1. Introduction

In response to increasing energy consumption since the Industrial Revolution, the amount of industrial waste heat has been rising. Energy-cascading technologies have been developed to save on energy consumption. In particular, the development of advanced technologies to utilize low-temperature heat energy below 373 K would be beneficial in terms of reducing waste heat emissions. Recently, adsorption chillers have been receiving much attention. The adsorption chiller is one of the most advanced pieces of equipment that has the potential to collect or upgrade waste heat at low- temperatures. An adsorption chiller can store waste heat at temperatures less than 353 K and supply cooling energy at levels around 283 K [1] - [3] . Furthermore, the refrigerant used in adsorption chillers is water vapor, which is environment friendly. However, adsorption chillers have some issues that make them difficult for practical use.

The most significant problem is that the coefficient of performance (COP) per unit is considerably smaller than that of an absorption chiller. The COP of an adsorption chiller is less than 1.0. The maximum value is approximately 0.65 at 358 K for a regenerative heat source. Adsorption chillers have a realistic heat loss between the adsorption steps and desorption steps. This is because adsorption chiller was a batch type for a chiller cycle. A previous study of an adsorption chiller proposed a high adsorption capacity of adsorbent as well as designed an adsorber for a heat exchanger in a combined cycle. In this study, we pay attention to a double-effect adsorption chiller cycle.

Figure 1 shows a schematic diagram of the double-effect adsorption chiller cycle. Figure 2 shows a schematic diagram of the equilibrium line for a double-effect cycle. The double-effect adsorption chiller cycle comprises four adsorbers, A1, A2, B1, B2 (first adsorber A1, B1 and second adsorber A2, B2), an evaporator, a condenser, valves, and metallic tubes. In the adsorption step, the heat cooling output was obtained in the evaporator when H₂O vapor moves to the adsorbers A1, A2, and the vapor adsorber, FAM-Z01. In the desorption step, the FAM-Z01, in the first adsorber B1, desorbs at a high heat source temperature. The desorbed H₂O vapor from B1 moves and condenses

in the heat exchanger tube of the B2 adsorber. The FAM-Z01 on adsorber B2 is desorbed by condensation heat from B1 (first adsorber). Because the double-effect cycle obtained a heat cooling output from the first and second adsorber approaching the input desorption heat for the first adsorber, the cycle COP was expected to have much higher value than a single-effect cycle.

In general, the adsorption quantity was large because the desorption temperature was high in the adsorption chiller [4] . In a double-effect adsorption chiller cycle, it is desirable for the desorption temperature and pressure of the first adsorber to be high because the condensation heat of H₂O vapor from the first adsorber was utilized for heat desorption in the second adsorber. The heat supply for the second adsorber was H₂O vapor, not liquid.

In previous reports, the adsorbent for a double-effect adsorption chiller cycle has been silica gel. Recently, Marlinda published papers that refer to the feasibility of a double-effect adsorption chiller for numerical analysis. As a result, a double-effect cycle can be operated at the heat source temperature of 90˚C [5] [6] . In addition, The COP of a double effect cycle is more than twice that of a single-effect cycle at a heat source temperature of 130˚C. Ashida has determined that H₂O vapor desorbed from the first adsorber was condensed and the adsorbent for the second adsorber was heated and desorbed. The double-effect adsorption chiller cycle with silica gel functions is desired.

In this paper, adsorption isotherm and adsorption/desorption rate characteristics of a double-effect cycle with FAM-Z01 have been investigated. We have determined the behavior of the first and second adsorbers based on the assumption that a double effect cycle was operated. The volumetric method was employed to evaluate the influence of the desorption temperature and the adsorption bed thickness for the adsorption/ desorption rate. In the desorption step for the second adsorber, the amount of adsorption in the second adsorber, while H₂O vapor was supplied, was determined using an estimate of the pressure and temperature of the first adsorber. We evaluated that the characteristic of COP with a double-effect adsorption chiller cycle was based on the amount of adsorption from experimental results.

2. Experimental

2.1. Adsorbent

FAM-Z01 of zeolite Adsorbent made by Mitsubishi Plastics, Inc. was selected for this study. FAM-Z01 was developed for adsorption chillers. The adsorption isotherms of water on FAM-Z01 are available for low temperature exhaust heat at approximately 60˚C.

In this experiment, the sample diameter was under 5 μm. Before the experiment, an adsorbent sample was out gassed at 90˚C for 24 h inside vacuum.

2.2. Experimental Apparatus

This experiment for the evaluation of the amount of adsorbent and rate at which it adsorbs is measured using a static volumetric method. Figure 3 shows a schematic diagram of the experimental apparatus. The experimental apparatus consisted of an H₂O vapor chamber, a reaction cell made from stainless steel, four valves, and a pressure gage (ULVAC CCMT-100, 1000 D), heaters, two Pt resistance temperature detector, the evaporators 1 and 2, and a vacuum pump. The experimental equipment was set up in a thermostat air box. The temperature of the H₂O chamber as well as all the connecting pipelines was maintained at 60˚C by a thermostat-controlled air box. The inner volume of the H₂O chamber and reactor is 10,000 mL and 10 mL, respectively. The H₂O vapor pressure was changed, and the adsorption and desorption were measured by a pressure gage with a maximum uncertainty of 20 Pa.

2.3. Experimental Procedure

The main features of a different sample tested are recorded in Table 1. The samples were prepared with different masses with the aim to evaluate the effect of the thickness on the adsorption and desorption rates in the first and second adsorbers. In addition, the mass of the samples and the H₂O chamber volume were selected specifically for each experiment. The mass of each sample was adjusted for the pressure fluctuation of the H₂O chamber to maintain the adsorption value below 5%. The thickness of the sample was measured by micrometer (Mitutoyo, DIGIMATIC MICROMETER, 293-140).

The adsorption and desorption rates with a double-effect adsorption cycle were evaluated based on the thickness of the adsorbent packed bed, so the thickness of the samples in this experiment were adjusted to change the packed bed area on the reaction cell for the remaining samples of the same weight.

First, the temperature of the reaction cell was set at 120˚C. The reaction cell and the H₂O chamber were degassed for 12 h. After that, all the valves were closed. Second, the H₂O vapor evaporating from evaporator 1 was introduced to the H₂O chamber, and the pressure was increased to the required pressure for adsorption. The temperature of the reaction cell was set at the appropriate adsorption temperature. The valve between the reactor and H₂O chamber was opened, and the H₂O vapor was introduced to the reactor. The pressure of the H₂O chamber reduced along with the rate of adsorption. The pressure change was measured by a data-logger. The duration of time taken for the pressure of the H₂O chamber to change was equal to the time required for adsorption equilibrium. Then, the valve was closed. Subsequently, the desorption step was initiated. The temperature of the reaction cell was raised to the desorption temperature. The pressure of the H₂O chamber was either set to 20 kPa (first adsorber) or 4.2 kPa (second adsorber) with the use of a vacuum pump and evaporator 1. The pressure of the H₂O chamber was raised to correspond to the pressure required for desorption. The time required for the pressure of the H₂O chamber to change occurred at the same time the desorption equilibrium was reached. When the rate of the first adsorber desorption was evaluated, the desorption heat was supplied by a medium heat at a high desorption temperature. However, when the rate of the desorption of the second adsorber was evaluated, the desorption heat was supplied by the condensation of the H₂O vapor. The

pressure of the H₂O vapor provided by evaporator 2 was set at the first adsorber desorption pressure. When the desorption step was started, the H₂O vapor was introduced to the reaction chamber with a reaction cell.

The amount of adsorption and the rates of desorption in this experiment were as follow:

(1)

(2)

3. Experimental Results and Discussion

3.1. Adsorption Step

Figure 4 demonstrates the typical pressure variation of adsorption in different H₂O chamber pressures. The pressure of the H₂O chamber decreased with adsorption, but the value of the pressure change was very small compared to the initial H₂O chamber pressure. Figure 5 shows the amount of adsorption compared to the duration of time at the adsorption temperature 30˚C. From the results, the amount of adsorption and the rate depends on the H₂O pressure. Figure 6 shows the effect of the packed bed thickness for the adsorption rate at the H₂O chamber pressure 1.0 kPa. The thickness was measured for each sample. From the experiment, the duration of time required to reach the adsorption = 0.1 kg/kg was 46 s at 50 - 100 μm, 164 s at 120 - 150 μm, 772 s at 180 - 200 μm, respectively. The time for 180 - 200 μm was seventeen times longer than for 50 - 100 μm. The adsorbent packed bed thickness was related to the thermal and mass transfer resistance. This is why the adsorbent packed bed thickness influenced the adsorption rate.

3.2. Desorption Step for the First Adsorber

In this section, we evaluated the desorption rate for the first adsorber in a double-effect cycle. In this experiment, the adsorption step was conducted at 30˚C, 1.0 kPa. Next, the H₂O chamber pressure was set at 20 kPa, which was equal to saturated H₂O vapor pressure at 60˚C. Figure 7 shows the effect of the desorption temperature on the desorption rate. From the experimental results, the desorption rate and the desorption equilibrium linearly increases with the desorption temperature. When the desorption temperature was over 100˚C, the amount of desorption was 0.14 g/g at 100˚C or more. It was expected that the dependency of the amount of desorption at temperatures over 100˚C would be small, in addition, the input heat energy of a realistic heat for the first adsorber increased with temperature. Therefore, it was investigated that the desorption temperature for the first adsorber was 100˚C in this study.

Figure 8 shows the desorption rates at 100˚C under different packed bed thickness conditions. The temperature of the packed bed surface at this time was 50 - 100 μm. From the experimental results, the duration of time required for desorption equilibrium at 180 - 200 μm was 100 sec. The duration of time required for desorption was rapid for all measured thickness conditions. The desorption temperature change with the duration of time was only 0.3˚C. The packed bed temperature was kept constant at 100˚C. In the desorption step for the first adsorber, the H₂O vapor was rapidly desorbed from the adsorbent, and it was expected that the condensation heat energy was quickly supplied to the second adsorber.

3.3. Desorption Step for the Second Adsorber

After the adsorption step at 1.0 kPa and 30 kPa was completed, the valves were closed. The H₂O chamber pressure was set at 4.2 kPa. The H₂O vapor condensation heat was the desorption heat of the sample introduced to the reactor chamber. The H₂O chamber could be set at any pressure by changing the evaporator 2 temperature. Figure 9 illustrates the amount of desorption with time as a function of the H₂O pressure. From the measurement results in Figure 9, it can be seen that the amount of desorption and the rate was large because the H₂O pressure was high. Figure 10 shows the amount of desorption with time at 20 kPa H₂O vapor and liquid temperature of 60˚C, which was the saturated temperature at 20 kPa. From the measurement results in Figure 10, the

desorption rate was same for both the supplied heat form conditions. The packed bed temperature change was under 0.5˚C for the duration of time because the effect of the supplied heat form was very small.

Figure 11 shows the effect of the packed bed thickness on the desorption rate. From the experimental results, it can be seen that the desorption rate was large because the packed bed thickness was thin. In comparison with the first adsorber desorption step, it was revealed that the adsorbent packed bed thickness had an influence on the desorption rate. This is why the H₂O vapor diffusion resistance on the adsorbent packed bed was different for 20 kPa in the first adsorber desorption step and 4.2 kPa on the second adsorber desorption step.

4. Coefficient of Performance

The coefficient of performance was calculated for a double effect adsorption cycle using the experimental quantity of adsorption and desorption. It was expected that the coefficient of performance for an adsorption chiller was capable of influencing realistic heat values for an adsorber heat exchanger. In this study, this value was investigated only for the adsorbent and the H₂O refrigerant for realistic heat. The value referred to was from Mitsubishi Plastics, Inc [7] . Figure 12 illustrated the effect of the desorption temperature on the coefficient of performance. In addition, the coefficient of performance for a single-effect cycle was shown in Figure 12. The value of the coefficient of performance for double-effect cycle was 1.1 at 100˚C. The coefficient of performance on a single- effect cycle at 60˚C was 0.7. It was expected that the realistic heat of the adsorbent and the H₂O had an influence on the coefficient of performance for temperatures over 100˚C and amount of adsorbent influenced the coefficient of performance for temperatures under 100˚C. The maximum value for the coefficient of performance on a double-effect cycle was 1.57 times higher than the value of a single-effect cycle.

5. Conclusions

In this study, a double-effect adsorption chiller cycle with FAM-Z01 was evaluated. From the experimental study, we obtained the following results.

1) A double-effect adsorption chiller cycle can be operated at the desorption temperature of 100˚C. At this temperature, the quantity of adsorption was 0.14 g/g for the first and the second adsorbers.

2) The COP was calculated from the quantity of adsorption for the experiment and the result was 1.1 at the desorption temperature 100˚C.