# Use of condensate drain to pre-cool the inlet air to the condensers: a technique to improve the performance of split air-conditioning units.

In hot ambient conditions, split air-conditioning (AC) systems experience a drop in their coefficient of performance and an increase in power consumption due to the direct relationship between the condensation temperature of the unit and the prevailing ambient temperature. In this work, the enhancement in the COP and the reduction in power consumption for a split AC unit are evaluated when a direct water spray evaporative cooling system is used. The study focuses on minimizing the amount of water needed to cool the condenser by utilizing the wasted water from the condensate drain, synchronizing the water injection with ON-OFF compressor operation, and determining the optimal hourly periods for the injection system to achieve a reduction in energy consumption. The objectives of this work are threefold. The first objective is to conduct experiments inside an environmental chamber to determine the effectiveness of the evaporative cooling system at different ambient conditions and to derive correlations that predict the outlet air conditions and the water consumption. The second objective is to develop an integrated mathematical model to simulate the performance of the evaporatively cooled split AC system in a controlled space under certain operational conditions (using steady-state performance data supplied by the manufacturer and taking into account the transient effect during the ON-OFF compressor cycling) and to experimentally validate this model The third and final objective is to evaluate the model under "real environment conditions" for a case study where the thermal conditions and energy demand of a typical office space in Beirut during three months (June, August, and October) are investigated and where the "sufficiency" of drain water is the limiting factor. The simulation results have shown that the drain water would be sufficient in October only, resulting in 5.3% energy saving throughout the whole day. On the other hand, the synchronized splay of water is found to last for six operating hours in a June day and eight hours in August; this results' in a total daily reduction in the consumed energy of 5% in June and 4.5% in August.Introduction

Split air-conditioning (AC) units are widely used in residential and office buildings to provide thermal comfort and dehumidification during the summer months. Unfortunately, during high ambient air temperature conditions, the performance of split units is known to drop substantially, thus limiting the ability of these units to provide the required comfort. At high ambient temperatures, the condensation temperature and pressure increase, leading to an increase in the compressor power consumption due to the higher pressure ratio in the vapor compression cycle. To overcome this shortcoming, manufacturers rely on over-sizing these units, which, unfortunately, leads to further increases in energy consumption and demand. It should be noted that the electric energy producing company in Lebanon, EDL, cannot cope with the continuing rise in the energy demand and has recently enforced a daily three hour electric power cut throughout the city of Beirut.

A potential solution to reducing the high electric energy demand would be to improve the efficiency of AC units during operation under high outdoor temperature conditions. This can be accomplished by decreasing the condensation temperature of the units by employing shading from direct sun radiation during sunny and high ambient temperatures (ElSherbini and Maheshwari 2010; Akbari 2002) or by pre-cooling the ambient air before it enters the condenser (Yu and Chart 2010; Hajidavalloo 2007; Waly et al. 2005; Lee et al. 2008; Chen et al. 2008). Most research work has focused on evaporative cooling, as this appears to be more effective in lowering the condensation temperature when compared to the shading approach. There are two systems commonly used for evaporative cooling: the direct and indirect water-injection methods (Yu and Chen 2010; Hajidavalloo 2007).

In direct water-injection systems, water is sprayed into the incoming air stream by means of a pump and nozzles; this atomized water spray evaporates easily and causes the incoming air temperature to decrease prior to entering the condenser. On the other hand, indirect water-injection systems use a cooling mesh, or a pad media, where water is sprayed on top of these pads; when the air passes through the mesh/media, it cools down before reaching the condenser. It is claimed that direct water injection may result in coil corrosion and fouling deposits (Hajidavalloo 2007). It should be noted that corrosion and fouling deposits may arise as a result of the salt content of the sprayed water; however, if drain water is used, it is likely that these effects may be avoided due to the very low salinity of this water. Hajidavalloo and Eghtedari (2010) conducted experiments using an evaporative cooling media pad with a 1.5-ton (5.27-KW) Mitsubishi split AC unit. In order to estimate the effect of the media pad on the performance of the AC unit under different ambient conditions, the experimental tests are performed in two stages, before and after applying the evaporative cooling.

It is found that the decrease in power consumption ranges from 11.7% to 20.3%, coupled with an increase in the cooling capacity in the range of 6.1% to 13.3% (Hajidavalloo and Eghtedari 20t0). Hwang et al. (2001) performed a series of experiments on a 9-KW (2.55-ton) split heat pump system in environmental chambers that resembled both indoor and outdoor conditions. It is noted that the capacity of the evaporative condenser increases over that of air-cooled condensers by 1.8%-8.1%. Experimental analysis on the use of different types of cooling pads to reduce the temperature of the ambient air was carried out by Waly et al. (2005). The experimental results, performed using a 2.8-ton (9.84-KW) AC unit, have shown that the use of a cellulose pad is more effective than a plastic pad in improving the cooling capacity and decreasing the power consumption of the DX unit (Waly et al. 2005). Since the use of pads in indirect water-injection systems hinders the movement of air and causes a significant drop in the air temperature, Yu and Chan (2010) employed mist pre-cooling on an air-cooled chiller system to enhance the coefficient of performance (COP) and thus to achieve energy savings. It is estimated that the use of mist spraying, coupled with condensing temperature control, could result in an annual decrease of 20% in electrical energy consumption (Yu and Chan 2010).

The research on evaporatively cooled condensers employing pads or spray mist is not limited to experimentation. Other researchers have developed models to investigate the performance of these units under different operating and environmental conditions. Techarungpaisan et al. (2007) developed a steady-state simulation model to predict the behavior and performance of a split AC unit with an integrated water heater. The mathematical model included all the components of the cycle (evaporator, compressor, capillary tube, condenser, receiver, and water heater) using fundamental principles of heat transfer, thermodynamics, and fluid mechanics, in addition to manufacturer's data. Techarungpaisan et al. (2007) performed experimental work to validate the model, and it was found that the experimental and simulation results are in good agreement. Kapadia et al. (2009) studied the transient characteristics of a split AC unit using R-22 and R-410a refrigerants. The transient model for the 1.5-ton (5.27-KW) split unit, which considered all four components of the unit (rotary rolling compressor, capillary tube, condenser, and evaporator), has been validated with measured and published experimental data.

From the above review, it can be concluded that energy efficiency improvement arising from pre-cooling the condenser air is well documented and is supported by both modeling and experimental work. Evaporative cooling, whether employed directly or indirectly, is effective in lowering the ambient air temperature but requires considerable amounts of water. In arid regions of the Middle East, the need for water to pre-cool the ambient air may result in evaporative cooling becoming infeasible and may negate the advantages that would arise from using such a technology. To achieve the efficiencies quoted in the work of Hajidavalloo and Eghtedari (2010), 6-8 l/h (2.2 x [10.sup.-2]-2.93 x [10.sup.-2] gal [UK]/min) of water have been used for the 1.5-ton (5.27-KW) Mitsubishi split AC unit. It is postulated that evaporative cooling would become more feasible if, instead of using "fresh" water resources, the water produced from the evaporator is collected over the operation hours of one day, stored, and is then used for evaporative cooling at scheduled times during the following day. This research aims to limit water usage in the evaporative cooling system to the water produced from the condensate of the operating AC system. The collected water from the condensate drain will be used efficiently by determining the optimal operation hours at which energy consumption is effectively reduced and by synchronizing the spray system with the compressor ON-OFF cycling.

Mathematical modeling

In order to meet the second objective of this manuscript, a mathematical model that integrates the office space, split direct expansion (DX) AC unit, and evaporative cooling system models is needed to predict the energy savings that could result from the coupling of evaporative cooling with the operation of an AC system serving the cooling needs of an office space. The office space model is used to predict the indoor thermal and moisture conditions in response to the various indoor and outdoor load profiles. The split DX unit model is implemented in the space model to determine the energy consumed in response to extracting the sensible and latent loads from the space, and the evaporative cooling system is used to predict the effect of water spray on lowering the ambient air temperature and enhancing the operation of the DX unit.

[FIGURE 1 OMITTED]

Space model

The thermal response of the space is modeled as a simple lumped resistance capacitor element, as shown in Figure 1. Temperatures [T.sub.i], [T.sub.o], and [T.sub.m] are the inner and outer air temperatures and the uniform wall temperature, respectively. The inner and outer thermal resistances [r.sub.i] and [r.sub.o] are expressed in terms the overall element resistance [r.sub.m] and the ratio f, which is based on the work of Laret (1980), and is a function of the overall wall thermal resistance [r.sub.m] and capacitance Cm:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (l)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (2)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (3)

where [x.sub.i] is the thickness of wall layer i, [c.sub.i] is the material capacitance of wall layer i, h is the convective heat transfer coefficient, and [[rho].sub.i] is the wall layer density. The external convective heat transfer coefficient is based on the external sol-air temperature [T.sub.ext] to account for direct solar radiation for external walls. The space wall temperatures [T.sub.w] can be then computed from the wall energy equation:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (4)

Given the temperature of the walls and the internal load, the space air temperature and humidity can be calculated as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (5)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (6)

In Equation 5, the term on the left-hand side of the equation represents the thermal capacitance of the indoor air space. The three terms on the right-hand side of the equation represent the heat transfer exchange with the six surfaces (four walls, ceiling, and floor), the internal energy released, and the sensible load of the AC unit, respectively. Parameter [A.sub.j] represents the surface area; [T.sub.a] is the air dry-bulb temperature of the space, [[??].sub.sen] is the sensible load of the DX unit, and [Q.sub.int] is the internal sensible heat generation. In the moisture conservation equation (Equation 6), w is the moisture concentration, [[??].sub.gen] represents the mass flow rate of the humidity generated in the space, and [[??].sub.HVAC] represents the mass rate removed by the AC system resulting from the latent load.

AC system modeling

The DX AC system is designed to remove the total peak load at its rated capacity to control the indoor temperature at its set-point. While at partial load, the DX system relies on the ON-OFF cycling of the compressor to maintain a temperature close to the desired set temperature, while the relative humidity is left uncontrolled. The rated capacity of the AC system varies as the indoor and outdoor conditions deviate from the manufacturer's rated conditions. In addition, the power consumption of the DX unit, computed from the energy input ratio (EIR), depends on the indoor and outdoor conditions. A model that can simulate the ON-OFF operation of the DX unit in response to changes in the rated capacity of the system and indoor space load conditions is needed. Transient modeling of the DX system that considers the dynamic behavior of the different cycle components (compressor. condenser, expansion valve, and evaporator) and their thermal capacitance will, from a computing time perspective, be quite time consuming, especially when air pre-cooling and water spray optimization are coupled with the operation of the DX system. In this study, the well-established steady-state modeling of the DX system will be adopted with a first-order correction for the transient cooling capacity (O'Neal and Katipamula 1991; Mara et al. 2005; Garde et al. 1997).

To this end, the DOE's (1982) empirical "curve fit" relations in combination with "apparatus" dew point temperature will be used to determine the steady-state AC capacity and the EIR during OFF design operation. The steady-state empirical characteristic curve performance of the DX AC system at OFF design conditions is a bi-quadratic function expressed in terms of two parameters: the wet-bulb temperature entering the coil and the outside air dry-bulb temperature (DOE 1982):

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (7)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (8)

The coefficients in the bi-quadratic function can be computed from manufacturer data points at OFF design conditions (see Table 1 for coefficients used in this work).

However, knowledge of the cooling capacity during OFF design conditions is not sufficient to determine the indoor relative humidity. This will require the determination of the sensible heat ratio (SHR) over the range of operating conditions. This is accomplished by determining the apparatus dew point temperature in conjunction with the coil by pass factor (Henderson and Rengarajan 1996). Alternatively, the SHR may be calculated from the manufacturer data at different indoor dry- and wet-bulb temperatures. As for the transient cooling capacity of the DX system [Q.sub.t], it is modeled as a first-order system with a single time constant [tau]:

[Q.sub.t] = Q(I - [e.sup.-t/[tau]]). (9)

Water evaporation model

The effect of direct water injection on decreasing the ambient air temperature is studied experimentally, since it is difficult to model the mixing of ambient air and the water spray droplets, which can vary with different experimental setups and different implemented techniques of air-water mixing. To this end, a simple evaporative cooling system that can easily be integrated with existing AC systems is built and fitted at a distance from the condenser air inlet of a typical split AC system of a rated capacity of 1 ton (3.52 KW) with a condenser fan flow rate of 1000 [ft.sup.3]/min (0.472 [m.sup.3]/s), as shown in Figure 2. Two empirical equations (Equations 10 and 11), which relate both the air temperature and relative humidity of the air leaving the evaporative water spray system to the ambient air temperature [T.sub.db-in], and relative humidity R[H.sub.in], are developed. The empirical relations are developed by regression analysis and are unique to the simple evaporative cooling design that is integrated to the condenser fan capacity specified above:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (10)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (11)

[FIGURE 2 OMITTED]

Details of the experimental setup, regression analysis, and final equations for the water evaporation model are given in Appendix A.

Simulation model

To simulate the synchronized ON-OFF compressor operation with the water spray system of an AC unit serving a space that is subjected to a variable external and internal loads, the following information is required: the space envelope construction material, the external and internal load schedules, and the capacity of the DX system and its operation constraints (ON/OFF minimum periods and time constant). In addition, the evaporative regression equations for the ambient air temperature and relative humidity will be needed to predict the potential energy savings and water consumption of the evaporative cooling system (see Appendix A).

Starting from arbitrary initial conditions and using a time step of I s, the space internal humidity, wall temperatures, and air temperature are then computed using the first-order Euler-Forward scheme. In addition, the DX AC total capacity, SHR, and power consumption, as well as the amount of condensate drain, which are dependent on the indoor dry- and wet-bulb temperatures and the outside air temperature, are also computed. Marching the simulation through time for a "typical" summer day, the integrated model results at the end of the 24-h operational period are then used as initial input conditions in the cyclic simulations until steady periodic convergence is achieved. The criteria for convergence is reached when the maximum percentage errors for the air temperature, wall temperatures, and relative humidity are all less than [10.sup.-4]% when values at time t and t + 24 h are compared. The simulation model flowchart is shown in Figure 3. The simulations are performed to determine the daily energy consumption for two cases: regular operation of the AC system without evaporative cooling and regular operation of the AC system coupled with evaporative cooling.

[FIGURE 3 OMITTED]

Experimental methodology and results

To validate the energy savings that result from integrating evaporative cooling with the operation of the compressor of an AC system, experiments are conducted inside climatic chambers where the environmental conditions at both the evaporator and condenser sides can be controlled. The condenser, with the evaporative cooling system as described in Appendix A, for the commercially available split DX unit (Quiet-Side QSH09, rated capacity of 9000 BTU/hr [2.64 kW], rated EIR of 0.256, compressor type is rotary, and refrigerant used is R410A) is placed in one chamber, whereas the evaporator is placed in another chamber with inner dimensions of 2.5 m x 2.75 m x 2.8 m (8.2 ft x 9 ft x 9.2 ft) and a room wall thermal conductance of 1 [+ or -] 0.1 W/m K (0.578 Btu/h.ft.[degrees]F [+ or -] 0.057 Btu/h.ft.[degrees]F). The experimental setup is shown in Figure 4. The experimental AC unit is characterized by the following: compressor minimum ON and OFF periods of 3 min, time constant r equal to 1.2 min, evaporator fan operating on 360 CFM (0.17 [m.sup.3]/s), and a condenser fan having a capacity of 1000 CFM (0.472 [m.sup.3]/s). The polynomial coefficients of Equations 7 and 8 are given in Table 1. The condenser chamber is maintained at a temperature of 30[degrees]C (86[degrees]F) and relative humidity of 65%, and the thermal loads inside the environmental chamber are kept at 1.7 KW (0.48 ton) to allow the cycling of the compressor when the thermostat in the evaporator chamber is set at 23[degrees]C (73.4[degrees]F). The air temperature and relative humidity of the room air are measured by a temperature and relative humidity logger (OMEGA OM-EL-USB-2-LCD). Based on manufacturer's data sheets, the error for the temperature measurements is [+ or -] 0.5[degrees]C (32.9[degrees]F) and that of the relative humidity is [+ or -] 3% of the reading. In addition, the compressor power consumption is recorded using a power analyzer made by CHAUVIN ARNOUX (C.A 8334B QUALIS-TAR) with an accuracy of [+ or -] 1%. Before the start of the experiment, the tank of the water spray system is filled with 15 L (3.29 gal [UK]) of water, and then the AC unit is turned on after adjusting the experimental conditions inside the two climatic chambers. The AC unit is left to run for several hours to allow the chambers and AC unit to reach steady-state conditions prior to collecting data for one full hour. Two experiments are conducted: one experiment without water spraying (conventional) and another experiment where the water spraying is synchronized with the ON-OFF cycling of the compressor.

[FIGURE 4 OMITTED]

As can be seen in Figures 5 and 6, the experimentally recorded air temperature in the evaporator chamber and the compressor power consumption compare well with the results of the simulation runs. The model is able to capture the varying air temperature (23.25[degrees]C-24.75[degrees]C [73.85[degrees]F-76.55[degrees]F]) during the ON and OFF cycling of the compressor as well as the compressor power consumption but only after the startup electric surge of the compressor (the power surge for the compressor is not included in the modeling). The hourly experimental energy consumption (compressor + water pump) for the AC unit during the evaporative operation is found to be 3.5% less when compared to the energy (compressor) consumed during conventional operation. On the other hand, the energy reduction of the evaporative operation is over predicted by the model and is found to equal 5.1%.

Test case

A case study is performed to test the effect of the evaporative air cooling system on the energy performance of a DX AC unit providing the cooling needs for a "typical" office space in Beirut, Lebanon. The evaporative cooling system will rely solely on the condensate drain obtained from the operation of the DX unit to lower the ambient air temperature and enhance its efficiency. The office space dimensions are 5 m x 6 m x 3 m (16.4 ft x 19.7 ft x 9.8 ft) with no windows. The floor, ceiling, and three of the four walls are assumed to be internal partitions, adjacent to conditioned spaces. The remaining wall facing the south is considered an external wall. The typical construction of Lebanese external walls is 15 cm (5.9 in.) of hollow concrete blocks ([rho] = 977 kg/[m.sup.[up arrow]3](61 [lbm/ft3], cp = 840 J/[kg.K](0.201 ([Btu/lb.[F])), sandwiched between two layers of plaster 1 cm (0.39 in.) each ([rho] = 1380 kg/[m.sup.[up arrow]3] (86.2[lbm/ft3], cp = 825 J([kg.K]) (0.197[Btu/lb.[F]]). The internal load component is based on the occupancy schedule, where maximum occupancy is four persons during the occupied period (8:00 a.m.-8:00 p.m.), and the minimum occupancy is zero persons during the non-occupied period. Each of the office occupants is assumed to generate a sensible heat load of approximately 75 W (55.31 ft-[lb.sub.f]/s) and a latent load of 55 W (40.56 ft-[lb.sub.f]/s). The total office internal sensible load ranges from 1000 W (0.28 tons) to 1700 W (0.48 tons) during the occupied period to 200 W (5.68 x [10.sup.-2] tons) during the non-occupied period, as shown in Figure 7. The latent load source is mainly from people and from the fresh air requirement of 10 L/s (21.18 CFM) per person (ASHRAE 2007). The external load on the south wall is calculated from the weather data for Beirut, Lebanon, during the summer season (Ghaddar and Bsat 1998).

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

The AC unit studied is a residential split DX air-to-air heat pump (Carrier 38/40 MVC (Q) 012-1) with a rated cooling capacity of 3.43 kW (0.975 KW), rated flow rate of 425 CFM (0.201 [m.sup.3]/s), and rated EIR of 0.27. The unit's compressor is of a rotary type with a condenser fan flow rate of 1000 CFM (0.472 [m.sup.3]/s). The compressor's minimum ON and OFF time periods are assumed to be 3 and 3 min, respectively. The unit indoor thermostatic temperature is set at 24[degrees]C (75.2[degrees]F) with a temperature variation of [+ or -] 0.5[degrees]C (32.9[degrees]F) serving the office cooling needs during the occupied period (8:00 a.m.-8:00 p.m.), while "night purging" will be used to prevent the office concrete structure from storing heat during the OFF operation of the DX unit. The coefficients (a, b, c, d, e, and f) required for modeling the cooling capacity and EIR of the studied DX unit at OFF design conditions are summarized in Table 1.

[FIGURE 7 OMITTED]

Results of numerical simulations and discussions

The evaluation of the synchronized operation of the evaporative cooling system with the ON-OFF compressor operation of the DX unit will be based on the integrated model simulations results for three typical summer days: June 15, August 15, and October 15, representing the average, mild, and hot summer conditions in the city of Beirut. Simulations are performed for the two cases described earlier, the conventional and the evaporative cooling operations, to compare the energy savings and enhancements of the evaporative system arising from a drop in the power consumption of the DX unit. It should be noted that integrating an evaporative cooling system, which is synchronized with the ON and OFF cycle of the compressor during the entire operating period (8:00 a.m.-8:00 p.m.) will, in some cases, necessitate a quantity of water greater than that collected from the condensate drain. For such cases, and in order to limit the amount of evaporative water to that generated from the drain, the simulations are performed during selected optimized operational hours only.

Tables 2, 3, and 4 present the simulation results for the three representative summer days for the two cases of conventional and evaporative air-cooled condenser operations. Tables 2, 3, and 4 summarize the hourly results for the compressor cycling rate (rounded to the nearest integer), average COP, compressor power consumption, and the amount of condensate drain for the conventional operation. The same information is presented for the evaporative cooling case, with additional information on the amount of evaporated water required to lower the ambient air temperature. For the conventional operation, the results in the tables show the same trend for the compressor cycling rate and the cooling load. The cycling rate of the compressor starts high in the early hours of the DX unit operation (8:00 a.m.-10:00 a.m.); it then decreases with the increase in the internal load and external air temperature, and it increases again at the end of the daily operating period when the office cooling load is decreased. The average hourly COP is influenced by both the outside ambient temperature and the cycling rate of the compressor. A lower ambient temperature and a lower compressor cycling rate will have a net positive effect on the COP, while a higher ambient temperature and a higher cycling rate will have a negative effect on the COP. These two factors are the main reasons for the hourly variation of the COP and explain why, for example, for the month of June, the COP at 8:00 a.m. (3.88) has the same value of the COP at 12:00 p.m. The results also indicate that the compressor energy consumption is inversely related to the number of cycles and that it reaches its highest value for the month of August, 8.08 KW-hr (27,570 BTU), where it is larger by more than 21.5% and 29.6% when compared to the daily energy consumed in the months of June and October, respectively. Similarly, the amount of condensate water drain is the highest for the month of August (14.48 kg [31.92 lb]) when compared to the months of June (10.03 kg [22.11 lb]) and October (8.51 kg[18.76 lb]).

Tables 2, 3, and 4 show that coupling of the evaporative cooling system with the operation of the conventional split unit has improved the performance of the DX unit and that it has led to an increase in the amount of condensate and of the COP and a decrease in the power consumption. For the three day simulations, the maximum improvement in the COP is observed during the operating period (12:00 p.m.-5:00 p.m.) when the outdoor air temperature is expected to rise, thus justifying the use of the evaporative cooling system. The month of June, characterized by higher air temperatures and mild humidity, shows the highest percentage increase in COP of 8.15% at 4 p.m., while the humid month of August has a maximum increase in the COP of 4.7% occurring at 2:00 p.m., and the less humid month of October has its maximum increase in COP of 5.7% at 12:00 p.m. Most importantly, the results show that the daily energy consumption (compressor + 15 W [11 ft-[lb.sub.f]/s])water pump) of the evaporative cooling system is lower than the conventional system. The daily percentage decreases in the energy consumption are 7.1%, 6.1%, and 5.3% for the months of June, August, and October, respectively. It should be noted that the total amount of water required for the synchronized operation of the evaporative cooling pump with ON and OFF operation of the compressor during the daily operating period (8:00 a.m.-8:00 p.m.) exceeds the amount of collected condensate drain for the months of June and August; this is not the case for October, where the amount of drain water is sufficient to meet the evaporative cooling requirements. Figures 8a-8c present the hourly energy consumption for the two cases of conventional and evaporative cooling operations of the DX unit.

To limit the water used in the evaporative cooling system to the amount of daily collected condensate drain, the priority of each operating hour interval, in terms of the drop that occurs in the power consumption within each hour, is labeled with numbers ranging from 1 to 12 (see Figures 8a-8c). The operating hour labeled with number 1 is the hour that is most likely to be selected, and the hour labeled with number 12 is the one that is least likely to be selected. The operating hourly intervals, where the evaporative cooling is applied, are selected starting from the hourly period of highest priority labeled 1, and then the second priority hour of label 2 is selected until the total amount required for evaporative cooling nearly matches the amount of collected condensate drain. The spraying hours that match the condensate drain are from 1:00 p.m.-7:00 p.m. (13-19 in Figures 8a and 8b) and from 11:00 a.m.-7:00 p.m. (11-19 in Figures 8a and 8b) for the months of June and August, respectively, while for October, and as stated previously, the drain water is for operation throughout the day (see Figure 8c). To assess the energy savings that result from the water spray system, when operated intermittently and not continuously, Figures 8a and 8b, which are based on the operation of either the conventional or evaporative DX unit throughout the whole period (8:00 a.m.-8:00 p.m.), cannot be used since the office conditions will change when switching from conventional to evaporative operation. Thus, the integrated model is simulated for the month of June from 8:00 a.m. to 12:00 p.m. with no water spray, and the simulation is continued from 1:00 p.m. to 7:00 p.m. with the water spray. Similar simulations are performed for the month of August utilizing the water spray system from 11:00 a.m. to 7:00 p.m. The resulting daily percentage decrease in the energy consumption drops to 5% and 4.5% for the months of June and August, respectively.

[FIGURE 8 OMITTED]

The values reported above are comparable but, on the whole, lower than those reported elsewhere in the literature, where values ranging from 8. 1%-20.3% have been given (Hajidavalloo 2007; Waly et al. 2005; Hajidavalloo and Eghtedari 2010). It should be noted, however, that the system described herein does not rely on external sources of water but is "self-sufficient" in terms of reusing the water generated by the split AC unit.

Conclusions

Experimental work has been carried out to determine the effectiveness of an evaporative cooling system at different ambient conditions and to derive correlations that predict the outlet air conditions and water consumption. For the conditions under consideration, regression analysis of the experimental data has resulted in two correlations that match the experimental value to a very high degree of accuracy ([R.sup.2] value of greater than 0.95 and P-values [less than or equal to] 0). An integrated mathematical model has been developed to simulate the performance of the evaporatively cooled split AC system in a controlled space under certain operational conditions (using steady-state performance data supplied by the manufacturer and taking into account the transient effect of the compressor cycling and the manufacturer's minimum ON-OFF timing). The model has been experimentally validated, where the results indicate differences of less than 5.1%. The mathematical simulation model is used to evaluate potential energy savings under "real environment conditions" for a case study where the thermal conditions and energy demand of a typical office space in Beirut during three months (June, August, and October) are investigated and where the "sufficiency" of drain water is the limiting factor. The simulation results have shown that the drain water would be sufficient in October only, resulting in 5.3% energy saving throughout the whole day. On the other hand, the synchronized spray of water is found to last for six operating hours in a June day and eight hours in August; this results in a total daily reduction in the consumed energy of 5% in June and 4.5% in August.

Nomenclature A = area, [m.sup.2] ([ft.sup.2]) [c.sub.i] = wall layer thermal capacitance, J/kg[degrees]C (Btu/lb.[degrees]F) [C.sub.m] = overall wall thermal capacitance, J/kg[degrees]C (Btu/ib.[degrees]F) COP = coefficient of performance [C.sub.p] = air specific heat, J/kg[degrees]C (Btu/lb.[degrees]F) Cyc. = compressor cycles DX = direct expansion AC system EIR = energy input ratio f = fraction h = convective heat transfer coefficient, W/[m.sup.2].K (Btu/[ft.sup.2].[degrees]F) k = thermal conductivity, W/m.K (Btu/hr.ft[degrees]F) [??] = mass flow rate, kg/s ([lb.sub.m]/hr) Q = AC total capacity, kW (Btu/hr) Qlat = AC latent capacity, kW (Btu/hr) Qsens = AC sensible capacity, kW (Btu/hr) r = wall thermal resistance RH = relative humidity, % [r.sub.i] = wall element thermal resistance, K/W (hr.[degrees]F/Btu) [r.sub.m] = overall wall thermal resistance, K/W (hr.[degrees]F/Btu) t = time, s T = temperature, [degrees]C ([degrees]F) V = volume, [m.sup.3] ([ft.sup.3]) w = humidity ratio, kg/kg ([lb.sub.m]/[lb.sub.m]) W = power consumption, kW.hr (Btu.hr) x = thickness, m (R) Greek symbols [rho] = density, kg/[m.sup.3] ([lb.sub.m]/[ft.sup.3]) [tau] = time constant, s Subscripts a = air cyc = cycling cons = consumption db = dry bulb evap = evaporator gen = generated in, i = inside out, o = outside purg = purging w = wall wb = wet bulb

DOI: 10.1080/10789669.2012.619395

Appendix A

Experiment are conducted on an available commercial split AC unit (Quiet-Side QSH09), where the condenser fan (35 W [25.81 ft-[lb.sub.f]/s]) for this unit has a typical flow rate of 1000 CFM (0.472 [m.sup.3]/s). The condenser is placed in an environmental chamber equipped with a controller for both the air temperature and relative humidity. An evaporative cooling system is manufactured and fitted at the condenser inlet as is shown in Figure 2. The system frame is consistent with the shape and dimensions of the condenser. The system consists of a water tank in which water (drained and non-evaporated) is collected, a duct in which the spraying action takes place, connecting pipes, a small pump (TR11604, 15 W [11.06 ft-[lb.sub.f]/s]) that delivers water from the tank to the nozzles, and two bi-directional nozzles (JKL-JS2031W) that spray the water at the top of the duct. The duct is made from galvanized steel and is 0.8 m (31.49 in.) long, 0.5 m (19.68 in.) wide, and 0.3 m (11.31 in.) deep. The water tank is 0.8 m (31.49 in.) long, 0.2 m (7.87 in.) wide, and 0.15 m (5.95 in.) deep, having a total water capacity of 24 L (5.27 gal [UK]). The connecting pipes are 0.5 cm (19.68 in.) in diameter. The pump has a rated power of 15 watts (11.06 ft-[lb.sub.f]/s), pressure head of 2 m (78.74 in.), and a maximum constant mass flow rate of 0.0373 kg/s (4.69 x [10.sup.-6] [lb.sub.m]/hr). The nozzles spray water at an angle of 45[degrees] and against the airflow direction to achieve maximum evaporation rate. The nozzles are centered at the top side of the duct at an offset distance of 25 cm (9.84 in.) from the sides of the duct, maintaining 30 cm (11.81 in.) between the two nozzles. As the water hits the air entering the condenser, part of the spray evaporates into the air, and the excess water drains back to the tank. The temperature and relative humidity sensors (SENSIRION, EK-H2), with a maximum measuring error of [+ or -] 2% for relative humidity and [+ or -] 0.3[degrees]C (32.5[degrees]F) for temperature, are placed inside the duct before and after the water spraying action.

Experimental procedure and data analysis

Typical outdoor summer conditions in the city of Beirut in the range of 24[degrees]C-35[degrees]C (24, 26, 28, 29, 32, 33, and 35[degrees]C [75.2[degrees]F-95[degrees]F] for three relative humidity profiles--45%, 60%, and 75%) are replicated inside the environmental chamber to measure the air conditions before and after the water spraying action. The air inside the chamber is circulated using a stand fan to ensure a homogenous temperature distribution.

The experiment commences by running the first relative humidity profile (45%) in the environmental chamber. As the relative humidity inside the chamber reaches its steady-state value, the first outdoor temperature value (24[degrees]C [75.2[degrees]F]) is set inside the chamber. As soon as the chamber set temperature reaches a stabilized value (approximately hail an hour to reach steady-state condition), the water is sprayed at the minimum pump mass flow rate, at which point the steady-state values for the temperature and relative humidity, after the spray, are logged. The experiment continues by maintaining the same humidity profile and setting another value for the outdoor temperature from the above-specified range. The water is then sprayed, and the steady state values for the temperature and relative humidity after the spray are collected. This step is repeated until all seven outdoor temperatures are set and results obtained for the first relative humidity value (45%). The experiments are then carried out for a new relative humidity profile is run (60% and then 75%), and the whole experimental procedure is repeated. The above combination of test settings results in 21 experimental values for the air conditions at the condenser outlet. Completing the 21 experiments, the whole procedure is then repeated but with a decreasing order of the temperature starting from 35[degrees]C (95[degrees]F) and ending up with 24[degrees]C (75.2). The resulting raw data are then analyzed using the SPSS statistical software package to generate linear regression relations representing the drop in the dry-bulb temperature and the increase in the relative humidity of the air after the spray.

[FIGURE A1 OMITTED]

The generated correlation for the drop in the dry-bulb temperature is given by

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (A1)

whereas the correlation representing the increase in the air relative humidity given by

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (A2)

[FIGURE A2 OMITTED]

Figures A1 and A2 show the regression analysis equation versus the experimental values for both the output dry-bulb temperature and output relative humidity, respectively. The graphs show good agreement between experimental and regression analysis values.

In order to further validate these results, the water evaporation rate is determined from the measurement of air temperature and humidity, before and after the spraying action, and these values are compared with the change in the amount of water inside the tank. After running the water spray system for 3 h at ambient air conditions of 30[degrees]C (86[degrees]F) and relative humidity of 60%, the two values (water evaporation and difference in the tank water weight) are found to be less than 5%.

Acknowledgment

The financial support of the Lebanese ASHRAE Chapter is highly acknowledged.

Received June 13, 2011; accepted August 24, 2011

References

Akbari, H. 2002. Shade trees reduce building energy use and CO2 emissions from power plants. Environmental Pollution 116:119-26.

ASHRAE. 2007. ASHRAE Standard 62.2-2007, Ventilation and Acceptable Indoor Air Quality in Low-rise Residential Buildings. Atlanta, GA: American Society of Heating, Refrigeration, and Air Conditioning Engineers.

Chen, H., W. Lee, and F. Yik. 2008. Applying water cooled air conditioners in residential building in Hong Kong. Energy Conversion and Management 49:1416-23.

DOE. 1982. DOE-2 Engineers Manual, version 2.1A LBL-11353. Berkeley, CA: Lawrence Berkeley National Laboratory.

ElSherbini, A.I., and G.P. Maheshwari. 2010. Impact of shading air-cooled condensers on the efficiency of air-conditioning systems. Energy and Buildings 42:1948-51.

Garde, F., H. Boyer, F. Pignolet, F. Lucas, and J. Brau. 1997. Multiple model approach and experimental validation of a residential air to air heat pump. Proceedings of CLIMA 2000 Conference, August 30-September 2, Brussels.

Ghaddar, N., and A. Bsat. 1998. Energy conservation of residential buildings in Beirut. International Journal of Energy Research 32:523-46.

Hajidavalloo, E. 2007. Application of evaporative cooling on the condenser of window-air-conditioner. Applied Thermal Engineering 27:1937-43.

Hajidavalloo, E., and H. Eghtedari. 2010. Performance improvement of air-cooled refrigeration system by using evaporatively cooled air condenser. International Journal of Refrigeration 33:982-8.

Henderson, H.I., and K. Rengarajan. 1996. A model to predict the latent capacity of air conditioners and heat pumps at part-load conditions with constant fan operation. ASHRAE Transactions 102:266-74.

Hwang, Y., R. Radermacher, and W. Kopko. 2001. An experimental evaluation of a residential-sized evaporatively cooled condenser. International Journal of Refrigeration 24:238-49.

Kapadia, R.G., S. Jain, and R.S. Agarwal. 2009. Transient characteristics of split air-conditioning systems using R-22 and R-410A as refrigerants. HVAC&R Research 15:617-49.

Laret, L. 1980. Use of general models with a small number of parameters: Part 1: Theoretical analysis. Proceedings of 7th International Congress of Heating and Air Conditioning CLIMA 2000, September 17-19, Budapest.

Lee, W., H. Chert, and F. Yik. 2008. Modeling the performance characteristics of water-cooled air conditioners. Energy and Buildings 40:1456-65.

Mara, T., E. Fock, F. Garde, and F. Lucas. 2005. Development of a new model of single-speed air conditioners at part-load conditions for hourly simulations. ASME Transactions 127(2):294-301.

O'Neal, D.L., and S. Katipamula. 1991. Performance degradation during on-off cycling of single-speed air conditioners and heat pumps: model development and analysis. ASHRAE Transactions 97(2):316-23.

Techarungpaisan, P., S. Theerakulpisut, and S. Priprem. 2007. Modeling of a split type air conditioner with integrated water heater. Energy Conversion and Management 48:1222-37.

Waly, M., W. Chakroun, and N.K. Al-Mutawa. 2005. Effect of pre-cooling of inlet air to condensers of air-conditioning units. International Journal of Energy Research 29:781-94.

Yu, F.W., and K.T. Chan. 2010. Simulation and electricity savings estimation of air-cooled centrifugal chiller system with mist pre-cooling. Applied Energy 87:1198-206.

R. Sawan is Graduate Student. K. Ghali, PhD, Associate Member ASHRAE, is Associate Professor. M. Al-Hindi is Assistant Professor.

R. Sawan, K. Ghali, * and M. Al-Hindi Department of Mechanical Engineering, American University of Beirut, P.O. Box 11-0236, Beirut 1107-2020, Lebanon

* Corresponding author e-mail. ka04@aub.edu

Table 1. Coefficients of performance for the DX units. a b c Experiment DX capacity 1.816 -9.627E-02 3.196E-03 EIR -0.452 0.106 -3.007E-03 Test case DX capacity 1.717 -8.127E-02 2.626E-03 EIR -0.368 9.315E-02 -2.522E-03 d e f Experiment DX capacity 6.758E-04 -1.318E-04 -1.203E-05 EIR 6.402E-03 5.102E-04 -4.613E-04 Test case DX capacity 9.365E-04 -1.392E-04 -5.387E-06 EIR 5.564E-03 5.033E-04 -4.271E-04 Table 2. Simulation results for the month of June. Conventional Evapo- rative [W.sub.cons], Cycling Kw.hr Drain, Cycling Time rate COP (BTU * [10.sup.3]) kg (lb) rate 8 5 3.88 0.50 (1.71) 1.06 (2.33) 5 9 5 3.76 0.54 (1.84) 0.89 (1.96) 5 10 5 3.71 0.50 (1.71) 0.75 (1.65) 5 11 4 3.84 0.55 (1.87) 0.86 (1.89) 4 12 4 3.88 0.59 (2.01) 0.84 (1.85) 4 1 3 3.89 0.62 (2.11) 0.89 (1.96) 4 2 3 3.86 0.62 (2.11) 0.83 (1.83) 4 3 3 3.87 0.64 (2.18) 0.89 (1.96) 4 4 4 3.61 0.58 (1.98) 0.77 (1.69) 4 5 5 3.53 0.57 (1.94) 0.72 (1.58) 5 6 5 3.58 0.53 (1.81) 0.72 (1.58) 5 7 6 3.64 0.52 (1.77) 0.81 (1.78) 6 Total 6.76 (23.06) 10.03 (22.06) Evaporative Evaporated Wcons, water kW.hr Drain, required, Time COP (BTU * [10.sup.3]) kg (lb) kg (lb) 8 4.08 0.48 (1.64) 1.09 (2.39) 0.68 (23.2) 9 3.96 0.52 (1.77) 0.91 (2.01) 0.82 (1.80) 10 3.91 0.48 (1.64) 0.77 (1.69) 0.79 (1.74) 11 3.99 0.53 (1.81) 0.84 (1.85) 1.13 (2.49) 12 4.15 0.56 (1.91) 0.87 (1.91) 1.30 (2.86) 1 4.19 0.55 (1.87) 0.84 (1.85) 1.35 (2.97) 2 4.17 0.55 (1.87) 0.84 (1.85) 1.30 (2.86) 3 4.12 0.56 (1.91) 0.83 (1.83) 1.74 (3.83) 4 3.94 0.53 (1.81) 0.80 (1.76) 1.99 (4.38) 5 3.81 0.53 (1.81) 0.74 (1.63) 1.68 (3.69) 6 3.76 0.50 (1.71) 0.72 (1.58) 1.46 (3.21) 7 3.94 0.49 (1.67) 0.78 (1.72) 1.17 (2.57) Total 6.28 (21.48) 10.03 (22.06) 15.43 (33.94) Table 3. Simulation results for the month of August. Conventional Evapo- rative [W.sub.cons], Cycling kW.hr Drain, Cycling Time rate COP (BTU * [10.sup.3]) kg (1 b) rate 8 3 3.81 0.63 (2.15) 1.60 (3.52) 3 9 3 3.58 0.63 (2.15) 1.15 (2.53) 3 10 3 3.67 0.63 (2.15) 1.07 (2.35) 3 11 3 3.68 0.63 (2.15) 1.03 (2.26) 3 12 2 3.73 0.75 (2.56) 1.30 (2.86) 3 1 2 3.78 0.72 (2.45) 1.23 (2.71) 2 2 2 3.76 0.72 (2.45) 1.21 (2.66) 2 3 2 3.85 0.72 (2.45) 1.26 (2.77) 2 4 2 3.65 0.68 (2.32) 1.15 (2.53) 2 5 3 3.50 0.69 (2.35) 1.16 (2.55) 4 6 4 3.52 0.64 (2.18) 1.10 (2.42) 4 7 4 3.49 0.65 (2.22) 1.21 (2.66) 4 Total 8.08 (27.57) 14.48 (31.85) Evaporative Evaporated [W.sub.cons], water kW.hr Drain, required, Time COP (BTU * [10.sup.3]) kg (lb) kg (lb) 8 3.98 0.61 (2.08) 1.64 (3.61) 1.50 (3.30) 9 3.75 0.60 (2.04) 1.17 (2.57) 1.60 (3.52) 10 3.85 0.59 (2.01) 1.09 (2.39) 1.37 (3.01) 11 3.86 0.59 (2.01) 1.06 (2.33) 1.33 (2.93) 12 3.92 0.68 (2.32) 1.22 (2.68) 1.82 (4.01) 1 3.98 0.68 (2.32) 1.28 (2.82) 1.70 (3.74) 2 3.97 0.68 (2.32) 1.25 (2.75) 1.82 (4.01) 3 4.02 0.68 (2.32) 1.26 (2.77) 1.76 (3.87) 4 3.88 0.64 (2.18) 1.19 (2.62) 1.65 (3.63) 5 3.75 0.61 (2.08) 1.11 (2.44) 1.98 (4.36) 6 3.76 0.60 (2.04) 1.13 (2.49) 1.66 (3.65) 7 3.66 0.62 (2.11) 1.25 (2.75) 1.68 (3.70) Total 7.58 (25.86) 14.66 (32.25) 19.88 (43.74) Table 4. Simulation results for the month of October. Conventional Evaporative [W.sub.cons], Cycling kWhr Drain, Cycling Time rate COP (BTU * kg (lb) rate [10.sup.3]) 8 7 3.88 0.43 (1.47) 0.80 (1.76) 7 9 7 3.85 0.43 (1.47) 0.65 (1.43) 7 10 6 4.04 0.46 (1.57) 0.70 (1.54) 6 11 6 3.92 0.50 (1.71) 0.66 (1.45) 6 12 5 3.96 0.53 (1.81) 0.69 (1.52) 5 1 4 4.07 0.54 (1.84) 0.70 (1.54) 5 2 4 4.01 0.55 (1.87) 0.73 (1.61) 5 3 4 3.98 0.58 (1.97) 0.75 (1.65) 4 4 5 3.77 0.52 (1.77) 0.64 (1.41) 5 5 5 3.86 0.53 (1.81) 0.75 (1.65) 5 6 6 3.80 0.49 (1.67) 0.71 (1.56) 6 7 6 3.80 0.45 (1.53) 0.74 (1.63) 6 Total 6.01 (20.51) 8.51 (18.72) Evaporative Evaporated [W.sub.cons], water kWhr Drain, required, Time COP (BTU * kg (lb) kg (lb) [10.sup.3]) 8 4.03 0.42 (1.43) 0.81 (1.78) 0.29 (0.64) 9 4.01 0.42 (1.43) 0.66 (1.45) 0.35 (0.77) 10 4.14 0.45 (1.54) 0.64 (1.41) 0.35 (0.77) 11 4.17 0.48 (1.64) 0.75 (1.65) 0.60 (1.45) 12 4.20 0.50 (1.71) 0.70 (1.54) 0.89 (1.96) 1 4.21 0.50 (1.71) 0.69 (1.52) 0.93 (2.04) 2 4.23 0.50 (1.71) 0.68 (1.49) 0.93 (2.04) 3 4.28 0.52 (1.77) 0.77 (1.69) 1.30 (2.86) 4 4.08 0.48 (1.64) 0.67 (1.47) 1.00 (2.2) 5 4.08 0.51 (1.74) 0.76 (1.67) 0.68 (1.49) 6 4.01 0.47 (1.60) 0.72 (1.58) 0.70 (1.54) 7 3.94 0.44 (1.50) 0.68 (1.50) 0.59 (1.30) Total 5.69 (19.41) 8.52 (18.74) 8.63 (18.98)

Printer friendly Cite/link Email Feedback | |

Author: | Sawan, R.; Ghali, K.; Hindi, M. Al- |
---|---|

Publication: | HVAC & R Research |

Geographic Code: | 7LEBA |

Date: | Jun 1, 2012 |

Words: | 8721 |

Previous Article: | A statistical pattern analysis framework for rooftop unit diagnostics. |

Next Article: | Model-based robust temperature control for VAV air-conditioning system. |

Topics: |