Published in the proceedings of the 5th International Conference Solar AirConditioning organized by OTTI, 25-27 September 2013, Bad Krozingen, Germany

The Potential of Solar Adsorption Air-Conditioning in Saudi Arabia: A Simulation Study Ahmed Al-Mogbel1, Patrick Ruch2, Abdulrahman Al-Rihaili1, Saleh Al-Ajlan1, Paul Gantenbein3, Andreas Witzig4, Bruno Michel2 1 King Abdulaziz City for Science and Technology (KACST), Riyadh, Saudi Arabia Phone: +966 1 481 4323, Fax: +966 1 481 3880 E-Mail: [email protected] Internet: http://www.kacst.edu.sa/ 2 IBM Research – Zurich, Säumerstrasse 4, 8803 Rüschlikon, Switzerland 3 Institut für Solartechnik SPF, Oberseestrasse 10, CH-8640 Rapperswil, Switzerland 4 Vela Solaris AG, Stadthausstrasse 125, CH-7400 Winterthur, Switzerland

1. Introduction The electricity consumption for air-conditioning in Saudi Arabia which is dominated by hot desert climate exceeds 70% of the total electricity consumption during the summer months1–3. The installed generating capacity is required to double in the timeframe 2007-2017 in order to meet the growth in electricity demand of 4-8% per year4. Therefore, there is a strong demand to implement cooling solutions with minimal electricity consumption, in particular during peak hours in the summer season. While evaporative cooling is highly effective in arid climates, the scarcity of fresh water supplies in Saudi Arabia prohibits excessive use of water for cooling. As an alternative, solar air-conditioning provides a means to utilize the high solar irradiance5 in Saudi Arabia to mitigate electrical loads associated with peak cooling demand. Solid sorption (adsorption) cooling has the greatest potential benefit for electricity reduction due to the absence of internal moving parts, leading to a high electrical coefficient of performance (COPel). Also, compared to liquid sorption (absorption) cooling, adsorption chillers can be driven at relatively low temperatures, typically 60°C or above, and can therefore be combined with low-cost flat plate collectors. However, the low thermal coefficient of performance (COPth) of these systems leads to high recooling needs. Further, the performance in terms of COPth, COPel and cooling power is highly dependent on the temperature level for heat rejection, which in Saudi Arabia tends to be significantly higher than the nominal middle temperature of commercial adsorption chillers. Therefore, there is a demand for studies of the feasibility of implementing solar-driven adsorption chillers to domestic air-conditioning in Saudi Arabia.

256

In this contribution, the solar-driven adsorption cooling systems were studied with the Polysun® software package. The effects of location, recooler (evaporative vs. dry) and adsorption chiller type (silica gel vs. zeolite adsorbent) on COPth and COPel were investigated and compared to the performance of a compression cooling system.

2. Methodology Simulations were carried out using Polysun® 5.9 (Vela Solaris, Switzerland) using statistical climate data based on the Meteonorm library (Meteotest, Switzerland). The system diagram is shown in Figure 1.

Figure 1: Solar cooling system diagram implemented in Polysun®.

The efficiency of the flat plate collector was described by a second order polynomial

(x) = c0 – c1x – c2Gkx2 with c0 = 0.8, c1 = 3.3 W/(m2K) and c2 = 0.01 W/(m2K2) and the total irradiated solar energy Gk. The tilt angle was near horizontal for all locations. The net collector area was set to 150 m2 (net area 135 m2), equal to the footprint of the building. A hot water storage of 8 m3 was implemented. The two adsorption chillers simulated were commercially available silica gel–water and zeolite–water chillers6,7. For the simulations involving a dry recooler unit, an electrical COP of 11 was assumed. Simulations were also carried out using an evaporative cooling tower with a design approach temperature of 2°C and an electrical COP of 50. The modeled building was a single-floor house with a heat loss coefficient of 0.5 W/(m2K) and an internal heat gain of 8.6 W/m2 due to people, lighting and equipment. The cooling set point temperature was defined as 24°C and the minimum driving temperature was set to 55°C. A constant flow rate of 40 L/(m2·hour) was selected for the collector array while the remaining flow rates

257

were set according to the chiller requirements6,7. The performance of the solar cooling system was compared to that of a conventional grid-connected compression cooling scheme in which the nominal COPel of the compressor was 4.3. The four Saudi locations studied included Riyadh, Jeddah, Dammam and Abha. The thermal COP of the simulated adsorption chiller systems was calculated as the ratio of cooling yield Qth,cool to thermal energy Qth,drive delivered to the chiller from the hot water storage tank, COPth = Qth,cool/Qth,drive. The electrical COP was evaluated as the ratio of the cooling yield to the sum of electrical energy consumed by the auxiliary components including the fan coil, recooler, pumps and chiller, COPel = Qth,cool/(Qel,fan + Qel,recool + Qel,pump + Qel,chill).

3. Results and Discussion The simulated annual specific cooling load of the building in Riyadh was 194 kWh/m2, which compares well with the reported value of 199 kWh/m2 for an insulated house8. The highest cooling demand was found in the week of August 8 to 15. The ambient temperature regularly exceeded 40°C and the maximum global irradiance approached 1 kW/m2 during this time. The thermal yield of the collector array was maximal at noon at around 80 kW while the cooling yield peaked after sunset at around 8.2 kW to meet demand. The need for sufficient hot water storage volume to serve as a daily buffer is evident from this result. The cooling water inlet temperature to the adsorption chiller varied strongly during this timeframe, typically between 35°C and 45°C, which is expected to have a strong influence on the COP. The building temperature was maintained at 23-24°C throughout the entire simulation and the tank was never fully depleted. The power consumption for all simulated scenarios and locations was categorized according to the fan coil, recooler, pumps and chiller electricity consumption (Figure 2). The electricity consumption of the solar-driven adsorption cooling system was found to be lower than the compression cooling reference system in all cases. While the chiller itself accounted for 70-80% of the overall power consumption in the compression cooling scenario, the dominating component in the adsorption cooling scenarios was the recooler which accounted between ca. 70% (evaporative recooler) and 80% (dry recooler) of the total power consumption. The efficacy of evaporative cooling was highly dependent on location. For the silica gel–water adsorption chiller In Riyadh, the simulated yearly power consumption was 7991 kWh for the dry recooler system and only 3508 kWh for the evaporative recooler system. The difference was less pronounced in Jeddah which is exposed to higher humidity. Here, the simulated yearly power consumption was 7626 kWh (dry recooler) and 4679 kWh (evaporative recooler). As expected, the cooling scenario in

258

Dammam is similar to that of Riyadh in terms of cooling demand. Abha exhibited the lowest cooling demand and therefore power consumption due to its mild climate.

Figure 2: Simulated monthly component power consumption in four Saudi locations for the silica gel–water adsorption chiller and comparison with a grid-connected compression cooled system.

The monthly thermal and electrical COPs of the simulated adsorption cooling systems are plotted in Figure 3. The thermal COP of the simulated adsorption

259

cooling scenarios fell well below the nominal value of 0.6, which is mainly attributed to the elevated ambient temperature. In particular during the summer months, COPth dropped to as low as 0.30 in Riyadh for the silica gel–water chiller equipped with a dry recooler. The combination of dry recooler with the zeolite–water chiller resulted in a higher minimum COPth of 0.38 during the summer months due to the better performance of this system at higher heat rejection temperatures. For either system, the COPth tended to improve when combined with the evaporative recooler. For the silica gel–water chiller, the average COPth in Riyadh over the entire year was 0.35 for dry recooling compared to 0.43 for evaporative recooling. For the zeolite–water chiller, the yearly COPth was found to be 0.40 (dry recooling) and 0.47 (evaporative recooling), respectivey. The electrical COP was also highly dependent on the location and recooler type. In

Figure 3: Simulated electrical and thermal COPs for the silica gel–water and zeolite–water chillers employing dry (a, c) and evaporative (b, d) recooling for four Saudi locations in comparison with grid-connected vapor compression cooling.

Riyadh, the silica gel–water chiller exhibited a yearly COPel of 3.55 (dry recooling) and 8.36 (evaporative recooling), respectively, while the zeolite–water chiller was

260

characterized by yearly COPel values of 3.94 (dry recooling) and 8.84 (evaporative recooling). Compared to the vapor compression reference case with COPel values of 3.11 (Riyadh, dry recooling) and 3.91 (Riyadh, wet recooling), these results suggests that solar-driven cooling using current adsorption chiller technology and dry recooling offers only a marginal advantage over vapor compression cooling in Saudi Arabia in terms of electricity consumption. With evaporative recooling, however, a yearly reduction in electricity consumption of up to 58% was found.

4. Conclusions Simulations suggest that solar adsorption cooling can provide a means to considerably reduce the electricity needed for air-conditioning in Saudi Arabia. However, currently available adsorption chillers are not optimized for the high ambient temperatures encountered in many highly populated locations such as Riyadh. Therefore, the simulated performance of the chillers was significantly worse than under nominal operating conditions and evaporative recooling had to be employed in order to obtain a noteworthy advantage over vapor compression cooling. In order to minimize water consumption and improve the overall system efficiency, it is recommended to study the use of hybrid recoolers and novel adsorption systems adapted to high heat rejection temperatures.

5. References 1.

F. Al-Sulaiman & S.M. Zubair, A survey of energy consumption and failure patterns of residential air-conditioning units in eastern Saudi Arabia. Energy 5442, 967–975 (1996)

2.

All sunshine makes a desert... Saudi on the solar drive, Asian Solar (April/May 2010)

3.

A. Heuzeroth & K. Ishida, A closer look at solar assisted cooling, Climate Control Middle East (2011), available online at http://www.climatecontrolme.com/en/2011/06/closer-solar-assisted-cooling/

4.

Saudi Arabia Power Report Q1 2012. Business Monitor International (2012)

5.

W. Alnaser et al., First solar radiation atlas for the Arab world. Renewable Energy 29, 1085–1107 (2004).

6.

SolCool final report (in French), SFOE project #102095 (2011)

7.

Invensor GmbH, http://www.invensor.com (2013)

8.

S. Said & A. Al-Hammad, Energy conservation measures on residential buildings in Saudi Arabia. International Journal of Energy Research 17, 327–338 (1993)

261