demands, the system can utilize the frequency conversion means and the dual control mode of efficiency priority and heating capacity priority to improve the heating performance coefficients and heating capacity of the heat pump system. Ma and Shao [5] from Beijing University Of Technology put forward a scroll compressor with auxiliary air intake, and the research results indicated that this system can run steadily with high heating capacity even under a low temperature. The Institute of Thermal Energy Research, Tianjin University [6] application of heat pump with low ambient temperature air source in northern China has been studied.The research shows that in terms of district heating, the heating demand is below 200 KW, the air source heat pump mostly uses multiple vortex compressor or rotor compressor, when the capacity is greater than 200 KW, mostly use multiple units in parallel or use screw unit. Ma et al. [7] from Shandong University verified the existence of cold island effect by numerical simulating the air flow field around the single module heat pump.The study shows that the smooth discharge of the air source heat pump is more conducive to resist the deterioration of the environmental air field, and the arrangement of a certain spacing of diversion blades around the air source heat pump unit has a good weakening effect on the heat pump cold island effect. Wang [8] used CFD (Computational Fluid Dynamics) software to simulate and analyze the mechanism of cold island effect, and proposed that the appropriate height at the bottom of the air source heat pump can not only significantly improve the cold island effect, but also increase the air circulation channel to prevent condensation or freezing the frost water.

Samsung company [9] introduced the turbo surcharge technology in the air conditioner field. It replaced the slotted fins of common air conditioning unit with diamond fins, greatly reducing the energy consumed in initialization phase by expediting the system’s temperature rise/drop velocity. Nobukatsu [10] from Japan raised a turbo compressor flash-tank heat pump system, which was able to enhance units’ heating capacity by about 15% under a low-temperature work condition. Sami and Tulej [11] from Canada brought forward a novel air source heat pump heating system for improving the temperature of the air entering the evaporator for heat transfer, and this system showed a high heating performance coefficients even under a low-temperature environment. Lee et al. [12, 13, 14] used numerical simulation method and integrated with experimental study to analyze the influence of different heat-transfer coil arrangements and fin allocation on the heating performance of heat pump condenser.

The above overseas and domestic research status reveals that all these studies of air source heat pump unit itself mainly concentrate on promoting the performance parameters of air source heat pump under nominal work conditions and have received significant results; in addition, among the studies of fin tube evaporator, most are analyzing the local heat transfer and flow characteristics of fin tubular heat exchanger. Though there have been some investigations about the impact of environmental wind field on air-conditioning air-cooled condenser, it still lacks of the research on the influence of the cold island effect of air source heat pump unit array on the operation of unit array as well as the technical measure for improving cold island effect. Based on this situation, this paper carried out research of the influence of cold island effect on the operation of air source heat pump.

2. Experimental study and model building

2.1 Experimental study

2.1.1 Experiment content and purpose

The experiment content mainly includes: testing the air field at the low temperature side of air source heat pump, collecting the data of nearby wind field during the operation of heat pump unit array, processing the data collected, analyzing the problems of air source heat pump unit array in the running process in winter, and verifying and comparing with the model built to guarantee its correctness and applicability. In the experimental process, the air field near heat pump unit array was measured by measuring apparatus and the experimental data was recorded, to prepare for the further analytical research.

The experiments include:

(1)
Measure and record the current atmospheric environment temperature and wind speed;
(2)
Collect the temperature data of each observation point of air source heat pump unit array.

2.1.2 Experimental facility

The paper investigated an air source heat pump unit array in a heat supply station in Shijiazhuang, which has a boiler room with overall heating floorage of 130,650 m ${}^{2}$ and uses finned radiator. The heating system is operated combining 74 sets of 60 KW air source heat pumps and 2800 KW electric boilers. A total of 74 sets of air source heat pump units covers an area of about 690 m ${}^{2}$ . Due to limited floor area, the units are arranged compactly with small space between them, so the unit array shows evident “cold island effect” in the process of operation. See basic information in the following tables.

Table 1
Basic information of the heat source station

No. of air source heat pumps	Electric boiler power	Total transformer capacity (KVA)	Transformer attached load
74	2800 KW	1600 $+$ 1600	The heat pumps are loaded by a 1600 KVA transformer
(60 KW)		FALSE	The electric boiler is loaded by two 1600 KVA transformers

Table 2

Parameters of required test equipment

Equipment name	QTY	Precision	Measuring range	Remark
Thermometer	3	$\pm$ 0.05 ${}^{\circ}$ C	$-$ 30–50 ${}^{\circ}$ C	Used to measure the ambient temperature of each observation point
Anemograph	1	$\pm$ 1 m/s	0–30 m/s	Used to measure the ambient wind velocity

2.1.3 Experimental method

Multiple observation points were arranged in the heat pump unit array, see the positions of observation points in Fig. 1. Before the experiment, the first thing was to determine the number of heat pumps started, and to check the operating state of heat pumps and if there was any unit in fault. After the unit run steadily for one hour, data can be collected; when the ambient wind field was relatively stable, the temperature of each observation point can be collected and recorded.

Figure 1.

Layout of observation points.

During the experiment, three thermometers worked together to quickly complete the measurement of all observation points, reducing the influence of ambient wind field on the data of observation points and guaranteeing the accuracy of data analysis. Under different ambient temperatures, data were measured several times to analyze the influence of the unit’s operation under different work conditions on the nearby wind field and verify the reasonability of the model built. In the process of experiment, the data of different observation points were read many times to take their averages so as to move the error from the reading process.

Set three observation points A, B and C at the height of 1 m, 2 m and 3 m to measure and obtain the experimental data.

Figure 2.

Temperatures of the observation points measured under an ambient temperature 3.1 ${}^{\circ}$ C, wind velocity 2.3 m/s and 36 units started.

Figure 3.

Temperatures of the observation points measured under an ambient temperature 2.5 ${}^{\circ}$ C, wind velocity 1.8 m/s and 36 units started.

Figure 4.

Temperatures of the observation points measured under an ambient temperature $-$ 0.9 ${}^{\circ}$ C, wind velocity 2.5 m/s and 48 units started.

2.1.4 Experimental results and analysis

After conducting many experiments, we measured the temperature of each observation point under an ambient temperature 3.1 ${}^{\circ}$ C, wind velocity 2.3 m/s, 36 units started; under an ambient temperature 2.5 ${}^{\circ}$ C, wind velocity 1.8 m/s, 36 units started; under an ambient temperature $-$ 0.9 ${}^{\circ}$ C, wind velocity 2.5 m/s, 48 units started; under an ambient temperature $-$ 3.8 ${}^{\circ}$ C, wind velocity 2.5 m/s, 48 units started; and under an ambient temperature $-$ 7.1 ${}^{\circ}$ C, wind velocity 3 m/s, 74 units started. Heights A, B and C are 1-1 to 3-4 where all measuring points are 1 m, 2 m and 3 m respectivelySee details in Figs 2–6.

Figure 5.

Temperatures of the observation points measured under an ambient temperature $-$ 3.8 ${}^{\circ}$ C, wind velocity 3 m/s and 74 units started.

Figure 6.

Temperatures of the observation points measured under an ambient temperature $-$ 7.1 ${}^{\circ}$ C, wind velocity 2 m/s and 74 units started.

By analyzing the data collected under different ambient temperatures, we can see that the temperatures at the observation points 2-2, 2-3, 3-2 and 3-3 are lower, bringing much cold air to accumulate and form the “cold island effect”. Taking the observation point 2-2 as an example, the temperature changes with height under different ambient temperatures, as shown in Fig. 7.

Figure 7.

Temperature changes of observation points with height.

The experimental data shows that in the central part of unit array (observation point 2-2), as the ambient temperature falls down, the temperature gradient increases with height; and as the heat load required at user side rises, the number of units started increases, and the “cold island effect” in central part of unit array also aggravates.

In winter, the heat pumps run under a great impact from ambient temperature, and the heating capacity and performance coefficients of air-cooled heat pump units are closely related to outdoor temperature, that’s they degrade with the falling temperature. And in most cases, the decrease of outdoor temperature means that the heating load of a building will also rise. So, the “cold island effect” as well as the impact of the cold air backflow generated during unit operation on unit operation process cannot be ignored.

2.2 Model building

Take the evaporator module of air source heat pump unit as the research object, and model and figure out the air source unit array module in this project.

2.2.1 Physical model

The paper takes the air source heat pump unit in Fig. 8 as its research object. Its main structural parameters are as follows: the evaporator uses inner grooved copper tube with diameter $\Phi=$ 7 mm $\times$ 0.26 mm and R22 refrigerant within the tube; copper tube bundles are arranged with row distance $d_{1}=$ 21 mm and column distance $d_{2}=$ 18.2 mm, a single tube’s effective length l is 1.4 m, and all tubes are arranged in 44 columns and 3 rows in total. The slotted blades on the tubes are arranged with a distance $d_{3}=$ 1.7 mm, and the fin thickness $\delta$ is 0.105 mm. Two heat transfer coils are arranged in a $V$ shape with an included angle of 45 ${}^{\circ}$ . During operation, the air flows through V-shaped finned evaporator by two sides for exchanging heat, and then cold air is exhausted out by the fans on top of evaporator into ambient environment.

Figure 8.

Air-cooled modular air source heat pump.

Compared to evaporator’s heat exchange tube bundles, the heat pump unit is very big in size, while the distance between tubes of evaporator’s tube bundle is small and intensive. If the overall heat exchanger is modeled and grid-divided, the analysis and calculation process will be very hard. Therefore, the fluid interchange part by evaporator’s air side in the air source heat pump is simplified based on the following assumptions:

The flow condition of the air within the calculation area is a steady-state turbulent flow, without considering the time term of governing equation.

The air is considered as an incompressible fluid, and the viscous dissipation of air is ignored, then the change of fluid density with temperature conforms to the Boussinesq’s Hypothesis, that’s:

$\displaystyle\rho=\rho_{0}\left[{1-\beta\left({T-T_{0}}\right)}\right]$ (1)

in which, $\rho_{0}$ is the fluid density under operating temperature, $T_{0}$ is the operating temperature, $\beta$ is the expansivity.

The heat exchanging region by the evaporator’s air side of air source heat pump is deemed as a porous media. When the air flows through the heat transfer tube bundles, the porous media is used to replace the evaporator tube bundles:

(1)

The influence of other components near the evaporator’s heat exchange tube bundles on fluid flow outside the tube is not taken into consideration;

(2)

The diffusing effect of air is ignored;

(3)

The thermal-conductive resistance of tube wall of finned evaporator’s tube bundles as well as the radiant heat exchange during heat transfer are ignored.

2.2.2 Mathematical model

The governing equations of fluid flow and heat transfer within the whole calculation area are as follows.

Mass conservation equation:

$\displaystyle\frac{\partial\left({\rho u}\right)}{\partial x}+\frac{\partial% \left({\rho v}\right)}{\partial y}+\frac{\partial\left({\rho w}\right)}{% \partial z}=0$ (2)

Momentum conservation equation:

$\displaystyle u\frac{\partial\left({\rho u}\right)}{\partial x}+v\frac{% \partial\left({\rho u}\right)}{\partial y}+w\frac{\partial\left({\rho u}\right% )}{\partial z}=-\frac{\partial p}{\partial x}+\frac{\partial}{\partial x}\left% ({\mu\frac{\partial u}{\partial x}}\right)+\frac{\partial}{\partial y}\left({% \mu\frac{\partial u}{\partial y}}\right)+\frac{\partial}{\partial z}\left({\mu% \frac{\partial u}{\partial z}}\right)$ (3)

$\displaystyle u\frac{\partial\left({\rho v}\right)}{\partial x}+v\frac{% \partial\left({\rho v}\right)}{\partial y}+w\frac{\partial\left({\rho v}\right% )}{\partial z}=-\frac{\partial p}{\partial y}+\frac{\partial}{\partial x}\left% ({\mu\frac{\partial v}{\partial x}}\right)+\frac{\partial}{\partial y}\left({% \mu\frac{\partial v}{\partial y}}\right)+\frac{\partial}{\partial v}\left({\mu% \frac{\partial v}{\partial z}}\right)$ (4)

$\displaystyle u\frac{\partial\left({\rho w}\right)}{\partial x}+v\frac{% \partial\left({\rho w}\right)}{\partial y}+w\frac{\partial\left({\rho w}\right% )}{\partial z}=-\frac{\partial p}{\partial x}+\frac{\partial}{\partial x}\left% ({\mu\frac{\partial w}{\partial x}}\right)+\frac{\partial}{\partial y}\left({% \mu\frac{\partial w}{\partial y}}\right)+\frac{\partial}{\partial v}\left({\mu% \frac{\partial w}{\partial z}}\right)-\rho g$ (5)

Energy conservation equation:

$\displaystyle u\frac{\partial\left({\rho T}\right)}{\partial x}+v\frac{% \partial\left({\rho T}\right)}{\partial y}+w\frac{\partial\left({\rho T}\right% )}{\partial z}=\frac{\partial}{\partial x}\left({\frac{\lambda}{C_{P}}\frac{% \partial T}{\partial x}}\right)+\frac{\partial}{\partial y}\left({\frac{% \lambda}{C_{P}}\frac{\partial T}{\partial y}}\right)+\frac{\partial}{\partial z% }\left({\frac{\lambda}{C_{P}}\frac{\partial T}{\partial z}}\right)$ (6)

in which, $\rho$ is the fluid density, in kg/m ${}^{3}$ ; $p$ is the pressure, in Pa; $\mu$ is the dynamic viscosity, Pa $\cdot$ s.

2.2.3 Fan model

The axial flow fan on top of the air source heat pump unit applies the simplified Fan model with lumped parameters, which simplifies the boundary as an infinitely thin plane and sets the model parameters by data obtained in the experiment, so as to ensure the effect of heat pump model conforming to actual situation.

2.2.4 Evaporator model

For the heat exchanger, the aggregate parameter model-radiator model (Radiator) of the heat exchanger unit in Fluent is often used to simulate the fin bundle of the air-cooled evaporator, which mainly considers the heat exchange coefficient and the resistance loss to calculate the flow of the flow field near the heat exchanger.

The pressure loss for the heat exchanger as:

$\displaystyle\Delta p=k_{L}\frac{1}{2}\rho v^{2}$ (7)

in which, $\Delta p$ is Pressure drop of air flow through the evaporator bundle, Pa; $\rho$ is air density, kg/m ${}^{3}$ ; $k_{L}$ is dimensionless loss coefficient can be constant, piecewise linear function and polynomial; $v$ is normal velocity component, m/s.

For the heat exchanger,

$\displaystyle q=h\left({T_{f}-T_{\textit{ext}}}\right)$ (8)

in which, $q$ is density of heat flow rate, $W/m^{2}$ ; $h$ is convection heat exchange coefficient, $W/(m^{2}\cdot K)$ , Constant, piecewise linear function and polynomial representation can be used; $T_{f}$ is reference temperature of the fluid; $T_{\textit{ext}}$ is evaporator temperature, $K$ .

For the heat exchanger, the lumped parameters model-radiator model is often used to simulate the air-cooled finned tube bundles, and the heat transfer coefficient and resistance loss of heat exchanger model are the main considerations in calculating the fluid interchange of liquid flow field near the heat exchanger.

In view of the existence of a large number of heat exchange tubes and fins in the heat exchanger, if the whole of it is grid-divided, the analysis and calculation process will be too complicated. To solve this problem, Patankar and Spalding [15] proposed a porous media model to simulate the heat transfer process of a heat exchanger. The function of the porous media model is to add a source item to the momentum equation for simulation. The source item is composed of an inertia resistance item and a viscous resistance item, expressed as:

$\displaystyle S_{i}=-\left(\sum_{j=1}^{3}D_{ij}\mu\nu_{1}+\sum_{j=1}^{3}C_{ij}% \frac{1}{2}\rho\nu_{\textit{mag}}\nu_{j}\right)$ (9)

in which, $C$ and $D$ are given matrices; $S_{i}$ is the source items of the $i$ -th momentum equation.

For simple porous medias, only the diagonal elements of matrices $C$ and $D$ are retained; for simple and even porous medias, the resistance characteristics are the same at each direction and the diagonal elements are equal. Their mathematical model is expressed as:

$\displaystyle S_{i}=-\left({\frac{\mu}{\alpha}\nu_{i}+C_{2}\frac{1}{2}\rho\nu_% {\textit{mag}}\nu_{j}}\right)$ (10)

in which, $\alpha$ is the permeability coefficient, $C_{2}$ is the inertia resistance coefficient, in $m^{-2}$ .

On account of the complex structure of finned tube evaporator, the paper referred to the test data in literature of Wang et al. [16], taking the inertia resistance coefficient as 209 m ${}^{-2}$ and the viscosity coefficient as 1.6 $\times$ 10 ${}^{-7}$ m ${}^{-2}$ .

The paper defined the viscous resistance directions of evaporator model as: the first direction is parallel to the direction of evaporator inlet plane, and the second direction is perpendicular to the direction of evaporator inlet plane. Due to the anisotropic porous media of evaporator, the media resistance of other two directions except the direction perpendicular to evaporator is infinite, set as 10,000 times of that perpendicular to evaporator direction. Regarding the heat transfer when the air flows through finned tube evaporator, the paper approximately considers it as porous media containing a cold source, so the air will have heat convection with it when flowing through the porous media.

2.2.5 Model building of heat pump unit

Heat pump unit grid division and boundary condition setting

The paper based on actual engineering situation to model the air source heat pump unit array with Gambit software. In simulated calculation, we chose to build models of induced axial fan and V-shaped finned tube bundle evaporator below it in a rectangular coordinate system, and used axial fan to drive the ambient air to exchange heat with evaporator’s tube bundle when flowing through the evaporator, then the cold air after heat transfer was discharged out from the unit top.

Figure 9.

Heat pump unit module model.

As shown in Fig. 9, an air source heat pump unit module was built with 2 m $\times$ 1 m $\times$ 0.8 m footing foundation, 2 m $\times$ 1 m $\times$ 1.2 m evaporator, and dia. 0.8 m fan. To guarantee the fluid field homogeneity surrounding the model and solving accuracy, the entire fluid domain space was set as 10 m $\times$ 10 m $\times$ 20 m, and the heat pump model was located at the central part of calculation area. The calculation area took the negative direction of axis x as the north. Under windless work condition, the four sides around and top surface were set as the pressure-outlet boundary condition, the ground and heat pump bottom foundation were set as the adiabatic wall, and ambient temperature was set as 266 K; under windy work condition, the boundary of wind entering side was set as the velocity-inlet boundary condition, and other sides around and top surface were still set as the pressure-outlet boundary condition.

The division of grid adopts the method of the method of partition division of grid to ensure the grid quality. For the flow field with large calculation area, it is suitable to use the tetrahedral unstructured grid generated by TGRID method, which has no obvious rows and columns in the spatial distribution. Take a grid division of different sizes for the heat pump unit and its surrounding flow field domain.

Validation of heat pump unit model

Carry out numerical calculation of this model in Fluent. Set ambient temperature as 266 K, ambient fluid field as wind coming from the north, wind velocity as 2 m/s. The standard turbulence model $k-\varepsilon$ was adopted in the calculation, the second order upwind discretization scheme was used in numerical calculation, and the pressure velocity coupling was solved by SIMPLE algorithm. To validate the precision of the mathematical physical model built and the selected calculation model, the paper compared the results of numerical simulation and the measured experimental data. In Fig. 10, it is found that the maximal and minimal velocity derivations along fan axis are 9.2% and 1.9%. These derivations are in the allowable range of engineering practice, so it can be considered that the model built is correct.

Figure 10.

Comparison of test and simulated wind velocity values along fan axis.

2.2.6 Model building of heat pump array

Take the real air source heat pump array in this engineering as the research object, and conduct numerical simulation on the air field flowing condition under its operating state. The current layout of this engineering (as below) is: the columns from west to east are defined as Columns 1 $\sim$ 5, and those from north to west are defined as Rows 1 $\sim$ 17. Column 1 is taken as a separate group, including six heat pumps; Columns 2 $\sim$ 5 include seventeen heat pumps each, and each column is divided into three groups, that’s Row 1 $\sim$ 6 (including six heat pumps), Row 7 $\sim$ 12 (including six heat pumps), Row 13 $\sim$ 17 (including five heat pumps). The heat pumps between groups are connected in series, and the heat pumps between rows are arranged with a spacing of 440 mm.

Figure 11.

Layout of air source heat pump units in the heat source station.

To ensure the homogeneity and accuracy of the entire fluid field, the overall calculation region of air-cooled platform is set as 30 m $\times$ 20 m $\times$ 10 m. Establish the unit array model as shown in the figure below.

Figure 12.

GAMBIT model of air source heat pump unit array.

The air source heat pump unit array model built received a simulated calculation in Fluent, with ambient temperature set as 266 K, north upwind, and wind velocity set as 2 m/s; the standard $k-\varepsilon$ turbulence model was adopted in the calculation, the second order upwind discretization scheme was used in numerical calculation, and the pressure velocity coupling was solved by SIMPLE algorithm. To validate the precision of the mathematical physical model built and the selected calculation model, the paper compared the results of numerical simulation and the measured experimental data, as shown in the Figureure below.

Figure 13.

Comparison of simulated data and measured value of each measuring point.

The above Measuring point height-Temperature curves indicate that, the major errors in the simulation are concentrated in measuring points 2-2 and 3-4. The error caused in measuring point 2-2 is resulted from that in the model establishment process, the heat pump unit efficiency decreases due to evaporator frosting under a low-temperature work condition, leading to a slightly higher measured value than simulated data; while, the error caused in measuring point 3-4 lies on the setting of north upwind. As we know, in practical measurement, we can’t ensure an absolutely stable ambient wind field. The change of ambient wind disturbs the low-temperature air discharged from the heat pump unit array, bringing a large error between measured value and simulated data. In other measuring points, the measured temperatures can well coincide with simulated data. These comparison and analysis demonstrate that in numerical simulation, the boundary conditions set for heat pump module as well as the simplified assumption equation of model are reasonable, and the numerical simulation results can be used to analyze the nearby environmental field condition under operation state of heat pump unit array in this engineering.

3. Analysis of periphery fluid field of air source heat pumps

In regional heating engineering, an air source heat pump system often uses multiple heat pump units arranged in an array. In practical arrangement, considering the site and space factors, units are always arranged very closely. For a single heat pump unit, subject to the impact of side wind, the cold air discharged from fan will deflect to one side, forming a turbulent flow near heat exchanging tube bundle at the leeward side of heat pump; but in the heat pump unit array, the cold air discharged form unit fan at weather side will have an effect on the air field near heat exchanging tube bundle of heat pump at the leeward side under the influence of side wind.

Figure 14.

Temperature distribution of heat pump unit section in Column 1.

Figure 15.

Temperature distribution of heat pump unit section in Column 2.

Figure 16.

Temperature distribution of heat pump unit section in Column 3.

Figure 17.

Temperature distribution of heat pump unit section in Column 4.

Figure 18.

Temperature distribution of heat pump unit section in Column 5.

From the temperature distribution curves of heat pump units in Columns 1 $\sim$ 5 at 2.5 meters height, we can see that the cold air mass over the units is smaller in Columns 1 and 5 surrounding the unit array, while the cold air mass over the units aggregates in Columns 3 and 4 in the central part of unit array, forming an obvious cold island effect. Suffering from the impact of cold island effect, for heat pump units in central part of unit array, the fluid field temperatures near their evaporators’ heat exchanging tube bundle inlet are lower than ambient atmospheric temperature, and the cold-temperature air temperature will further fall down flowing through evaporator, greatly influencing the heat transfer performance of units in central region.

Though both Columns 1 and 5 are at the outmost edge of unit array, they contain different number of units. The six heat pumps in Column 1 exhaust cold-temperature air with little variation in temperature; but from Row 6 of Column 5, the units are affected by other units in the array, so the temperatures of their exhausted low-temperature air are lower than those by unit outlets at the upwind side. Thus, the more the units are arranged, the greater the impact between units will be, the lower the temperature of air field around central units will be.

When the units are running, the ambient air at both ends of the array under the effect of pressure flows from relatively high-pressure environment to the low-pressure area above unit array; on contrary, the low-pressure area in central region under the effect of pressure gradient makes the low-temperature air exhausted from surrounding units to flow into the central part, leading to the accumulation of cold air and the formation of cold mass in central region. Besides, a low-pressure region generates near the evaporator inlet of central units under the effect of axial fans, making the cold air mass concentrating in central region to flow through its evaporator’s heat exchanging tube bundles for heat convection, which brings a circulated deterioration to fluid field temperature near heat pumps.

In a word, in an air source heat pump unit array, the intensively arranged units result in uneven distribution of pressure field in the array and severe backflow phenomenon of cold air in central region. The low-temperature air exhausted from surrounding heat pump units’ fans accumulates in central region, forming an obvious cold island effect. Furthermore, due to many heat pump units arranged in the array, the fresh air in the air duct cannot meet the fresh air demand of heat pump unit array for heat transfer, giving rise to a decline of aspiration capacity of central units. This eventually produces a bad impact on the overall operation of air source heat pump unit array, and leads to an insufficient output of units, thus failing to satisfy the heating needs in practical operation.

4. Technical measures for improving cold island effect of air source heat pump unit array

To improve the cold island effect of air source heat pump unit array, three transformation schemes are proposed on theoretical basis: (1) increase the distance between heat pump units in the array; (2) add isolation baffles; (3) add isolation baffles and elevate the heat pump units, carry out relevant calculation on each transformation scheme, and discuss on the distribution of peripheral fluid field of unit array under different schemes as well as the influence of different schemes on periphery fluid field during unit operation.

4.1 Increase the distance between heat pump units in the array

When the heat pump unit array is arranged intensively, the low-temperature air exhausted from the fans of peripheral heat pumps always flows back to the heat pump’s evaporator inlet in central region, affecting the heat transfer of heat pump units in central region and further affecting the units efficiency. Taking the different arrangement distances of heat pump units as contrast, a research analysis was made on the operation condition of heat pump array with distance arrangement of 0.5 m, 0.7 m, 0.9 m, 1.1 m, 1.3 m and 1.5 m. The entire watershed space is set at 30 m 10 m 20 m, and the heat pump model is located at the center of the calculated area.The model schematic diagram is shown below.

Figure 19.

Schematic diagram of heat pump array model.

The row spacing W is 0.5 m, 0.7 m, 0.9 m, 1.1 m, 1.3 m, 1.5 m and takes the axis section of the second column fan as the object, and the temperature distribution map is intercepted as follows.

Figure 20.

The temperature distribution cloud diagram of the axis section of the fan in the second column with a row spacing of 0.5 m.

Figure 21.

The temperature distribution cloud diagram of the axial cross section of the fan in the second column with a row spacing of 0.7 m.

Figure 22.

The temperature distribution cloud diagram of the axis section of the fan in the second column with a row spacing of 0.9 m.

Figure 23.

The temperature distribution cloud diagram of the fan axis section in the second column of 1.1 m row spacing.

Figure 24.

The temperature distribution cloud diagram of the axis section of the fan in the second column with a row spacing of 1.3 m.

Figure 25.

The temperature distribution cloud diagram of the axis section of the fan in the second column with a row spacing of 1.5 m.

From the figure above comparison, can clearly in hours between the layout, the mutual influence between heat pump units is serious, cross air flow directly affects the air temperature, low, low temperature air flow through the evaporator heat transfer, located in the center of the array heat pump unit operating at low temperature environment conditions, low temperature mass accumulation in the array center area, forming obvious cold island effect.With the increase of the spacing, the exhaust temperature of the air outlet of the heat pump unit in the central area increases, and the cold island effect gradually attenuates.

In order to visually measure the cold island effect degree of heat pump unit array when it is operated in winter, a non-dimensional parameter uniformity $k$ is defined to describe the non-uniformity of air temperature at the heat pump’s evaporator inlet in different positions of array. According to the definition of standard deviation, the equation of non-uniformity is introduced as below:

$\displaystyle k=\sqrt{\frac{\mathop{\sum}\nolimits_{i=1}^{n}\delta^{2}\left(i% \right)}{n-1}}\times 100\%$ (11) $\displaystyle\delta\left(i\right)=\frac{T_{i}-T_{\textit{avg}}}{T_{\textit{avg% }}}$ (12)

in which, i refer to the ith heat pump unit, n refers to total number of heat pump units in the array; T represents the average temperature at evaporator inlet of heat pump unit; $\delta(i)$ refers to the relative deviation between actual and mean values of average temperature parameter at evaporator inlet of each heat pump unit.

Take the average temperature of evaporator inlet of heat pump unit array with different row distances, substitute it into above equations, Figureure out the non-uniformity of average temperature at evaporator unit inlet when the heat pump units with different row distances are running, and work out the maximum difference between ambient atmospheric temperature and the average temperature of evaporator inlet under different arrangement distance. See the curves as below.

Figure 26.

The change curves of temperature non-uniformity of evaporator inlet with row distance.

Above curves indicate that when row distance increases from 0.5 m to 0.7 m, the air temperature non-uniformity of evaporator inlet has little change; but when the row distance increases from 0.9 m to 1.1 m, the non-uniformity increases largely, and the cold island effect is relieved obviously. Thereby, when the unit row distance is less than 1 m, the units may produce a great impact on each other, resulting in a local low-temperature region in the array. While, as the row distance increases, the temperature distribution gradually becomes uniform, the lowest temperature in the array gradually rises, and the cold island effect shall gradually be weakened. When row distance is between 1.3 m and 1.5 m, the non-uniformity change rate is slowed down and the non-uniformity degree reduces below 0.5; when row distance increases from 0.5 m to 1.3 m, the temperature non-uniformity of evaporator inlet declines by 62.9%; when row distance increases to 1.5 m, the temperature non-uniformity of evaporator inlet declines by 69.3%, and the difference between ambient atmospheric temperature and average temperature of evaporator unit inlet in central region is already lower than 1.5 k, proving that the cold island effect has been reduced effectively. From these data, it is believed that in practical engineering, an arrangement distance of heat pump unit array less than 1.3 m can effectively mitigate the cold island effect generated during operation in winter.

4.2 Add isolation baffles in heat pump unit array

In practical engineering, due to many heat pump units, it is difficult to provide a satisfactory site for reasonable arrangement distance. To solve the interaction effect between units in the array and effectively guide wind field energy, it is proposed to add isolation baffles in the array. Isolation baffles can separate air outlet and inlet of heat pump units, prevent low-temperature air turbulent flow exhausted from air outlet to flow back near the evaporator inlet, and stop the low-temperature air mass aggregating above heat pump unit array to sink near evaporator inlet under the effect of buoyancy lift, thereby improving the impact of cold island effect generated by heat pump array during heating operation in winter on the running of heat pumps.

Regarding the 4 $\times$ 10 heat pump unit array with row distance 0.5 m and column distance 2 m before and after adding isolation baffles, the evaporator inlet’s temperature non-uniformity, fan’s air aspiration non-uniformity and total air aspiration of unit array fan during the operation of heat pump unit are compared.

Figures 27 and 28 respectively show the temperature distribution of the fan axis section of the second column of heat pump units when the row spacing is 0.5 m. As can be seen from the two figure comparison, the heat pump unit in the middle area, near the temperature of the temperature is significantly lower compared with the temperature of the heat pump evaporator inlet, and the temperature of the whole heat pump evaporator inlet temperature distribution is relatively uniform.After installing the separator above the heat pump array still appear low temperature gas accumulation formation cold island effect, but due to the existence of the separator, the heat pump evaporator inlet area and fan outlet area, low temperature air mass can not settle to the heat pump evaporator inlet area, make the array of each heat pump unit evaporator inlet area temperature distribution is more uniform.

Figure 27.

The temperature distribution cloud diagram of the second column of the fan axis section with a baffle plate installed at 0.5m row spacing.

Figure 28.

Temperature distribution cloud diagram of the axis section of the fan in the second column with 0.5 m row spacing.

Similar to the calculation of temperature non-uniformity at evaporator inlet, it is feasible to take the air aspiration amount of each heat pump in the array and substitute them into the non-uniformity equation to obtain the air aspiration non-uniformity of heat pump fan in the array. For a heat pump unit array with row distance 0.5 m before and after adding isolation baffles, its evaporator inlet’s temperature non-uniformity, fan’s air aspiration non-uniformity and total air aspiration of unit array fan during unit operation are shown in Table 3.

Table 3

Data comparison before and after adding isolation baffles

	Evaporator inlet’s temperature non-uniformity	Fan’s air aspiration non-uniformity	Fan’s total air aspiration kg/s
Before adding isolation baffles	1.1301	0.0623	405.4
After adding isolation baffles	0.0377	0.1660	385.9

Above table demonstrates that isolation baffles added significantly promote the temperature uniformity of evaporator inlet during unit operation. On contrast, the isolation baffles seriously bring down the air aspiration uniformity of heat pump fans, and total air aspiration capacity declines by 4.9%, which reduce the heat absorption efficiency of heat pumps at the air side.

4.3 Add isolation baffles and elevate the units

In case of a heating engineering project demanding for larger heat load and more heat pump units, adding isolation baffles will restrict fresh air obtained during unit operation, leading to a reduction of both total air aspiration capacity of fan during operation and heat transfer performance of heat pump unit.

For the sake of increasing fresh air flowing into unit array and meeting the need of fresh air in heat exchange during unit operation, it is proposed to elevate the unit array in addition to adding isolation baffles. Through expanding air ducts, the lacking of air flow in central region during operation can be improved.

Figure 29.

Layout of elevated heat pump unit array.

Regarding the 4 $\times$ 10 heat pump unit array with row distance 0.5 m and column distance 2 m after adding isolation baffles, all units are elevated 2.5 m above the ground. A fresh air region covering about 500 m ${}^{3}$ is set at the bottom of unit array, and the fresh air flow velocity required at side surfaces is 2 m/s $\sim$ 3 m/s, which can satisfy the required air flow for heat exchange during operation. The operation conditions of heat pump unit array before and after elevation is compared and analyzed.

The data of elevated and non-elevated heat pump unit array during operation is shown in Table 4.

Table 4

Data comparison of elevated and non-elevated unit arrangement

	Evaporator inlet’s temperature non-uniformity	Fan’s air aspiration non-uniformity	Fan’s total air aspiration kg/s
Add isolation baffles	0.0377	0.1660	385.9
Add isolation baffles and elevate the units for 2.5 m	0.0225	0.0665	409.3

Above table shows that, by comparing the data of heat pump unit array added with isolation baffles with that of elevated units added with isolation baffles, the temperature non-uniformity of evaporator inlet during unit operation only differs by 0.0152. And due to the existence of isolation baffles, the interaction between units during operation has been weakened to a large extent, so elevating the units has little impact on the temperature non-uniformity of evaporator inlet in the array. However, after elevating the units and setting air ducts at the bottom of unit array, fresh air is greatly promoted in the air duct during unit operation, air aspiration non-uniformity of each heat pump fan in the array decreases significantly, and total air aspiration rises evidently.

Table 5 displays the data of three different arrangement forms of air source heat pump unit array during operation in winter.

Table 5

Data comparison of three different technical measures

	Evaporator inlet’s temperature non-uniformity	Fan’s air aspiration non-uniformity	Fan’s total air aspiration kg/s
Distance 0.5 m	1.1301	0.0623	405.4
Distance 1.5 m	0.3537	0.0301	409.4
Distance 0.5 m, isolation baffles added	0.0377	0.1660	385.9
Distance 0.5 m, isolation baffles added, elevated for 2.5 m	0.0225	0.0665	409.3

Table 5 compares the evaporator inlet’s temperature non-uniformity, fan’s air aspiration non-uniformity and fan’s total air aspiration of heat pump unit array with four arrangement forms: (1) Distance 0.5 m; (2) Distance 1.5 m; (3) Distance 0.5 m, isolation baffles added; and (4) Distance 0.5 m, isolation baffles added, elevated for 2.5 m. Among the four arrangement forms, the fourth form obtains the lowest temperature non-uniformity at evaporator inlet, the second form obtains the lowest air aspiration non-uniformity of fan, and the forms 4 and 2 obtain similar total air aspiration.

5. Conclusions

Regarding the cold island effect occurred in the air source heat pump unit array of one heat source station in Shijiazhuang during heating operation in winter, this paper took the evaporator of air source heat pump units as research object, used experimental research method combined with numerical simulation, to simulate the distribution condition of temperature and pressure fields of peripheral air field when the air source heat pump unit array of the engineering is operated in winter, and to analyze the physical mechanism of cold island effect as well as its influence on air source heat pump heating system. On this basis, different technical measures such as different arrangement distance, adding isolation baffles and elevating the units were analyzed, and the advantages and disadvantages of all these optimization measures were compares, finally the optimum application condition of different optimization measured were found out.

If the site permits, the arrangement distance between units should be increased, and the distance between any two units shall be no less than 1.3 m. After adding isolation baffles to the heat pump unit array, the evaporator inlet temperature non-uniformity reduces by 66.7%, effectively limiting the generation of cold island effect. After adding isolation baffles and meanwhile elevating the units, the problem of lacking fresh air in central region when only isolation baffles are added to air source heat pump array can be effectively solved, and the air aspiration of overall units increases by 6.1% compared to the arrangement mode of adding isolation baffles without elevating the units. Nevertheless, this design method requires to consider many factors and costs higher, so it is suitable for the heating projects with greater heating load and more heat pump units that are installed in some building intensive sites such as downtown and large-scale community, to guarantee the stable operation of air source heat pump units.

Footnotes

Acknowledgments

This work is supported by the Central Government Guides Local Funds for Scientific and Technological Development Project (206Z4502G).

This work is supported by the Central Government Guides Local Funds for Scientific and Technological Development Project (216Z5201G).

References

The Xinhua News Agency. The 14th Meeting of the Central Leading Group for Financial and Economic Affairs[EB/OL]. (2016-12-21) [2017-08-17].

Liu

Hou

. Study on performance of household air source heat pump water-heater. Construction Science and Technology. 2019; 376(2): 42-44.

Shen

Song

. Analysis and technical measures of application expansion of air source heat pump. The HVAC Unit is Available. 2002; 32(6): 37-39.

Tian

Shi

Wang

. Study on two-stage compressed variable frequency air source heat pump applied in cold region. Acta Energiae Solaris Sinica. 2004; 25(3): 288-293.

Shao

. Study on heat pump for air conditioning in cold regionActa. Energiae Solaris Sinica. 2002; 23(1): 17-21.

Chen

Yang

. A novel layout of air-cooled condensers to improve thermo-flow performances. Applied Energy. 2016; 165.7.

Wang

Luo

. Application of Low Environment Temperature Air Source Heat Pump in North District Heating. District Heating. 2018(4): 14-18.

Wang

. Study on Cold Island Effect and Influence of Environmental Wind Field on Operating Performance of Air Source Heat Pump Array. Shandong University, 2019.

Lei

Chen

. Mechanism and control measures of cold island effect in the cluster-arranged heating system of air source heat pump. Energy saving and Environmental Protection. 2021(5): 51-54.

10.

Nobukatsu

. Scroll Compressor and its Application to Packaged Air-conditioner. Hitachi Comments. 1983; 65(6).

11.

Sami

Tulej

. A new design for an air-source heat pump using a ternary mixture for cold climates. Heat Recovery Systems and CHP. 1995; 15(6).

12.

Lee

T-S

W-C

Chuah

Y-K

Wang

S-K

. An improvement of airflow and heat transfer performance of multi-coil condensers by different coil configurations. International Journal of Refrigeration. 2010; 33(7).

13.

Lee

T-S

W-C

Wang

S-K

. Improved energy performance of air-cooled water chillers with innovative condenser coil configurations – Part I: CFD simulation. International Journal of Refrigeration. 2012; 35(8).

14.

Lee

T-S

W-C

Jiang

J-C

. Improved energy performance of air-cooled water chillers with innovative condenser coil configurations – Part II: Experimental validation. International Journal of Refrigeration. 2012; 35(8).

15.

Patankar

Spalding

. A calculation procedure for heat, mass and momentum transfer in three-dimensional parabolic flows. Pergamon. 1972; 15(10).

16.

Wang

Ren

. Optimization of V-shape arranged heat exchanger. Refrigeration and Air-Conditioning. 2013; 13(9): 35-37+40.

Technical measures for improving the cold island effect of air source heat pump unit array

Abstract

Keywords

1. Introduction

2. Experimental study and model building

2.1 Experimental study

2.1.1 Experiment content and purpose

(1) Measure and record the current atmospheric environment temperature and wind speed; (2) Collect the temperature data of each observation point of air source heat pump unit array. 2.1.2 Experimental facility

Table 1 Basic information of the heat source station

2.2.1 Physical model

2.2.4 Evaporator model

4.1 Increase the distance between heat pump units in the array

Footnotes

Acknowledgments

References

(1)
Measure and record the current atmospheric environment temperature and wind speed;
(2)
Collect the temperature data of each observation point of air source heat pump unit array.

2.1.2 Experimental facility

Table 1
Basic information of the heat source station