Humidiﬁcation–Dehumidiﬁcation (HDH) Desalination System with Air-Cooling Condenser and Cellulose Evaporative Pad

: This paper presents a humidiﬁcation–dehumidiﬁcation (HDH) desalination system with an air-cooling condenser. Seawater in copper tubes is usually used in a condenser, but it has shown the drawbacks of pipe erosion, high cost of the copper material, etc. If air could be used as the cooling medium, it could not only avoid the above drawbacks but also allow much more ﬂexible structure design of condensers, although the challenge is whether the air-cooing condenser can provide as much cooling capability as water cooling condensers. There is no previous work that uses air as cooling medium in a condenser of a HDH desalination system to the best of our knowledge. In this paper we designed a unique air-cooling condenser that was composed of closely packed hollow polycarbonate (PC) boards. The structure was designed to create large surface area of 13.5 m 2 with the volume of only 0.1 m 3 . The 0.2 mm thin thickness of the material helped to reduce the thermal resistance between the warm humid air and cooling air. A fan was used to suck the ambient air in and out of the condenser as an open system to the environment. Results show that the air-cooling condenser could provide high cooling capability to produce fresh water e ﬃ ciently. Meanwhile, cellulous pad material was used in the humidiﬁer to enhance the evaporative process. A maximum productivity of 129 kg / day was achieved using the humidiﬁer with a 0.0525 m 3 cellulous pad with a water temperature of 49.5 ◦ C. The maximum gained output ratio (GOR) was 0.53, and the maximum coe ﬃ cient of performance (COP) was 20.7 for waste heat recovery. It was found that the system performance was compromised as the ambient temperature increased due to the increased temperature of cooling air; however, such an e ﬀ ect could be compensated by increasing the volume of the condenser.


Introduction
Fresh water is very important for human life regarding development in industry, agriculture and livelihood. The scarcity of fresh water in some countries and areas has driven the research and development in desalination in recent decades. Various types of technology have been developed for desalination, such as multi-stage flash (MSF), multi-effect desalination (MED), reverse osmosis (RO), and humidification and de-humidification (HDH) [1]. Some of these technologies have been commercialized at large scales; however, there are the drawbacks such as high capital cost and high operation cost [2]. HDH systems are advantageous in many ways, such as flexible scales, simple layout, low temperature operation, low capital cost and operation cost, and could be combined with renewable energy resources such as solar energy [3,4].
Many research groups have presented their design and results of HDH desalination systems in recent decades. The performance of these system depends on fluid properties and conditions, including water is much less than the surface area of the plastic fillings. Therefore, a cellulose pad is used as an alternative filling material in the humidifier. The cellulose pad is designed as a cellulose-bound cardboard structure and of a cross-fluted type. The capillary force results in the strong absorption of water into the cellulose pad, which makes the cellulose pad a water dispersing device once saturated. Such design greatly increases the contact area between air and water and can generate vapor very efficiently [17,18].
The rest of the paper is organized as follows.-Section 2 describes the experimental setup, Section 3 contains the thermodynamic analysis, Section 4 presents the results and discussion, and finally Section 5 is a short conclusion.

Experiment Setup
Our humidifier-dehumidifier (HDH) system was a closed-water-open-air (CWOA) system and the schematic diagram of the experiment setup is shown in Figure 1. The storage tank contained hot water that was heated by an electric heater. Hot water was pumped into a tank on top of the humidifier and a big chuck of sponge was placed at the exit of the tank to help evenly distribute hot water throughout the humidifier. The humidifier contained a cellulose pad with a large surface area in direct contact with the hot water, which generated vapor efficiently. The rejected water was collected at the bottom tank and cycled back to the storage tank by another pump.
bound cardboard structure and of a cross-fluted type. The capillary force results in the strong absorption of water into the cellulose pad, which makes the cellulose pad a water dispersing device once saturated. Such design greatly increases the contact area between air and water and can generate vapor very efficiently [17,18].
The rest of the paper is organized as follows.-Section 2 describes the experimental setup, Section 3 contains the thermodynamic analysis, Section 4 presents the results and discussion, and finally Section 5 is a short conclusion.

Experiment Setup
Our humidifier-dehumidifier (HDH) system was a closed-water-open-air (CWOA) system and the schematic diagram of the experiment setup is shown in Figure 1. The storage tank contained hot water that was heated by an electric heater. Hot water was pumped into a tank on top of the humidifier and a big chuck of sponge was placed at the exit of the tank to help evenly distribute hot water throughout the humidifier. The humidifier contained a cellulose pad with a large surface area in direct contact with the hot water, which generated vapor efficiently. The rejected water was collected at the bottom tank and cycled back to the storage tank by another pump.
The hot humid air was carried to the dehumidifier through an air duct under the drive of the blower located at the dehumidifier side. Cooling air from the ambient was driven by a fan located at the top of the dehumidifier and the flow direction was from the bottom to the top of the dehumidifier. The dehumidifier was composed of closely packed hollow polycarbonate (PC) boards that allowed humid air to flow between the PC boards while allowing the cooling air to flow inside the channels of hollow PC boards in a cross direction. The vapor in the humid air condensed on the walls of the PC boards and was collected by the container underneath the dehumidifier. After passing the dehumidifier, the cooled humid air flew out of the system as an exhaust.  The hot humid air was carried to the dehumidifier through an air duct under the drive of the blower located at the dehumidifier side. Cooling air from the ambient was driven by a fan located at the top of the dehumidifier and the flow direction was from the bottom to the top of the dehumidifier. The dehumidifier was composed of closely packed hollow polycarbonate (PC) boards that allowed humid air to flow between the PC boards while allowing the cooling air to flow inside the channels of hollow PC boards in a cross direction. The vapor in the humid air condensed on the walls of the PC boards and was collected by the container underneath the dehumidifier. After passing the dehumidifier, the cooled humid air flew out of the system as an exhaust.
A detailed description of the experimental setup is as follows. Three electric heaters with 2 kW each were used to heat the water in the storage tank. A control system was employed to keep the Water 2020, 12, 142 4 of 14 water temperature constant at the experiment conditions and kept the amount of water in the storage tank constant with feed-in water. The hot water temperature was limited to 50 • C, as the expected temperature of heated water by the photovoltaic thermal (PVT) collector was about in this range. The power of the two pumps that cycled the hot water were 50 W and 70 W, respectively.
The size of the cellulose pad in the humidifier, as shown in Figure 2a, was 50 cm in height, 35 cm in width and 30 cm in length, with a volume of 0.0525 m 3 . The size of the condenser made of hollow polycarbonate (PC) boards in the dehumidifier was about 50 cm in height, 40 cm in width and 50 cm in length, with a volume of 0.1 m 3 . The hollow PC boards contains channels with a cross section of 6 mm × 6 mm, as shown in Figure 2b. Twenty-eight pieces of PC board were assembled in parallel with an even spacing of 5 mm by fixture and epoxy, as shown in Figure 2c. The total surface contact area was about 13.5 m 2 . The system was designed such that the hot humid air flowed between the PC boards, while the cooling air flowed through the channels inside the PC boards in the cross direction, without direct contact between the two fluids.
A detailed description of the experimental setup is as follows. Three electric heaters with 2 kW each were used to heat the water in the storage tank. A control system was employed to keep the water temperature constant at the experiment conditions and kept the amount of water in the storage tank constant with feed-in water. The hot water temperature was limited to 50 °C, as the expected temperature of heated water by the photovoltaic thermal (PVT) collector was about in this range. The power of the two pumps that cycled the hot water were 50 W and 70 W, respectively.
The size of the cellulose pad in the humidifier, as shown in Figure 2a, was 50 cm in height, 35 cm in width and 30 cm in length, with a volume of 0.0525 m 3 . The size of the condenser made of hollow polycarbonate (PC) boards in the dehumidifier was about 50 cm in height, 40 cm in width and 50 cm in length, with a volume of 0.1 m 3 . The hollow PC boards contains channels with a cross section of 6 mm × 6 mm, as shown in Figure 2b. Twenty-eight pieces of PC board were assembled in parallel with an even spacing of 5 mm by fixture and epoxy, as shown in Figure 2c. The total surface contact area was about 13.5 m 2 . The system was designed such that the hot humid air flowed between the PC boards, while the cooling air flowed through the channels inside the PC boards in the cross direction, without direct contact between the two fluids.
The temperature and humidity were measured by a T-type thermocouple and psychrometer at multiple points across the section of the air duct and at the exit of the dehumidifier. The air velocity was measured by an anemometer. The flow rate of the water was measured by a measuring cup.
A photograph of the actual system was shown in Figure 3.  The temperature and humidity were measured by a T-type thermocouple and psychrometer at multiple points across the section of the air duct and at the exit of the dehumidifier. The air velocity was measured by an anemometer. The flow rate of the water was measured by a measuring cup.
A photograph of the actual system was shown in Figure 3.

Thermodynamic Analysis
Several parameters were used to evaluate the performance of our HDH desalination system. The applied principles of thermodynamics are mass and energy conservation and the entropy increase in each process. A diagram for thermodynamic analysis was shown in Figure 4.
The mass ratio (MR) is a ratio between the hot water mass rate and the air mass rate as follows: where m a is the mass rate of air and m w,i is the mass rate of water in the humidifier inlet.

Thermodynamic Analysis
Several parameters were used to evaluate the performance of our HDH desalination system. The applied principles of thermodynamics are mass and energy conservation and the entropy increase in each process. A diagram for thermodynamic analysis was shown in Figure 4. The mass ratio (MR) is a ratio between the hot water mass rate and the air mass rate as follows: where is the mass rate of air and , is the mass rate of water in the humidifier inlet. The gained output ratio (GOR) is commonly used to evaluate the performance of a HDH system. It is defined as the ratio of the latent heat of generated fresh water over the input heat (while the work input of pumps and fans is negligible) as follows where is the mass rate of produced water in dehumidification, ℎ is latent heat of water and is heat input of this system. is evaluated with the energy balance of the water storage tank as

Thermodynamic Analysis
Several parameters were used to evaluate the performance of our HDH desalination system. The applied principles of thermodynamics are mass and energy conservation and the entropy increase in each process. A diagram for thermodynamic analysis was shown in Figure 4. The mass ratio (MR) is a ratio between the hot water mass rate and the air mass rate as follows: where is the mass rate of air and , is the mass rate of water in the humidifier inlet. The gained output ratio (GOR) is commonly used to evaluate the performance of a HDH system. It is defined as the ratio of the latent heat of generated fresh water over the input heat (while the work input of pumps and fans is negligible) as follows where is the mass rate of produced water in dehumidification, ℎ is latent heat of water and is heat input of this system. is evaluated with the energy balance of the water storage tank as The gained output ratio (GOR) is commonly used to evaluate the performance of a HDH system. It is defined as the ratio of the latent heat of generated fresh water over the input heat Q in (while the work input of pumps and fans is negligible) as follows where m d is the mass rate of produced water in dehumidification, h f g is latent heat of water and Q in is heat input of this system. Q in is evaluated with the energy balance of the water storage tank as where m w,i and m w,o are the hot water mass flow rate at the inlet and outlet of humidifier, respectively, and h w,i and h w,o are the enthalpy of the hot water at the inlet and outlet of the humidifier, respectively, m e is the evaporation rate in humidifier which equals to the feed water rate at steady state condition, and h f ,0 is the enthalpy of the feed water at ambient temperature. The recovery ratio (RR) is a ratio of the mass rate of condensed water in the dehumidifier over the mass rate of evaporated water. RR is defined as where m e is the mass rate of evaporated water in the humidifier and calculated by mass conservation as where w a,1 is the absolute humidity at the inlet of the humidifier and w a,2 is the absolute humidity at the outlet of the humidifier. The coefficient of performance (COP) is a coefficient to obtain performance of a system, typically used in cooling systems such as heat pumps. It is defined as where W in is the total power consumption of this system, excluding Q in .
To evaluate the performance of our system, we defined the effectiveness of the HDH system. The commonly used effectiveness is diverse in the simultaneous heat and mass exchanger or heat exchanger. We considered that the performance of HDH greatly depends on the fluid property described as enthalpy, so the definition of effectiveness for HDH is as [14]: where ∆H is the actual enthalpy change of either water steam or air stream, and ∆H max is the maximum possible enthalpy change.
The above Equation is based on mass and energy conservation, so the effectiveness is the ratio of real enthalpy change over maximum possible enthalpy change, which could be achieved in either stream. We name the enthalpy of any stream(x) in ideal state h x,ideal . Thus, the humidifier effectiveness (ε h ) could be defined as the following [13]: The max symbol is to make sure that the definition is not against the second law of thermodynamics. Similarly, the dehumidifier effectiveness (ε d ) could be defined as where the temperature of h f is assumed to be T a,2 +T a, 3 2 and m d h f ideal is calculated when the air temperature at the dehumidifier outlet (T a,3,ideal ) is set to be the ambient temperature. Table 1 lists some selected previous work with similar configuration of HDH systems to compare with ours. The hot water temperature in our system is set at 40 to 50 • C to simulate hot water from a photovoltaic thermal (PVT) collector. Taking into account the size difference of the lab systems, a calculation was done by setting the feed water flow rate to be the same and modifying the productivity proportionally. At the temperature range of 40-50 • C results showed that the productivity of our system was the highest among the listed previous work in Table 1. The amount of fresh water produced per day, or productivity, is an important measure of a HDH desalination system. The correlation between productivity and the mass ratio of hot water and air (MR) is shown in Figure 5. Productivity, measured by the produced fresh water, increased as MR increased, which was attributed to the increasing amount of vapor carried by the air at higher water flow rates. As the water flow rate increased, the contact area between the hot water and the cellulose pad increased, since a larger volume of the cellulose pad was soaked. It is also shown in Figure 5 that a higher water temperature improved the productivity, as the evaporation rate increased at higher water temperatures. The increase in productivity was more significant at higher water temperatures, as indicated by the slope of the trend lines.

Results and Discussion
Water 2020, 12, x FOR PEER REVIEW 7 of 14 productivity proportionally. At the temperature range of 40-50 °C results showed that the productivity of our system was the highest among the listed previous work in Table 1.

Part 1. Effect of Mass Ratio
The amount of fresh water produced per day, or productivity, is an important measure of a HDH desalination system. The correlation between productivity and the mass ratio of hot water and air (MR) is shown in Figure 5. Productivity, measured by the produced fresh water, increased as MR increased, which was attributed to the increasing amount of vapor carried by the air at higher water flow rates. As the water flow rate increased, the contact area between the hot water and the cellulose pad increased, since a larger volume of the cellulose pad was soaked. It is also shown in Figure 5 that a higher water temperature improved the productivity, as the evaporation rate increased at higher water temperatures. The increase in productivity was more significant at higher water temperatures, as indicated by the slope of the trend lines. Table 2 lists the experiment data for some example runs.    The measured average temperature at the outlet of the humidifier T a,2 increased as MR increased, as shown in the left y-axis of Figure 6. The absolute humidity w a,2 in the air duct, as shown in the right y-axis of Figure 6, represented the amount of the vapor generated by the humidifier. It was expressed as vapor mass over air mass and was calculated from the measured temperature and relative humidity across the air duct at the outlet of the humidifier. The results in Figure 6 show a positive correlation between the generated vapor and the mass ratio, i.e., the higher the mass ratio, the higher the temperature and the absolute humidity at the exit of the humidifier.  The measured average temperature at the outlet of the humidifier Ta,2 increased as MR increased, as shown in the left y-axis of Figure 6. The absolute humidity wa,2 in the air duct, as shown in the right y-axis of Figure 6, represented the amount of the vapor generated by the humidifier. It was expressed as vapor mass over air mass and was calculated from the measured temperature and relative humidity across the air duct at the outlet of the humidifier. The results in Figure 6 show a positive correlation between the generated vapor and the mass ratio, i.e., the higher the mass ratio, the higher the temperature and the absolute humidity at the exit of the humidifier. Figure 6. Left y-axis shows the average air temperature in the air duct , as a function of MR; Right y-axis shows the absolute humidity in the air duct ( , ) as a function of MR. The data were calculated based on the measured temperature and relative humidity in the air duct and averaged across the section.
The gained output ratio (GOR), as an important measure of HDH system performance, varied with the mass ratio, as shown in Figure 7. A higher mass ratio resulted in a higher GOR. A higher water temperature also improved GOR. The results show that the recovery ratio (RR) was relatively constant as MR varied, as shown in Figure 8. In other words, the capability to transform the vapor into water in the condenser was constant. The same portion of the vapor could be extracted in the humidifier, in spite of the amount of vapor carried by the air. This was probably attributed to the Figure 6. Left y-axis shows the average air temperature in the air duct T a,2 as a function of MR; Right y-axis shows the absolute humidity in the air duct (w a,2 ) as a function of MR. The data were calculated based on the measured temperature and relative humidity in the air duct and averaged across the section.
The gained output ratio (GOR), as an important measure of HDH system performance, varied with the mass ratio, as shown in Figure 7. A higher mass ratio resulted in a higher GOR. A higher water temperature also improved GOR. The results show that the recovery ratio (RR) was relatively constant as MR varied, as shown in Figure 8. In other words, the capability to transform the vapor Water 2020, 12, 142 9 of 14 into water in the condenser was constant. The same portion of the vapor could be extracted in the humidifier, in spite of the amount of vapor carried by the air. This was probably attributed to the high cooling capability of the condenser, and it was evidenced by the constant dehumidifier effectiveness which will be discussed in a later section.
Water 2020, 12, x FOR PEER REVIEW 9 of 14 high cooling capability of the condenser, and it was evidenced by the constant dehumidifier effectiveness which will be discussed in a later section.  The humidifier effectiveness and dehumidifier effectiveness , as defined in Equations (8) and (9), varied with MR as shown in Figure 9. The humidifier effectiveness was positively correlated with MR. A higher mass rate of hot water resulted in a larger contact area, which raised the air temperature and the absolute humidity as the hot water mixed with the incoming air in the high cooling capability of the condenser, and it was evidenced by the constant dehumidifier effectiveness which will be discussed in a later section.  The humidifier effectiveness and dehumidifier effectiveness , as defined in Equations (8) and (9), varied with MR as shown in Figure 9. The humidifier effectiveness was positively correlated with MR. A higher mass rate of hot water resulted in a larger contact area, which raised the air temperature and the absolute humidity as the hot water mixed with the incoming air in the The humidifier effectiveness ε h and dehumidifier effectiveness ε d , as defined in Equations (8) and (9), varied with MR as shown in Figure 9. The humidifier effectiveness ε h was positively correlated with MR. A higher mass rate of hot water resulted in a larger contact area, which raised the air temperature and the absolute humidity as the hot water mixed with the incoming air in the humidifier. The dehumidifier effectiveness ε d did not show much change with MR. The reason was probably because the cooling capability of the condenser was sufficient to condense the vapor under all the experiment conditions. This may be attributed to the efficient heat exchange between the hot humid air and the cooling air in the unique design of the condenser.
Water 2020, 12, x FOR PEER REVIEW 10 of 14 humidifier. The dehumidifier effectiveness did not show much change with MR. The reason was probably because the cooling capability of the condenser was sufficient to condense the vapor under all the experiment conditions. This may be attributed to the efficient heat exchange between the hot humid air and the cooling air in the unique design of the condenser. One of the reasons we think that the air-cooling condenser could work is that the ambient air as the cooling medium has no energy cost and the system can take as much air as it needs to cool down the warm humid air, while a water-cooling condenser needs to match the flow rate of water in the condenser with that in the humidifier, and limits how fast the cooling water (or preheated water) could flow. In other words water-cooling condensers recycle the heat, but compromise on the cooling capability of the condenser. The air-cooling condenser worked more independently, and it could condense the water vapor efficiently with a large surface area.

Part 2. Effect of Power Consumption on COP
The coefficient of performance (COP) is an important measure to evaluate how much useful thermal energy we can obtain provided the amount of work required. In our case, it was calculated as the ratio of the latent heat of fresh water over the work of the whole system, including pumps, fans etc., but excluding for the application of waste heat recovery. The maximal COP of 20.7 has been achieved with our system. The correlation between the COP and the total power consumption (Pt) is shown in Figure 10. The power consumption of the humidifier fan (PH) and dehumidifier fan (PDH) varied to find the maximum COP, while the power consumption of the rest of the system (pumps) was kept constant. The humidifier fan was the blower that delivered the vapor-carrying air through the system and the dehumidifier fan was the fan that delivered the cooling air into the dehumidifier. The left half of Figure 10 shows that the COP was positively correlated to PDH when PH was kept constant. This was probably because the higher PDH resulted in a higher cooling capability of the dehumidifier and increased the pure water production rate significantly. The right half of Figure 10 shows that the COP was negatively correlative with PH when PDH was kept constant. The reason was probably that a larger PH did not increase the fresh water production much, but the total power consumption increased, resulting a lower COP. The data of some experimental runs are presented in Table 3. One of the reasons we think that the air-cooling condenser could work is that the ambient air as the cooling medium has no energy cost and the system can take as much air as it needs to cool down the warm humid air, while a water-cooling condenser needs to match the flow rate of water in the condenser with that in the humidifier, and limits how fast the cooling water (or preheated water) could flow. In other words water-cooling condensers recycle the heat, but compromise on the cooling capability of the condenser. The air-cooling condenser worked more independently, and it could condense the water vapor efficiently with a large surface area.

Part 2. Effect of Power Consumption on COP
The coefficient of performance (COP) is an important measure to evaluate how much useful thermal energy we can obtain provided the amount of work required. In our case, it was calculated as the ratio of the latent heat of fresh water over the work of the whole system, including pumps, fans etc., but excluding Q in for the application of waste heat recovery. The maximal COP of 20.7 has been achieved with our system. The correlation between the COP and the total power consumption (P t ) is shown in Figure 10. The power consumption of the humidifier fan (P H ) and dehumidifier fan (P DH ) varied to find the maximum COP, while the power consumption of the rest of the system (pumps) was kept constant. The humidifier fan was the blower that delivered the vapor-carrying air through the system and the dehumidifier fan was the fan that delivered the cooling air into the dehumidifier. The left half of Figure 10 shows that the COP was positively correlated to P DH when P H was kept constant. This was probably because the higher P DH resulted in a higher cooling capability of the dehumidifier and increased the pure water production rate significantly. The right half of Figure 10 shows that the COP was negatively correlative with P H when P DH was kept constant. The reason was probably that a larger P H did not increase the fresh water production much, but the total power consumption increased, resulting a lower COP. The data of some experimental runs are presented in Table 3. Figure 10. The coefficient of performance (COP) of the system versus the total power consumption of the system. The power consumption of the humidifier fan and de-humidifier fan is specified as a secondary axis. Pure water production is also shown as a reference. Another measure of the system performance is the performance ratio (PR), recently developed using an exergy approach by K. C. Ng et al. [22]. The calculated result of PR for our system is about 9.8, based on the Equation given in the article.

Part 3. Effect of Ambient Temperature
We did experiments at different ambient temperatures and found that the ambient temperature influenced the system performance in two main aspects. One was the carrying air temperature and the other was the cooling air temperature. The two effects influenced the system performance in opposite ways. The former one resulted in a larger amount of vapor generation in the humidifier and a higher temperature at the outlet of the humidifier as the ambient temperature increased. The latter one, on the other hand, resulted in a lower cooling capability of cooling air in the dehumidifier and hence lower productivity as the ambient temperature increased. The net effect of the two factors could be evaluated by the correlation of GOR and ambient temperature. As shown in Figure 11, GOR decreased as the ambient temperature increased, implying that the latter effect was more significant than the former one. Although the results suggest that it is more desirable to operate the system in cool weather or at night, the system performance could be compensated or improved by increasing the size of the dehumidifier, more specifically by adding more PC boards to further increase the surface area of the condenser. Figure 10. The coefficient of performance (COP) of the system versus the total power consumption of the system. The power consumption of the humidifier fan and de-humidifier fan is specified as a secondary axis. Pure water production is also shown as a reference. Another measure of the system performance is the performance ratio (PR), recently developed using an exergy approach by Ng, K.C., et al. [22]. The calculated result of PR for our system is about 9.8, based on the Equation given in the article.

Part 3. Effect of Ambient Temperature
We did experiments at different ambient temperatures and found that the ambient temperature influenced the system performance in two main aspects. One was the carrying air temperature and the other was the cooling air temperature. The two effects influenced the system performance in opposite ways. The former one resulted in a larger amount of vapor generation in the humidifier and a higher temperature at the outlet of the humidifier as the ambient temperature increased. The latter one, on the other hand, resulted in a lower cooling capability of cooling air in the dehumidifier and hence lower productivity as the ambient temperature increased. The net effect of the two factors could be evaluated by the correlation of GOR and ambient temperature. As shown in Figure 11, GOR decreased as the ambient temperature increased, implying that the latter effect was more significant than the former one. Although the results suggest that it is more desirable to operate the system in cool weather or at night, the system performance could be compensated or improved by increasing the size of the dehumidifier, more specifically by adding more PC boards to further increase the surface area of the condenser. Water 2020, 12, x FOR PEER REVIEW 12 of 14 Figure 11. Effect of ambient temperature T0 on GOR.

Part 4. Cost Comparison
The cost comparison includes two aspects-the material cost and manufacturing cost of the aircooling condenser and a water-cooling condenser with the same volume. The former used the polycarbonate boards which are readily available as greenhouse covers on the market. The manufacturing process was to cut the boards and stack them together with epoxy. The latter used copper material which is commonly seen in other HDH systems. The manufacturing process involves piping, welding, and screw mounting. The total cost of the air-cooling condenser was about 100 USD, which is about half of the cost of the water-cooling condenser.

Conclusions
We have presented a HDH desalination system with an air-cooling condenser and a cellulose evaporative pad. A maximum productivity of 129 kg/day, GOR of 0.53 and COP of 20.7 were achieved.
The main features of our system are (1) efficient air cooling through hollow polycarbonate boards (better than traditional water cooling); (2) the use of the cellulous pad as a humidifier (more efficient and more compact); (3) high productivity per unit volume of dehumidifier; (4) high COP for waste heat recovery; (5) lower cost of the condenser.
It was found that productivity, GOR and humidifier effectiveness increased as MR increased and water temperature increased, while dehumidifier effectiveness did not vary much with MR or water temperature. The system performance was compromised as the ambient temperature increased, due to the increased temperature of the cooling air; however, such an effect could be compensated by increasing the volume of the condenser, or specifically by adding more PC boards to further increase the surface area of the condenser.

Part 4. Cost Comparison
The cost comparison includes two aspects-the material cost and manufacturing cost of the air-cooling condenser and a water-cooling condenser with the same volume. The former used the polycarbonate boards which are readily available as greenhouse covers on the market. The manufacturing process was to cut the boards and stack them together with epoxy. The latter used copper material which is commonly seen in other HDH systems. The manufacturing process involves piping, welding, and screw mounting. The total cost of the air-cooling condenser was about 100 USD, which is about half of the cost of the water-cooling condenser.

Conclusions
We have presented a HDH desalination system with an air-cooling condenser and a cellulose evaporative pad. A maximum productivity of 129 kg/day, GOR of 0.53 and COP of 20.7 were achieved.
The main features of our system are (1) efficient air cooling through hollow polycarbonate boards (better than traditional water cooling); (2) the use of the cellulous pad as a humidifier (more efficient and more compact); (3) high productivity per unit volume of dehumidifier; (4) high COP for waste heat recovery; (5) lower cost of the condenser.
It was found that productivity, GOR and humidifier effectiveness increased as MR increased and water temperature increased, while dehumidifier effectiveness did not vary much with MR or water temperature. The system performance was compromised as the ambient temperature increased, due to the increased temperature of the cooling air; however, such an effect could be compensated by increasing the volume of the condenser, or specifically by adding more PC boards to further increase the surface area of the condenser.