The liquid desiccant system has been fully operational since April 2003, and its operation was monitored throughout the summer of 2003 (27 May through November 6).Numerous experimental runs were conducted prior to this period, to test various components and instruments and to make final adjustments in the control system. As it turned out, weather during the spring of 2003 in Haifa was relatively dry, and the desiccant system could not perform in a meaningful way until the end of May.

The complete set of data collected during the monitoring period contains 30955 lines, each representing a record of the system’s performance at a particular time. Normally, data was taken at one-minute intervals. Each record contains 28 instrument readings of temperatures, humidities, flowrates, pressures, etc. at various parts of the system. A computerized data acquisition system was used, showing the location and type of the various sensors and measuring points in the system. A log book of events was maintained during the monitoring period and all irregular events were listed. These include electric power interruptions, failed equipment etc.

For the purpose of the following discussion, three typical days were selected from the large volume of data — one for July, one for August and one for September. These are days during which the system operated rather smoothly, representing what may be expected ultimately after all the various operational and control problems have been fixed. Characteristic measured data representing the total monitoring period is as follows:

Heating water temperature Heating water flow rate Cooling water temperature Cooling water flow rate Air flow rate through desorber Air flow rate through absorber

Maximum solution concentration Minimum DPT reached

Air flow rate through desorber/absorber fans [m3/sec] Average air density through absorber/desorber Desorber fan power @ 32 Hz (40Hz)

Absorber fan power @ 20 Hz (40 Hz)

Air/Air Heat exchanger power

Solar collectors circulation pump power

Solution pumps power @ absorber/desorber sides

Hot water pump power

Cooling tower fan/pump power

Total parasitic power

Figure 2 describes the variation of three air humidities as functions of time for the same selected typical day (21 August 2003): outside air (Win), desorber outlet air (Woutl) and absorber outlet air (Wout2). Note that the absorber outlet air humidity is also that of the supply air to the conditioned space. As evident, the outside air humidity has remained approximately constant during the whole day, at about 16 g/kg, with a slight increase toward the evening. The absorber outlet humidity was equal to that of the outside air when the liquid desiccant system was put in operation at about 10:00, and was reduced to about 8 g/kg within 20 minutes. The machine was able to keep this humidity steady throughout the day. The desorber outlet air humidity, which is the control parameter, shows considerable variations. The control system shut the desorber off about 5 minutes after the start of operation, when the temperature of the hot water did not reach the set minimum limit; it turned the desorber back on and then turned it off when its outlet air humidity went below 30 g/kg; this is an indication that the solution in the desorber becomes too concentrated in LiCl, which may lead to crystallization. The same sequence continues several times during the day. The on-off cycling of the desorber makes it possible to maintain the supply air humidity at the desired and steady values.

In view of the ambiguity often encountered in the literature regarding the role of parasitic losses in the performance of desiccant systems (and other heat pumps and HVAC systems) we have introduced four types of coefficient of performance:

1) COP1 is a strictly thermal COP, not including parasitic losses.

2) COP2 includes in the denominator the sum of (solar) heat input and parasitic losses.

3) COP3 is the same as COP2, but the parasitic losses in the denominator are converted to their equivalent heat value, assuming power plant electricity generation at 40% efficiency.

4) COP4 is the ratio of latent heat removed from the process air to the electric power equivalent of the input energy, consisting of (solar) heat and parasitic.

Figure 3 describes the four COP’s as functions of time, for the data of 21 August 03, discussed previously. Clearly, COP4 shows the highest values and COP3 the lowest, based on the above definitions. Except for local fluctuations, all COP’s seem relatively

steady at periods of operation. The on-off operation of the desorber is clearly reflected in the chart.

A summary of the calculated data representing the total monitoring period is as follows: Average heat balance error (desorber side) 10%

Average heat balance error (absorber side) 20%

(where the Temperature uncertainty of the PT100 sensors is 0.2oC and the Relative Humidity uncertainty is 2% )

Heat transfer effectiveness:

Desorber solution/water heat exchanger Absorber solution/water heat exchanger Solution/solution heat exchanger Desorber side air/air heat exchanger Mass transfer coefficients:

0.3 kg/sec 0.9 kg/sec

81%

CONCLUSION

The objective of this project has been to construct a solar-driven liquid desiccant system for cooling, dehumidification and air conditioning — to test the concept, identify problems, carry out preliminary design optimization and measure performance. The prototype system, built at the Technion campus in Haifa, is designed to air-condition a group of offices on the top floor of the Energy Engineering Center. The design process involved initial measurements to determine unknown parameters, along with extensive performance simulations. With many unknown factors, the initial design of the system underwent several changes during the development period. The characteristic performance of individual components, analyzed theoretically in the simulation, was studied experimentally. Measurements have provided much-needed realistic data about heat and mass transfer coefficients. Important information was obtained about practical design aspects of the key components — dehumidifier and regenerator — as well as quantitative
data about their performance. The final prototype, including controls, has been fully operational since April 2003. The system functioned well, with 12 kW dehumidification capacity, and its performance was monitored throughout the summer of 2003 — from the end of May till the beginning of November. The data analysis indicates a thermal COP of about 0.8, with parasitic losses on the order of 10%.

The COP calculations performed on the monitoring data have yielded satisfactory results, particularly with regard to the thermal COP. By more elaborate design in the future, it is anticipated that parasitic losses could be minimized and better overall COP’s could be achieved.