**6. Empirical validation**

Following the predictions obtained from the calculation model, empirical validation would need to be conducted on the solar pond constructed. Thus, the model calculations reported in this study were simultaneously compared and validated against the temperature gradient data gathered from a salinity gradient solar pond constructed as aforementioned. The solar pond of interest was constructed with the dimensions shown in **Figure 3**.


#### **Table 4.**

*Calculated temperatures of the lower half of the middle (NCZ) layer of the salinity gradient solar pond, divided into 10 sub-layers.*

#### **Figure 3**

*The dimension of the side view (top) and front view (bottom) of the salinity gradient solar pond used for empirical validation of the developed calculation model. The dimensions detailed were used to produce the thermal predictions for the salinity gradient zones of the pond.*

loses heat to the layers immediately below and above dissimilarly. A solar pond operating in the geographical area of interest is predicted to reach a maximum temperature of approximately 79.5°C in the summer months with lows of 40°C in the winter. **Tables 2–4** show the diverse temperatures expected with the months of the year based on the energy balance models discussed in the previous section.

*Calculated temperatures of the upper half of the middle (NCZ) layer of the salinity gradient solar pond,*

**Month Ti1 Ti2 Ti3 Ti4 Ti5** January 21.13 23.74 25.96 27.95 29.82 February 22.31 25.95 29.12 32.00 34.73 March 24.98 29.09 32.68 35.94 39.03 April 29.43 33.91 37.78 41.30 44.64 May 34.44 39.38 43.64 47.50 51.17 June 36.93 42.19 46.74 50.87 54.79 July 38.26 43.26 47.59 51.54 55.30 August 38.26 42.96 47.04 50.76 54.29 September 36.05 40.42 44.24 47.73 51.04 October 32.35 36.29 39.75 42.93 45.95 November 27.93 31.38 34.44 37.27 39.96 December 23.78 26.99 29.86 32.51 35.03

*Calculated temperatures of the top (UCZ) and bottom (LCZ) layers.*

**Month Tatm TUCZ TLCZ** January 19.50 19.50 42.36 February 20.10 20.10 51.69 March 22.50 22.50 58.07 April 26.70 26.70 65.63 May 31.40 31.40 74.51 June 33.70 33.70 79.48 July 35.20 35.20 78.65 August 35.40 35.40 76.23 September 33.40 33.40 71.31 October 30.00 30.00 63.98 November 25.90 25.90 55.58 December 21.90 21.90 49.51

Following the predictions obtained from the calculation model, empirical validation would need to be conducted on the solar pond constructed. Thus, the model

**6. Empirical validation**

**Table 3.**

**50**

**Table 2.**

*Low-temperature Technologies*

*divided into 10 sub-layers.*

#### **Figure 4**

*A comparison of the major ion concentration gradient achieved with the constructed solar pond. The samples were obtained using an extraction pump at different depths. The surface data refers to a sample obtained at approximately 5–10 cm below the surface of the pond to allow for the pump to be completely immersed. The bittern sample refers to the ion concentration of the bittern used before the addition of MgCl2.*

The presence of a well-defined salinity gradient is paramount to the effectiveness of the solar pond's capability for heat storage. Thus, samples were taken at different levels of the pond to observe the concentrations at the expected upper,

*Comparison of the calculation model predictions vs. the empirical temperature data obtained over the initial 300 days. The NCZ data for the model are the mean values for Ti values reported in Tables 3 and 4. The*

*Solar Pond Driven Air Conditioning Using Seawater Bitterns and MgCl2 as the Desiccant Source*

Seawater bittern was the saline material used for the solar pond with additional MgCl2 dissolved in the saline material to ensure a more pronounced salinity gradi-

The increase in the salinity of the ponds towards the lower-convective zone can be seen in **Figure 4** and shows promise towards the solar pond's thermal storage potential. An interesting observation of the different ion concentrations were those of calcium and magnesium. The highest concentration cations in the upper and nonconvective zones were those of magnesium. It was the case until the deeper depths of the lower-convective zone at which the calcium and magnesium concentrations

To validate the thermal properties model, the temperatures of the upperconvective, non-convective and lower-convective zones (UCZ, NCZ and LCZ, respectively) were measured periodically as an empirical comparison of the initial 300 days of operation. The temperature measurements were taken at different depths along the centre of the pond; by using RTD with an accuracy of 0.1°C. **Figure 5** shows the temperature measurements obtained from the constructed solar pond while **Figure 6** shows a comparison of the model predictions vs. the actual

The temperature of the lower-convective zone was not expected to reach the heights predicted by the model as the winter month predictions begin with a solar pond with an LCZ already at temperatures of over 45°C stored from the previous summer months. However, the gradual increase in heat storage measured across the

ent was achieved. Future trials were envisioned only to use desalination reject brines as this would not only make use of an industrial waste product, without the conventional energy-intensive processes needed, but it would also

*depths of each of the 'real' were obtained at the depths reported in Figure 5 caption.*

improve the potential economic viability of the technology by using an

lower and non-convective zones.

*DOI: http://dx.doi.org/10.5772/intechopen.89632*

inexpensive resource.

**53**

**Figure 6**

were much more comparable.

solar pond temperatures measurements.

**Figure 5**

*Temperature log of the different zones of the solar pond over the first 300 days of operation. The UCZ, NCZ and LCZ sample were taken at depths of 10, 70 and 150 cm, respectively.*

*Solar Pond Driven Air Conditioning Using Seawater Bitterns and MgCl2 as the Desiccant Source DOI: http://dx.doi.org/10.5772/intechopen.89632*

**Figure 6**

**Figure 4**

*Low-temperature Technologies*

**Figure 5**

**52**

*A comparison of the major ion concentration gradient achieved with the constructed solar pond. The samples were obtained using an extraction pump at different depths. The surface data refers to a sample obtained at approximately 5–10 cm below the surface of the pond to allow for the pump to be completely immersed. The*

*Temperature log of the different zones of the solar pond over the first 300 days of operation. The UCZ, NCZ and*

*LCZ sample were taken at depths of 10, 70 and 150 cm, respectively.*

*bittern sample refers to the ion concentration of the bittern used before the addition of MgCl2.*

*Comparison of the calculation model predictions vs. the empirical temperature data obtained over the initial 300 days. The NCZ data for the model are the mean values for Ti values reported in Tables 3 and 4. The depths of each of the 'real' were obtained at the depths reported in Figure 5 caption.*

The presence of a well-defined salinity gradient is paramount to the effectiveness of the solar pond's capability for heat storage. Thus, samples were taken at different levels of the pond to observe the concentrations at the expected upper, lower and non-convective zones.

Seawater bittern was the saline material used for the solar pond with additional MgCl2 dissolved in the saline material to ensure a more pronounced salinity gradient was achieved. Future trials were envisioned only to use desalination reject brines as this would not only make use of an industrial waste product, without the conventional energy-intensive processes needed, but it would also improve the potential economic viability of the technology by using an inexpensive resource.

The increase in the salinity of the ponds towards the lower-convective zone can be seen in **Figure 4** and shows promise towards the solar pond's thermal storage potential. An interesting observation of the different ion concentrations were those of calcium and magnesium. The highest concentration cations in the upper and nonconvective zones were those of magnesium. It was the case until the deeper depths of the lower-convective zone at which the calcium and magnesium concentrations were much more comparable.

To validate the thermal properties model, the temperatures of the upperconvective, non-convective and lower-convective zones (UCZ, NCZ and LCZ, respectively) were measured periodically as an empirical comparison of the initial 300 days of operation. The temperature measurements were taken at different depths along the centre of the pond; by using RTD with an accuracy of 0.1°C. **Figure 5** shows the temperature measurements obtained from the constructed solar pond while **Figure 6** shows a comparison of the model predictions vs. the actual solar pond temperatures measurements.

The temperature of the lower-convective zone was not expected to reach the heights predicted by the model as the winter month predictions begin with a solar pond with an LCZ already at temperatures of over 45°C stored from the previous summer months. However, the gradual increase in heat storage measured across the early summer months is promising. Except for the lower-convective zone, the thermal behaviour across the initial months of the solar ponds was increased at a similar rate to the predicted data albeit at lower temperatures. The initial increase was hypothesised to be due to the initial heat of dissolution because of adding the salts to the solar pond.

*Aw i*ð Þ corresponding surface area of the wall at the given layer *AwU* surface area of side wall corresponding to the UCZ.

*Solar Pond Driven Air Conditioning Using Seawater Bitterns and MgCl2 as the Desiccant Source*

*Pu* vapour pressure of water at the upper layer surface temperature

*ext* heat extraction rate (load) from the LCZ by heat exchanger

*<sup>G</sup>* heat loss rates through the bottom wall (ground) of the pond

*Rw* total thermal resistance of the individual resistance of the thermal

*TGU* ground temperature of the surrounding soil at the layer in consider-

*Pw* partial pressure of the water vapour in the ambient air

*<sup>s</sup>* heat loss rates through the side wall of the pond

*Uc* convective heat loss rate from the pond's surface

*RH* average monthly relative humidity at location

*Tg* annual average earth temperature at the location

*Tg*ð Þ *z*,*t* average soil temperature at depth z (m) and time t (d). *Ta n*ð Þ ,*<sup>t</sup>* ambient air temperature at time t of *nth* day of the year

*Cw* thermal conductance of the composite wall

*kp* thermal conductivity of the insulation material *kc* thermal conductivity of the concrete wall

*hc* convective heat transfer coefficient

*CPa* humid heat capacity of air

*DOI: http://dx.doi.org/10.5772/intechopen.89632*

*I* solar radiation intensity

*QLN* heat lost from LCZ to NCZ *Q NU* heat gained from NCZ to UCZ

*Qsolar* solar energy absorbed by the layer *QU* total heat losses from the UCZ

*Ue* evaporative of heat loss rate

*Ur* heat loss rate due to radiation

*Us* heat loss through the side wall

insulation and concrete

*Sp* thickness of insulation material *T***<sup>1</sup>** temperature in the first layer *Tamb* ambient air temperature *T <sup>f</sup>* temperature in the last layer

*Rs* specific humidity

*RH* relative humidity *Sc* thickness of concrete

ation *Tsky* sky temperature. *z***<sup>1</sup>** depth of the UCZ **ɛ** emissivity of water

*θ<sup>i</sup>* angle of incidence *θ<sup>r</sup>* angle of refraction **Φ** latitude angle **ω** hour angle **δ** declination angle

*τ* fraction

**55**

**σ** Stefan Boltzmann's constant *V* average monthly wind speed **β** reflectivity of the radiation

**λ** latent heat of water evaporation

*h* local time

*N* day of the year *Patm* atmospheric pressure

*Q*\_

*Q*\_

*Q*\_

*Q*\_

*Q*\_

*Q*\_

*Q*\_

In a nutshell, though the temperature profile of the pond (through the months of operation) follows a similar pattern to the model predictions, the measured temperatures are lower than predicted. The discrepancy can be partly attributed to the fact that the weather data employed in the model is in variance with the actual weather condition experienced during the pond's operation. In addition, the effects of shading of the side walls and turbidity (clarity of the water) were not considered in the model, but in reality, contribute to the reduction of solar radiation received by the pond. The shading due to the side walls tends to reduce the effective surface area of the pond available to receive incident radiation thus consequently resulting in reduced temperature. The pond is prone to dust (due to the geographical location). Increased dust spread on the pond reduces the clarity of the water thereby reducing the penetration of solar radiation into the pond, consequently reducing the pond temperature.

#### **7. Conclusions**

Thermal behaviour analysis and a prediction model have been developed which can be effectively used for the construction and operation of a solar salinity pond. For the developed numerical was split into three sections: the UCZ, NCZ and LCZ. However, the NCZ consisted of a much larger depth than the other zones, thereby resulting in much more significant variance in its salinity and density. It was split into 10 sub-sections. By employing the numerical model in a set calculation procedure, the heat transfer coefficient could be first determined followed by the physical parameter of the pond saline material. Following this procedure, the temperature of each layer (and NCZ sub-layer) could be determined for any period in the year.

In this study, the average temperature of each layer was calculated for each month in the year when exposed to the Qatari temperatures. With the high heat climates of Doha, LCZ temperatures were predicted to reach its highest thermal storage potential at temperatures of about 70°C in June with the lowest in January reaching around 40°C.

The developed numerical model is planned for solar desiccant cooling applications in which the salinity gradient solar pond would be used for thermal storage as well as the storage and regeneration of the liquid desiccant used in the proposed air conditioning system.

### **Acknowledgements**

The author acknowledges Qatar National Research Funds (QNRF) for supporting this research through NPRP 7-332-2-138.

#### **Nomenclature**


*Solar Pond Driven Air Conditioning Using Seawater Bitterns and MgCl2 as the Desiccant Source DOI: http://dx.doi.org/10.5772/intechopen.89632*


**55**

early summer months is promising. Except for the lower-convective zone, the thermal behaviour across the initial months of the solar ponds was increased at a similar rate to the predicted data albeit at lower temperatures. The initial increase was hypothesised to be due to the initial heat of dissolution because of adding the

In a nutshell, though the temperature profile of the pond (through the months of operation) follows a similar pattern to the model predictions, the measured temperatures are lower than predicted. The discrepancy can be partly attributed to the fact that the weather data employed in the model is in variance with the actual weather condition experienced during the pond's operation. In addition, the effects of shading of the side walls and turbidity (clarity of the water) were not considered in the model, but in reality, contribute to the reduction of solar radiation received by the pond. The shading due to the side walls tends to reduce the effective surface area of the pond available to receive incident radiation thus consequently resulting in reduced temperature. The pond is prone to dust (due to the geographical location). Increased dust spread on the pond reduces the clarity of the water thereby reducing the penetration of solar radiation into the pond, consequently reducing the

Thermal behaviour analysis and a prediction model have been developed which can be effectively used for the construction and operation of a solar salinity pond. For the developed numerical was split into three sections: the UCZ, NCZ and LCZ. However, the NCZ consisted of a much larger depth than the other zones, thereby resulting in much more significant variance in its salinity and density. It was split into 10 sub-sections. By employing the numerical model in a set calculation procedure, the heat transfer coefficient could be first determined followed by the physi-

temperature of each layer (and NCZ sub-layer) could be determined for any period

In this study, the average temperature of each layer was calculated for each month in the year when exposed to the Qatari temperatures. With the high heat climates of Doha, LCZ temperatures were predicted to reach its highest thermal storage potential at temperatures of about 70°C in June with the lowest in January

The developed numerical model is planned for solar desiccant cooling applications in which the salinity gradient solar pond would be used for thermal storage as well as the storage and regeneration of the liquid desiccant used in the proposed air

The author acknowledges Qatar National Research Funds (QNRF) for

*Ae i*ð Þ effective surface area of the layer that receives the solar radiation

supporting this research through NPRP 7-332-2-138.

*Ae u*ð Þ effective area that receives the solar radiation

cal parameter of the pond saline material. Following this procedure, the

salts to the solar pond.

*Low-temperature Technologies*

pond temperature.

**7. Conclusions**

in the year.

reaching around 40°C.

conditioning system.

**Acknowledgements**

**Nomenclature**

**54**

*Low-temperature Technologies*
