Challenge of Using Groundwater for Buildings Air Conditioning in Subtropical Areas

Abstract

Using circulating groundwater to cool air-conditioning is not new in high latitude regions but difficult in subtropical areas. Different from only using fans to remove the heat from indoor air for drier air in the high latitude region, the latent heat inside the humid air in subtropical areas makes the operation more difficult. Latent heat inside the humid air must remove away by air-conditioning including compressor and fan for cooling indoor air, which means more electrical power is required for the operation. To save total electrical power for the air-conditioning system is the main goal of this study. To use the advantage of groundwater with lower temperature to lower down the work of compressor, this research compared two ways, close/open types of water/groundwater circulation, both using groundwater to remove the heat generated by a 15RT (45 kW) air-conditioning. Full-scale tests and simulations were performed in this study to evaluate the efficiency of transferring the heat produced by air-conditioning systems to stably flowing groundwater in a grave stratum under Taipei Basin. With a closed circulating cooling water system, this study found that a 15RT air conditioner could only operate continuously for 4 h before it had to be shut down due to overheating. Additionally, groundwater must carry the heat away within the following 20 h. In changing the closed circulating water system to an open one, a system that uses a circulatory method to extract groundwater upwards and conduct heat exchange with an air conditioning system can enable the continuous operation of such a system with the same heat production condition. Numerical simulations for the heat dissipation behavior of two circulatory systems were performed herein. The results verified the aforementioned phenomena observed from both tests. The result showed both systems can provide air-conditioning working well. The total electrical power for a 15RT air-conditioning in sub-tropical areas can be reduced by 22% using circulating groundwater. Considering the system optimization, the total power consumption can be reduced by about 28%.

1. Introduction

Due to its basin topography, the metropolitan area of Taipei must rely on air conditioning systems to reduce indoor temperatures to a comfortable level. Consequently, a considerable amount of waste heat is emitted by air conditioning equipment to the atmosphere and such waste heat cannot be easily dissipated through atmospheric flow, causing considerably increased time- and intensity-related usage of air conditioning systems. As a result, a marked annual increase in the environmental temperature of Taipei Basin has been observed [1], resulting in a heat island effect. Similar problems have been observed in large cities worldwide [2,3,4,5]. The cooling towers of buildings have been used as a heat exchange medium between air conditioners and outdoor air and make the cooling cycle more efficient for air condensation in subtropical areas. However, this practice may discharge large amounts of waste heat into the air and causes noise pollution. Considering the difficulty of airflow within basins, cooling tower operation can prompt the surrounding air to quickly reach dynamic heat saturation. This rapidly increases temperature and reduces the heat dissipation efficiency of cooling towers. Consequently, the sizes of cooling towers have been increased repeatedly, and this has resulted in land usage problems and complaints from residents. To use the alternative and capable medium for cooling air-conditioning system is an important work to reduce the heat-island issue. Kappler et al. and Emad et al. also studied such an idea using groundwater as a heat exchanger [6,7]. Aprianti et al. perform a comparison of ground and air source heat pumps and found the cost-of-performance of groundwater is better than air [8].

To use ground or groundwater is considered to deal with the above issue. A similar idea may be carried out using such a BTES or ATES system. BTES, Borehole thermal energy storage, is one of the common methods used for seasonal TES currently employed around the world. BTES involves using the ground as the storage medium, allowing heat to be added to the ground during the summer months and extracted to meet the heating demands in the winter heating season. ATES, Aquifer thermal energy storage, is an innovative open-loop geothermal technology. It relies on the seasonal storage of cold and/or warm groundwater in an aquifer. The technology was developed in Europe over 20 years ago and is now in use at over 1000 sites, mostly in the high latitudes’ region. According to international observations, the relatively large temperature difference can be leveraged to achieve cooling without the flow of groundwater because the temperatures of strata are sufficiently low. However, many urban such as Taipei city in Taiwan are located between subtropical and tropical zones that have relatively high temperatures and humidity during summer. Therefore, air conditioning systems are necessary to achieve comfortable temperatures. Such systems have stringent requirements related to the medium for removing indoor heat. Piga et al., Sedaghat et al., Todorov et al., and Al-Madhlom et al. studied the use of TES for different areas [9,10,11,12,13]. Song et al. used an indoor sandbox test to simulate different operation modes of pumping and recharging well and found the results were well [14].

Taipei Basin contains about 85 billion tons of groundwater [15]. This water has a low temperature of 24–28 °C [16] and could be harnessed to replace the recycling water of cooling towers for the original central air conditioning systems. That is, using the water circulated between the air conditioning systems and aquifers to diffuse the heat of water used for cooling. The groundwater of the Taipei Basin remains at a low temperature throughout the year. By contrast, areas, where water towers are used, are usually exposed to sunlight and high temperatures (the daytime temperature during summer can easily exceed 35 °C). Therefore, groundwater from Taipei Basin has considerable advantages for use in air conditioning equipment in terms of electricity conservation. Additionally, if the required extraction pump and heat exchanger for circulating groundwater can be appropriately designed and applied, energy conservation can be achieved in both air conditioners and extraction pumps. Notably, the temperature of groundwater is stable all year round, and usually 5 °C to 7 °C lower than the air temperature in summer, which enables electricity cost savings related to air conditioning equipment.

To evaluate the feasibility of using circulating groundwater to cool air-conditioning and the behavior of dissipated heat carried by groundwater in a gravel aquifer, this study performed a series of numerical simulations and full-scale experiments.

2. Groundwater of Taipei Basin

The area under the 20-m surface elevation of Taipei Basin is approximately 243 km². The main river within the basin is Tamsui River, which consists of a confluence of Keelung River, Dahan River, and Xindian River. Tamsui River is located at the junction of the confluence of Dahan River and Xindian River. Keelung River joins the confluence of the two aforementioned rivers near Guandu.

In terms of the planning of alluvial sequences under the Taipei Basin, the Chingmei Formation mainly has thick-bedded lateritic gravels of alluvial fan facies with slate gray gravels placed between river facies. A significant number of conglomerates are present in the southeast portion of the basin, the majority of those are quartz, followed by sandstone, with igneous rocks appearing on occasion. The northwest portion of the basin is composed of medium and small gravel. The thickness of such sedimentary rocks can exceed 50 m. The ground surface of the Chingmei Formation is the bottom surface of the Songshan Formation, whereas the bottom surface of the Chingmei Formation is the ground surface of the Wugu Formation; both surfaces are composed of sand and mud. The ground surface of the Chingmei Formation gradually deepens from south to north, reaching a depth exceeding 100 m between Wugu and Luzhou. As for other districts in Taipei City, the ground surface has an average depth of 40–55 m. The thickness of the ground surface of the Chingmei Formation can reach 135 m in Chingmei District and gradually decreases from south to north. The surface does not reach Songshan District, which is located to the north of Chingmei [17]. The main recharge source for the groundwater of the Chingmei Formation is the Xindian River, followed by the Dahan River. The vertical leakage of highlands and tablelands from the basin margin is minimal, and recharge from Keelung River is negligible [18,19]. The annual recharge of groundwater from such a basin is approximately 1.2 × 108 m³ [15].

This study collected drilling data from the Central Geological Survey to construct a hydrogeological model of the Chingmei Formation. Water level data for 1996–2002 from the Groundwater Monitoring Network of the Water Resource Agency were used to calibrate relevant parameters for numerical simulation. This study found that after the groundwater flowed from the self-recharging zone to the north and into Taipei Basin, it gradually changed direction to the west before flowing into the deepest area of the Chingmei Formation, located in Wugu District (Figure 1). By contrast, the water elevation of the Chingmei Formation (piezometric head in the confined aquifer) from its head to its deepest location ranged between elevation level (EL) 7 m and (EL) −14 m. That is, the pressure distribution difference of the confined aquifer was 21 m, and the hydraulic gradient (i) was estimated to be 0.0076%. Based on the value, the estimated water budget value of the underground basin underground was 3.54 × 104 cmd (m³/day). The National Taiwan University of Science and Technology (Taiwan Tech) is located at the southeast corner of the Taipei Basin. In November 2011, current flow velocity and flow direction tests were conducted in a deep on-campus well that penetrates the Chingmei Formation. The results showed that this groundwater mostly flows in southeast and northwest directions. The flow velocity of the well was approximately 0.005 m/s, 1000 times higher than the simulation result. The velocity was possibly due to the considerable amount of pumping within the basin at that time. A pumping test found that hydraulic conductivity (T) was approximately 0.0003 m²/s.

3. Experimental System and Arrangement

The gravel stratum underneath the study site was located 39 to 60 m below the ground surface. Inside the gravel stratum, there was a varve clay layer of 4 m thick in ground level (GL.) −46 to −50 m dividing the gravel layer into the upper layer (Chingmei stratum) and lower layer (Banciao stratum). The groundwater levels of the upper and lower layers were located at GL −12.5 m and GL −10 m respectively during the period of testing. The test well was 1 m in diameter and 60 m in depth. To reduce the expense of good drilling while increasing the area of heat exchange between cooling water and strata, this study connected one well with two pipes in series in the closed system experiment (lengths: 40 and 60 m; Figure 2). The internal diameter of the 40-m pipe was 4 inches and that of the 60-m pipe was 5 inches; both pipes were made of stainless steel. The 100-m plumbing system can provide a total heat exchange surface area of 36.69 m². However, the effective heat exchange surface area (i.e., the surface area of the water-inlet section of a pipe) varied according to groundwater level changes at a particular location. The buried circulating pipes had a double pipe design, and the inner pipe was composed of PVC (external diameter: 2 inches). An outer-well pipe (PVC; internal diameter: 8 inches) was placed outside the two stainless steel pipes. The pipe has apertures from G.L.−41 to −45 m and from G.L. −52 to −60 m. This circulated groundwater inside and outside the 8-inch PVC pipe, thereby increasing the heat dissipation effect. Gravel stones were used to backfill the space between the 8-inch PVC pipe and the wall of the well. The open system retained the 8-inch PVC pipe outside the well and pumped groundwater from the upper gravel terrace to conduct heat exchange before discharging it back to the lower gravel terrace (Figure 3).

Heat exchange for the coolant source of two circulation types was conducted using the cooling water of the plate-type heat exchangers and air conditioners for heat exchange.

This experiment aimed to provide air conditioning to Taiwan Power Company’s power intake room in the basement of one building in Taiwan Tech. The area and height of the room were approximately 124 m² and 3.5 m, respectively. Several substation equipment sets were installed in the power intake room. Without having any ventilation or air conditioning, the estimated heat production rate within the entire space was 0.25 °C per hour. This quickly increased indoor temperature and affected the normal operation of substation equipment sets. One 15-RT (approximately 52.5 kW) package air conditioner was used to conduct testing, to determine the relationship between the energy consumption of compressors and the amount of groundwater circulation under the continuous operation of air conditioners and compressors. Moreover, the experimental results can be used to calibrate the heat exchange model obtained from the heat exchanger design program. Next to the power intake room, the system control room contained two pressure pumps, one for the air conditioner and the other for the closed circulating water system. Both pumps were used for driving the circulation of the cooling water. The backwater end of the closed circulating water system was equipped with one mechanical water meter to record the flow rate of the circulating water. Additionally, a plate-type heat exchanger was employed for heat exchange between the cooling water of the air conditioner and the circulating water of the geothermal energy pile. When the closed system was converted to an open system, a submersible pump was used to extract groundwater from the upper layer of the stratum, which was then used to conduct heat exchange with the circulating water of the air conditioner in the plate-type heat exchanger. Subsequently, the water was transferred back to the lower layer of the stratum.

4. Experimental Results

4.1. Closed Circulating Water System

Figure 4 shows changes in water temperature in a typical closed circulating water system when the water flowed through the heat exchanger. The curve can be separated into two parts, namely the temperature rising and temperature dropping curves. The early experimental period dealt with the temperature rising part. A significant amount of heat emitted by air conditioners immediately raised the temperature of the circulating water of the geothermal energy pile to approximately 54 °C; the high pressure of the refrigerant caused the system to shut down, which took around 4 h. Under such circumstances, the heat from the air conditioning equipment (including the compressor of the air conditioner and pump that drives water circulation) far exceeded the heat that can be discharged to strata (including soil and groundwater) from the circulating water pipeline of this experimental site. Consequently, the temperature of the circulating water increased. The air conditioner was shut down, but the circulating water pump had to continue operating to enable continuous heat exchange between the circulating water and strata. The heat produced by the circulating water pump was below the heat that the groundwater was capable of transferring away, resulting in a temperature decrease after system shutdown (Figure 4). At this time, the circulating water pump still continued to produce heat, and thus the curve of decreasing temperature did not return to the original level (4–25 °C).

4.2. Open Circulating Groundwater System

Figure 5 shows a typical example of the various changes in the temperature of open circulating groundwater of this system when it flowed into the heat exchanger. The figure shows that the outdoor temperature changed from 20 °C in the early morning to 30 °C when it reached its peak of the day, but this does not affect the various types of measured temperature of the system. Groundwater temperature was constantly maintained at approximately 23.5 °C to 24.5 °C following system operation. The air conditioner underwent full load operation (operating with both compressors) approximately 10 times within 24 h, and this lowered the temperature of the air conditioner’s outlet to 7–8 °C. During the remaining time, the air conditioner was operating at half load (one compressor) and this maintained the temperature of the air conditioner’s outlet at 12.5–16.5 °C. The aforementioned operating method constantly maintained the indoor temperature at approximately 17–18 °C. Figure 5 shows that the cooling water that flowed into the air conditioner from the plate-type heat exchanger was maintained at a temperature of 32–33 °C. The water flowing into the plate-type heat exchanger was maintained at 33–34 °C after absorbing heat from the air conditioner. Compared with the cooling towers, this system was not affected by the air temperature. Additionally, after the heat exchange of groundwater, the water temperature was maintained at 25.5–27.5 °C when it was transferred back to strata.

4.3. The Amount of Heat Produced by the Air Conditioner and That Absorbed by the Stratum

The air conditioner adopted in this study had a nominal capacity of 15 RT (approximately 52.5 kW) and cooling water flow of between 188 lpm (when the inflow and outflow water temperatures were 30 °C and 35 °C, respectively) and 196 lpm (when the inflow and outflow water temperatures were 32 °C and 37 °C, respectively). The air conditioner contained two scroll compressors, both of which had a nominal capacity of 24.03 kWh. Each compressor had an operating current of 11A and used refrigerant R22. An inflow water temperature higher than 40 °C would overload the compressor, make it unstable, and even cause damage to it. During the experiment, the compressor would trip as soon as the cooling water of the air conditioner reached a temperature of approximately 52 °C. According to information provided by the original manufacturer of this air conditioner, if the congelation temperature (the removal of heat from the refrigerant by cooling water) is 37.8 °C, the compressor can remove 31.1 kW of heat. If the congelation temperature increases to 48.9 °C, the compressor’s heat removal capacity is reduced to the equivalent of 28.0 kW. This demonstrates that the capacity of the compressor decreases as the cooling water temperature increases.

To estimate the heat exchange rate of a closed circulating water system in an experimental setting, this study conducted a log mean temperature difference (LMTD) measurement through the heat exchange of both sides (i.e., hot and cold) of a plate-type heat exchanger. This measurement was chosen because the temperature of water changes with time. By Equation (1) LMTD can calculate the heat transferred to the stratum from the air conditioner. From system startup to shut down (within 4 h), the average heat production of the air conditioner compressor was approximately equivalent to 48.7 kW and the heat absorbed by the stratum was 6.83 kW. In terms of the heat exchange in the open circulating groundwater system experiment, groundwater in strata could completely transfer the heat produced in this system over an entire day to strata. This is because the temperature differences of water did not increase over time.

This demonstrated that the heat exchange method between the strata and the system prohibited the use of groundwater as a coolant in an air conditioning system.

$$Q=UA\Delta T_m=UA\frac{\left(T_{h,0}-T_{c,i}\right)-\left(T_{h,i}-T_{c,o}\right)}{\ln \left(\frac{T_{h,0}-T_{c,i}}{T_{h,i}-T_{c,o}}\right)}$$

(1)

where $T_{h,i}$ is the temperature of high-temperature water when it enters the heat exchanger; $T_{h,0}$ is the temperature of high-temperature water when it exits the heat exchanger; $T_{c,i}$ is the temperature of the low-temperature water when it enters the heat exchanger; and T_c,o is the temperature of the low-temperature water when it exits the heat exchanger. $U$ represents the average total thermal conductivity and $\Delta T_m$ is the temperature difference between the two fluids.

5. Simulation of the Heat Dissipation Capacity of Groundwater

This study adopted a three-dimensional numerical simulation software program, Simulator for Heat and Mass Transport (SHEMAT), to construct a model for the aforementioned experiment and predict stratum reaction during the system’s subsequent operation. The software program simulated the thermal and mass flow of the groundwater aquifer, as well as connected fluid flow and the heat transfer model [20]. The three-dimensional heat transfer model of the United States Geological Survey [21] was integrated with the Processing Modflow interface [22] for producing the ‘Processing SHEMAT’, using Equation (2) as the governing equation.

$$\nabla \left(\lambda \nabla T-{\rho }_f{c}_f{T}_v\right)=\frac{\partial T}{\partial t}\left(\varnothing {\rho }_f{c}_f+\left(1-\varnothing \right){\rho }_m{c}_m\right)-H$$

(2)

where $\lambda$ is the thermal conductivity, $T$ is the temperature, $\varnothing$ is the porosity, ${\rho }_f$ is the groundwater density, $c_f$ is the groundwater specific heat, ${\rho }_m$ is the stratum density, $c_m$ is the stratum specific heat, and $H$ is the heat. The crystalline minerals in the Chingmei Formation were mainly quartzose sandstones. A parameter sensitivity analysis revealed that groundwater flow velocity had a much stronger effect on the heat transfer velocity of the system than the thermodynamic parameters of the materials. The relevant parameters in the model of this study included an effective porosity of 0.25, hydraulic conductivity of 0.0003 m²/s, thermal capacity of 1.875 MJ/m³·K, and thermal conductivity of 1.308 MJ/m³·K [23].

Figure 6 presents the simulated temperature boundaries of two different systems. Background velocity was simulated using a velocity of 0.005 m/s. The horizontal distribution of changes in stratum temperature suggested that all the heat of the closed circulating water system remained near the well and only a limited amount of heat was transferred to the stratum. In contrast, the open circulating water system allowed more heat to be transferred to the stratum and the heat to be brought downstream smoothly, allowing the system to operate continuously.

6. Evaluation and Planning of the System’s Energy Conservation Effectiveness

The aforementioned simulation and experimental results showed that under an open circulating model, the heat dissipation ability of strata was sufficient to absorb the heat produced by the air conditioner. Generally, every 1 °C drop in cooling water temperature can reduce the energy consumption of a compressor by 1–2%. Accordingly, using low-temperature groundwater can reduce the energy consumption of an air conditioner’s compressor. The relationship between the measured cooling water temperature ($T_{h,0}$) and compressor energy consumption in each experiment is presented in Equation (3). During the summer, the water supplied by cooling towers has a temperature of at least 33 °C. The cooling water temperature can be lowered to approximately 26–27 °C using a plate-type heat exchanger. This temperature decrease enabled a reduction in energy consumption by approximately 5.94–6.75% (0.49–0.56kWh), equivalent to monthly electricity savings of 352.8–403.2 kWh. Further discussions are provided on utilizing the low temperature (lower than air temperature) and stability of groundwater during summertime. Further energy conservation in the air conditioner compressor can be achieved by using a plate-type heat exchanger to effectively reduce the temperature ($T_{h,0}$) of the cooling water reentering the air conditioner.

$$P_{comp}=0.0024T_{h,0}^2-0.0582T_{h,0}+7.7337$$

(3)

Another source of energy consumption was the pump providing circulating groundwater. If the amount of groundwater in circulation (C_c) can be reduced appropriately without compromising the operation of the air conditioner, the energy consumption of the pump can be reduced. This experiment employed a frequency converter to control the pump’s rotation speed to reduce flow capacity. Figure 7 shows the relationship between the pump’s specific energy consumption, current frequency, and energy conservation rate. The specific energy consumption was determined by the ratio of the energy consumption under a decreased current frequency and that under a current frequency of 60 Hz (full load). Figure 7 shows that when the pump’s operation frequency was decreased from 60 Hz to 35 Hz, energy consumption could be reduced to 38.6% of that when operating with a full load. However, to protect the motor in actual practice, the operating frequency of the pump should not be lower than 40 Hz. For example, if the current frequency of the system was 40 Hz in this experiment, the energy conservation rate could reach up to 55.7%. Regarding the pump affinity laws in electrical engineering theories, specific operating frequency (N) and had a positive linear correlation with specific flow rate (Q) and a positive cubic correlation with specific energy consumption (P), as shown in Equation (4).

$$\left(\frac{P_1}{P_2}\right)\propto {\left(\frac{Q_1}{Q_2}\right)}^3$$

(4)

A reduction in the groundwater flow rate for circulation (C_c) enabled the pump to conserve energy but increased the groundwater’s temperature difference (T_c,o − T_c,i). In other words, when the inflow groundwater temperature (T_c,i) was fixed, the outflow groundwater temperature (T_c,o) increased. However, when the outflow groundwater temperature (T_c,o) was close to the inflow cooling water temperature (T_h,i), the heat exchange rate decreased. Therefore, increasing the inflow cooling water temperature (T_h,i) or cooling water temperature difference (T_h,I − T_h,o) increases the outflow groundwater temperature (T_c,o), that is, the groundwater temperature difference (T_c,o − T_c,i) increases. Appropriately decreasing the cooling water’s flow rate (C_h) can meet the aforementioned requirements and conserve energy from the cooling water pump. However, if the inflow cooling water temperature (T_h,i) is overly high, it may have a detrimental effect on components such as the air conditioner’s compressor. An overly high groundwater temperature (T_c,o) may result in the endothermic saturation of an aquifer and reduce its heat dissipation speed. However, the cooling water flow rate (C_h) and circulating water flow rate (C_c) would not reach below the level of inability to provide the hydraulic difference required for heat exchange.

When one 15-RT air conditioner operates at full load, the nameplate circulating cooling water flow rate is approximately 196 lpm (A 1-RT air conditioner requires 13 lpm). Therefore, this study set the upper limit of the circulating groundwater flow rate as 196 lpm (the actual flow rate is still affected by groundwater level) when the pump is set to operate at full load (60 Hz). To reduce the energy consumption of the pump, this study gradually reduced the flow rate and measured the relationship between the circulating water flow rate (C_c) and the air conditioner’s inflow cooling water temperature (T_h,o). The plate-type heat exchanger’s outflow cooling water temperature (T_h,o) and circulating water’s flow rate (C_c) mostly had a quadratic inverse relationship. Figure 8 shows that after the groundwater flow rate reached the air conditioner’s nameplate flow rate requirement (196 lpm), a further increase in the groundwater flow rate had only a limited influence on decreasing the cooling water temperature (T_h,o).

Therefore, by increasing the effective heat exchange area of the heat exchanger (e.g., increasing the number of plates), the aforementioned conditions can be satisfied to achieve the following goals: 1. Reducing the outflow cooling water temperature (T_h,o) to enable the energy conservation of the compressor; and 2. Reducing circulating groundwater flow rate (C_i) to enable energy conservation in the groundwater pump.

This study organized (1) the relationship between the air conditioner compressor’s energy consumption (P_comp) and cooling water temperature (T_h,o), (2) the relationship between cooling water temperature (T_h,o) and circulating groundwater flow rate (C_c), and (3) the relationship between groundwater flow rate (C_c) and the pump’s energy consumption (P_pump). Additionally, the cooling water temperature (T_h,o) and groundwater flow rate (C_c), both of which affect the energy consumption of an air conditioner compressor and pump, was incorporated into the illustration of a diagram for these four relationships (Figure 9). Four relationship curves in Figure 9 reveal that the required total energy consumption of the cooling water provided by air conditioner water temperatures of 32 °C and 30 °C was 9.83 kW and 10.01 kW, respectively. The aforementioned result demonstrates that decreasing water temperature increased the overall energy consumption.

This result shows that if the energy consumption of the pump is not considered, the proposed method could conserve energy in the air conditioner compressor by using groundwater with a temperature lower than the air temperature. Because the heat exchanger can only exchange heat between two fluids without mixing them, the reduction in the cooling water temperature is limited. If all efforts are placed into pump energy conservation, a decrease in circulating groundwater flow rate would occur, conversely increasing cooling water temperature, which is disadvantageous to energy conservation.

Calculations based on experiment results suggest that if a cooling tower relative to the elevation of the air conditioner is at an equivalent height to the lift head of the pump of the proposed method, the approach adopted by this study would conserve more energy than using a cooling tower. However, it is vital to determine the optimal condition for simultaneously conserving energy from the compressor and pump. Optimal conditions can be achieved by increasing the heat exchange area (the number of plates) of a plate-type heat exchanger.

7. Open System Power Consumption Optimization Calculation

According to the aforesaid test results, as the groundwater temperature is lower than the air temperature and stable, using the open circulating groundwater to cool down the air conditioning system can reduce the power consumption of the compressor. However, the heat exchanger only exchanges the heat energy of two fluids, without mixing them, so that the cooling effect of cooling water is limited. If only the power consumption of a water pump is reduced, the cooling water temperature would rise as the circulating groundwater discharge decreases, which is unfavorable to energy saving. A better solution is to determine the optimum condition for reducing the compressor power consumption and water pump power consumption concurrently, which could be achieved by enlarging the heat exchange area of the plate heat exchanger.

To reduce the cooling water quantity (C_h), the cooling water side temperature difference (ΔT_h) is increased (T_h,i increases and T_h,o increases). In the case of fixed heat exchange capacity, when the LMTD increases, the groundwater side temperature difference (ΔT_c) increases accordingly (T_c,i remains and T_c,o increases). When the groundwater side temperature difference (ΔT_c) is large, less groundwater discharge (C_c) is required, thus lowering the water pump power consumption. To achieve the aforesaid goal, the following settings were set:

• The overall system heat exchange only occurs in the air conditioner, plate heat exchanger, and stratum;
• The heat source only comes from the air conditioner, and the air conditioner heat output remains at 78.98 kW;
• The groundwater inlet temperature (T_c,i) remains at 23.6 °C;
• The maximum cooling water inlet temperature (T_h,i) is 37 °C;
• The maximum groundwater outlet temperature (T_c,o) is only 0.5 °C lower than the cooling water outlet temperature (36.5 °C);
• The circulating groundwater discharge (C_c) shall not be higher than 196 lpm (required rated cooling water quantity for a 15RT packaged air conditioning);
• The fluid pressure loss on both sides of the heat exchanger is lower than 50 kPa, the flow velocity is lower than 6 m/s.

The calculation results in

Table 1. The relationship between power consumption by the compressor and pump under different temperature of cooling water.

Fixed Flow of Cooling Water (A)				Variable Flow of Cooling Water (B)				Ratio of Power Saving
Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P(A) [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P(B) [kW]	100 − P(B)/P(A) [%]
36.3	1.46	8.79	10.25	33.0	1.81	8.43	10.24	2.2
33.3	2.07	8.46	10.53	32.0	1.94	8.33	10.27	3.7
31.7	2.39	8.30	10.69	31.0	2.12	8.24	10.36	5.2
30.4	2.91	8.19	11.10	30.0	2.39	8.15	10.54	5.5
29.6	3.74	8.12	11.86	29.0	2.73	8.07	10.80	5.7
28.0	6.67	7.99	14.66	28.0	3.28	7.99	11.27	23.1
27.0	10.51	7.92	18.43	27.0	4.27	7.92	12.18	33.9
26.0	22.61	7.85	30.46	26.0	6.61	7.85	14.46	52.5

show that under the requirement for reducing (T_h,o) to 27 °C, the overall power consumption can be reduced by 52.5% by lowering (C_h). As shown in Table 1, even if the (C_h) is changed, the required (C_c) is still 270 lpm under the requirement for reducing (T_h,o) to 27 °C. This result does not meet the constraint that the (C_c) shall not be higher than 196 lpm, so the heat exchange system shall be improved by increasing the number of plates. Under the optimum condition of the test heat exchanger, the four-phase relationship of compressor power consumption, cooling water temperature, water pump power consumption, and circulating groundwater flow is shown in Figure 10. This figure also represents the relationships among the compressor power consumption, water pump power consumption, and total power consumption of the system.

Under the same heat load requirement, enlarging the effective heat exchange area of the plate heat exchanger, meaning increasing the number of plates, can achieve the same goal at lower (C_c) under the same (T_h,o) requirement.

Table 2. The relationship between available (T_h,o) and (C_c) of different effective heat exchange areas.

Heat Exchange Areas [m2]	Th,i [°C]	Th,o [°C]	Ch [lpm]	Tc,i [°C]	Tc,o [°C]	Cc [lpm]
3.12	37.0	33.0	285.2	23.6	33.7	112.2
	37.0	32.0	228.1	23.6	33.0	120.6
	37.0	31.0	190.1	23.6	32.2	132.0
	37.0	30.0	162.9	23.6	31.2	149.4
	37.0	29.0	142.5	23.7	30.2	171.6
	37.0	28.0	126.6	23.6	29.1	206.4
	37.0	27.0	114.0	23.6	27.8	270.0
	37.0	26.0	103.6	23.6	26.3	420.0
4.56	37.0	33.0	285.2	23.6	34.6	103.2
	37.0	32.0	228.1	23.6	34.0	109.2
	37.0	31.0	190.1	23.6	33.2	117.6
	37.0	30.0	162.9	23.6	32.5	128.4
	37.0	29.0	142.5	23.7	31.4	146.4
	37.0	28.0	126.6	23.6	30.3	169.2
	37.0	27.0	114.0	23.6	28.9	213.0
	37.0	26.0	103.6	23.6	27.4	295.8
	37.0	25.0	94.9	23.6	25.6	571.2
9.84	37.0	33.0	285.2	23.6	35.9	92.4
	37.0	32.0	228.1	23.6	35.4	96.0
	37.0	31.0	190.1	23.6	34.9	100.8
	37.0	30.0	162.9	23.7	34.2	106.8
	37.0	29.0	142.5	23.6	33.4	115.8
	37.0	28.0	126.6	23.6	32.4	128.4
	37.0	27.0	114.0	23.6	31.3	148.2
	37.0	26.0	103.6	23.6	29.9	180.0
	37.0	25.0	94.9	23.6	28.0	257.4
16.8	37.0	33.0	285.2	23.6	36.5	88.2
	37.0	32.0	228.1	23.6	36.2	90.0
	37.0	31.0	190.1	23.7	35.8	93.0
	37.0	30.0	162.9	23.6	35.4	96.6
	37.0	29.0	142.5	23.6	34.8	101.4
	37.0	28.0	126.6	23.6	34.1	108.0
	37.0	27.0	114.0	23.6	33.1	118.8
	37.0	26.0	103.6	23.6	31.9	137.4
	37.0	25.0	94.9	23.6	30.0	177.0
21.84	37.0	32.0	228.1	23.7	36.4	88.2
	37.0	31.0	190.1	23.6	31.0	90.0
	37.0	30.0	162.9	23.6	35.8	93.0
	37.0	29.0	142.5	23.6	35.4	96.6
	37.0	28.0	126.6	23.6	34.7	102.0
	37.0	27.0	114.0	23.6	33.9	109.8
	37.0	26.0	103.6	23.6	32.8	123.6
	37.0	25.0	94.9	23.6	31.0	153.0
29.28	37.0	31.0	190.1	23.6	36.4	88.2
	37.0	30.0	162.9	23.6	36.2	90.0
	37.0	29.0	142.5	23.6	35.8	93.0
	37.0	28.0	126.6	23.6	35.3	97.2
	37.0	27.0	114.0	23.6	34.6	103.2
	37.0	26.0	103.6	23.6	33.6	114.0
	37.0	25.0	94.9	23.7	32.0	135.0
40.8	37.0	30.0	162.9	23.6	36.5	88.2
	37.0	29.0	142.5	23.6	36.2	90.0
	37.0	28.0	126.6	23.6	35.8	93.0
	37.0	27.0	114.0	23.6	35.3	97.2
	37.0	26.0	103.6	23.7	34.4	105.0
	37.0	25.0	94.9	23.6	33.0	120.6
60.48	37.0	28.0	126.6	23.6	36.3	89.4
	37.0	27.0	114.0	23.7	35.8	92.4
	37.0	26.0	103.6	23.6	35.2	97.8
	37.0	25.0	94.9	23.6	34.0	109.2
100.8	37.0	28.0	126.6	23.7	36.5	87.6
	37.0	27.0	114.0	23.6	36.3	88.8
	37.0	26.0	103.6	23.6	35.9	92.4
	37.0	25.0	94.9	23.6	35.0	99.6

shows the relationship between available (T_h,o) and (C_c) of different effective heat exchange areas. The relational graph is shown in Figure 11. The following phenomena can be observed in Table 2 and Figure 11:

• The heat exchange area of the original plate heat exchanger for the test is 3.12 m² (number of plates = 28). When the minimum (T_h,o) is 26 °C, the required circulating groundwater discharge decreases to 420 lpm, which is lower than the required circulating groundwater discharge = 1446 lpm when (T_h,o) is 26 °C in the simulated scenario (A) in Table 2;
• When the heat exchange area is larger than 4.56 m² (number of plates = 40), the available minimum (T_h,o) is 25 °C, but it cannot be lower even if the heat exchange area (number of plates) increases to 100.8 m² (850 plates);
• When the heat exchange area is larger than 21.84 m² (the number of plates = 184), the available maximum (T_h,o) decreases, and the available (T_h,o) is only 28 °C when the area is 100.8 m² (number of plates = 850);
• The aforesaid results show when the heat exchange area of the heat exchange model is larger than 21.84 m² (number of plates = 184), the difference among different heat exchange areas in the required (C_c) for the same available (T_h,o) decreases.

The power consumption of compressor (P_comp) and power consumption of water pump (P_pump) of the heat exchangers with different heat exchange areas in the calculation results are shown in

Table 3. The relationship between power consumption of compressor (P_comp) and water pump (P_pump) of the heat exchangers with different heat exchange areas.

Heat Exchange Areas [m2]
3.12				4.56				9.84
Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]
33.0	1.81	8.43	10.24	33.0	1.67	8.43	10.10	33.0	1.50	8.43	9.93
32.0	1.94	8.33	10.27	32.0	1.76	8.33	10.09	32.0	1.55	8.33	9.89
31.0	2.12	8.24	10.36	31.0	1.89	8.24	10.13	31.0	1.63	8.24	9.87
30.0	2.39	8.15	10.54	30.0	2.06	8.15	10.21	30.0	1.72	8.15	9.87
29.0	2.73	8.07	10.80	29.0	2.34	8.07	10.41	29.0	1.86	8.07	9.93
28.0	3.28	7.99	11.27	28.0	2.70	7.99	10.69	28.0	2.06	7.99	10.05
27.0	4.27	7.92	12.18	27.0	3.38	7.92	11.30	27.0	2.37	7.92	10.28
26.0	6.61	7.85	14.46	26.0	4.67	7.85	12.52	26.0	2.86	7.85	10.71
25.0	-	-	-	25.0	8.97	7.78	16.75	25.0	4.07	7.78	11.86
Heat Exchange Areas [m2]
16.8				21.84 (184)				29.28
Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]
33.0	1.43	8.43	9.86	33.0	-	-	-	33.0	-	-	-
32.0	1.46	8.33	9.79	32.0	1.43	8.33	9.77	32.0	-	-	-
31.0	1.51	8.24	9.75	31.0	1.46	8.24	9.70	31.0	1.43	8.24	9.67
30.0	1.56	8.15	9.72	30.0	1.51	8.15	9.66	30.0	1.46	8.15	9.61
29.0	1.64	8.07	9.71	29.0	1.56	8.07	9.63	29.0	1.51	8.07	9.58
28.0	1.74	7.99	9.73	28.0	1.65	7.99	9.64	28.0	1.57	7.99	9.56
27.0	1.91	7.92	9.83	27.0	1.77	7.92	9.69	27.0	1.67	7.92	9.58
26.0	2.20	7.85	10.05	26.0	1.99	7.85	9.83	26.0	1.84	7.85	9.68
25.0	2.82	7.78	10.60	25.0	2.44	7.78	10.23	25.0	2.16	7.78	9.95
Heat Exchange Areas [m2]
40.8				60.48				100.8
Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]	Th,o [°C]	Ppump [kW]	Pcomp [kW]	Total P [kW]
33.0	-	-	-	33.0	-	-	-	33.0	-	-	-
32.0	-	-	-	32.0	-	-	-	32.0	-	-	-
31.0	-	-	-	31.0	-	-	-	31.0	-	-	-
30.0	1.43	8.15	9.58	30.0	-	-	-	30.0	-	-	-
29.0	1.46	8.07	9.53	29.0	-	-	-	29.0	-	-	-
28.0	1.51	7.99	9.50	28.0	1.45	7.99	9.44	28.0	1.42	7.99	9.41
27.0	1.57	7.92	9.49	27.0	1.50	7.92	9.41	27.0	1.44	7.92	9.36
26.0	1.69	7.85	9.54	26.0	1.58	7.85	9.43	26.0	1.50	7.85	9.35
25.0	1.94	7.78	9.72	25.0	1.76	7.78	9.54	25.0	1.61	7.78	9.39

. The relationships between different effective heat exchange areas of PHE and the compressor power consumption, cooling water temperature, power consumption of water pump, and circulating groundwater flow are shown in Figure 12. It is observed that the required (C_c) for the same (T_h,o) decreases as the heat exchange area (the number of plates) of the heat exchanger increases so that the power consumption of the water pump (P_pump) decreases. The minimum total power consumption is shown in Table 3 in Bold and Italic. The power saved ratio can be calculated by each minimum total power consumption of the heat exchange area. The increment of power-saving ratio can be calculated by minus each power-saving ratio. The values of the calculations were listed in

Table 4. The relationship between power consumption of minimum total power consumption (compressor and water pump, P_comp + P_pump) of the heat exchangers with different heat exchange areas.

heat exchange areas [m2]	3.12	4.56	9.84	16.8	21.84	29.28	40.8	60.48	100.8
Min. total power consumption [kW]	10.24	10.09	9.87	9.71	9.63	9.56	9.49	9.41	9.35
Power saving ratio [%]	0	1.46	3.61	5.18	5.96	6.64	7.32	8.11	8.69
Increasement of power saving ratio [%]	-	1.46	2.15	1.57	0.78	0.68	0.68	0.79	0.58

and the relationship of minimum total power consumption and heat exchange area was plotted in Figure 13. It can be found that the maximum total power-saving ratio is up to 8.69% by increasing the heat exchange areas of PHE. The increment of power-saving ratio was flattened out after the heat exchange areas of PHE were larger than 21.84 m². Meanwhile, considering the limitation of space required for installing PHE; the heat exchange areas of PHE larger than 21.84 m² would be an optimum option to obtain the power-saving ratio of 5.69%.

8. Comparison with Cooling Tower Method

The volume ratio of the plate heat exchanger for this test to the actual cooling tower with the same refrigerating capacity is 1.36% (=56,672/4,171,510 cm²). Even if different models of heat exchangers are used, the plate heat exchanger + circulating groundwater well can save more than 98.1% (=1 – 56,672 × 1.34/4,171,510) of installation space compared with cooling tower, meaning about 8.7% (=1 − 9.35/10.24) of electricity can be saved after system optimization.

The power consumption of a compressor is related to the cooling water temperature, the compressor power consumption can be reduced by about 2% when the water temperature is reduced by 1 °C. In the summer, the air temperature often as high as 37 °C in Taipei Basin makes the temperature of the cooling water in the cooling tower for the compressor is close to this value. While the temperature of the circulating groundwater beneath the Taipei Basin is 26 °C, the power consumption of the compressor can be reduced by 22%.

Considering the aforesaid power saved after system optimization, the total power consumption can be reduced by about 28% (=22% + 5.96%) for a 15RT air-conditioning system using circulating groundwater.

9. Conclusions

Two experiments were conducted in this study to evaluate the effectiveness of using groundwater as a coolant for air conditioning systems in terms of reducing electrical power consumption. Specifically, groundwater from Taipei Basin was used as the coolant for an air conditioning system to obtain the air conditioning load, various water temperature changes, and the heat absorption of strata. The following conclusions can be drawn.

• Regarding a 15-RT air conditioner using a closed circulating cooling system, the heat exchange between the closed cooling water and the strata are slower than the rate of heat production of an air conditioner. Therefore, an air conditioner can only operate 4 h a day, and it needs to be shut down before starting operation again the next day. However, all-day air conditioning is possible after the simulation of stratum temperature distribution if several units of closed systems and their locations can be arranged appropriately to operate under different schedules.
• For a 15-RT air conditioner using an open circulating system, employing circulating groundwater with a temperature lower than that of the nameplate cooling water can discharge heat to strata and keep the air conditioner operating continuously for 24 h. This heat dissipation effect of strata can prevent a compressor from breaking down due to overheating. The results showed that every 1°C decrease in cooling water temperature enabled the air conditioner compressor to make energy savings of approximately 1% to 2%. This could achieve superior energy conservation than using cooling water provided by a cooling tower. No exhaust heat is discharged into the air by using the proposed circulating groundwater system. It can help mitigate the worsening heat-island effect in the Taipei Basin. The power consumption of the compressor can be reduced by 22%, considering the aforesaid energy saved after system optimization, the total power consumption can be reduced by about 28%.
• The volume ratio of the plate heat exchanger for this test to the actual cooling tower with the same refrigerating capacity is 1.36%. Even if different models of heat exchangers are used, the plate heat exchanger and circulating groundwater well can save more than 98.1% of installation space compared with the cooling tower, meaning about 8.7% of electricity can be saved after system optimization.
• To save the cost of drilling wells for the proposed method, the wells installed for dewatering to stabilize deep excavation are suggested to be used herein since they would be shut off after the engineering.