Abstract:
As of the end of 1995 about 500 photovoltaic(PV) waterpumping stations have been installed in Thailand. Most units are ofdirect-coupling types utilizing AC motor-pumps and inverters.Insufficient water supply occurs during low solar radiation periodsand inverter breakdowns.This thesis adopted the PV water pumping design methodologydeveloped by L.Rosenblum to size a direct-coupled(DC a battery-connectedpumping systems constructed under this researchproject for techno-economic assessment. A unit,installed at KMITT,was designed to deliver water of 13 m3/day at 7 m head, and could beconfigured to operate in either a direct-coupled mode or a battery-connectedmode.The designed unit would consist of one 282 Wp(1.94 mZ) PV array,a 3.2 kWh battery unit,10 m of 40 mm 0 waterpipes.In actual installation the PV array wattage was increasedto 376 Wp(2.54 m2); eight M75 Solartron panels were used-Two locallymanufactured flood-type lead acid batteries, each at 12V 100 Ah and20 hr discharge rate,were adopted. This resulted in reducing thebattery size to 2.4 kWh. One 150307 DSU AY McDonald,at 7OOW,wasselected. Water pipe diameter was reduced to 35 mm.Research work consisted of testing of subsystems i.e.motor-pump and battery, instantaneous and daily performances ofSystems. On the prediction of long term performance and economicanalysis average daily radiation and those derived from theUtilizability were employed.is found that the overall efficiency of the directcoupling system depends on the radiation intensity at a low celltemperature, less than 40 'C, whereas that of the battery-connectedsystem the radiation intensity as well as the state of the batteryare determining factors. Being in a discharge state the systemefficiency exhibited more dependence on radiation intensity than ina charging state.The battery-connected system has higher instantaneousefficiency at low radiation and can operate over a wider radiationrange.Because of mismatches between power ratings of the PVarray and the motor-pumps, and its being used before, the overallsystem efficiency is low.Modelling results of instantaneous performance of thedirect coupling system were in good agreement with actualperformance as far as qualitative description and prediction ofwater flow rate are concerned. At high radiation intensity, modelledvalues of the array voltage was 10% higher,whereas the array currentwas 5% lower, than the actual values.It is not successful to model the battery-connected systemdue to, perhaps, inadequate understanding of battery characteristicsresulting in incomplete model of the battery. On computer modelling,results became inconsistent.The battery was set to oprate 7 hours each day.It is foundthat the daily pumping capacity of the battery-connected system wasbigger than that of the direct-coupled system on days having dailytotal radiation (on the PV plane) lower than 6 kWh/m2-day.On long term techno-economic analysis, we find that(i) the battery connected system, set to oprate 7 hourseach day, would provide more water than the direct coupled systemwhether the radiation derived from the utilizability methodology orthe average daily radiation was used. In the first instance it was11% higher,the second 12%.(ii) upon comparing the cost of water pumped by the batteryconnected system, the direct-coupled system undertaken by thisresearch work and the 6 other systems of KMITT, the unit cost ofwas 9-10 baht/m3 for the systems that deliver 6-12 m3 a day atstatic heads of 7-15 m.(iii) a sensitivity analysis indicates that the unit costof water is primarily dictated by the discount rate more than the PVwattage price.It is concluded that the battery-connected PV water pumpingsystem has better performance than the direct coupling system.Battery connected systems are feasible technically and economically.