Thermodynamic and economic considerations in solar desalination

ELSEVIER Desalination 129 (2000) 63-89

DESALINATION

www.elsevier.com/locatc/desal

Thermodynamic and economic considerations in solar desalination

Mattheus F.A. Goosen a*, Shyam S. Sablani b, Walid H. Shayya b, Charles Paton °, Hilal A1-Hinai"

aDepartment of Mechanical and Industrial Engineering, hDepartment of Bioresouree and Agricultural Engineering Sultan Qaboos University, Muscat, Sultanate of Oman

Tel. +968 515300; Fax +968 515416; email: [email protected] CLightworks Ltd., UK

Received 7 February 2000; accepted 16 February 2000

Abstract

The thermodynamic efficiency of single-basin and multiple-effect solar water desalination systems was critically reviewed with special emphasis on humidification~lehumidification processes. Solar energy may be used, either directly or indirectly, to produce fresh water. The concept of using the humidification-dehumidification process in combination with the growth of crops in a greenhouse system, however, is relatively new. System economics was also covered since it affects the final cost of produced water. While a system may be technically very efficient, it may not be economic. The challenges and opportunities of solar energy were also briefly discussed. The paper closes with a summary of key factors affecting system performance and recommendations for future areas of investigation and development.

Keywords: Desalination; Humidification~:lehumidification; Solar energy; Solar distillation; Solar stills

1. Introduction

With significant increases in population and industrialization over the past few decades, the world 's fresh water resources have come under renewed pressure. Arid regions, in particular, are already facing the reality that a shortage of

potable water can hinder economic development. Industrialized nations are also not immune to the problems caused by fresh water shortages. There is a growing realization in both arid and non-arid countries that the long-term solution lies in a coordinated approach involving water management, purification, and conservation [ ! ].

*Corresponding author.

0011-9164/00/$- See front matter © 2000 Elsevier Science B.V. All rights reserved PII: S001 I -9164(00)00052-7

64 A4F.A. Goosen et al. / Desalination 129 (2000) 63-89

Filling device

~f

Glass cover

I l a t e trough

~ II] ~%~N.......~.X~ Drain

Brine drain

(a) eur~t.

/71/ IP

ol ol

='-/ / I I / ,.,l I I

C ~ ,',.'~ -

/ /..~" ..~o°° \ .J'W~ ¢"

..-i" / \ ..N

/¢-/ .....-"i'

M.F.A. Goosen et al. /Desalination 129 (2000) 63-89 65

D Solar radiation

111 ,o,

Single t~sm

(b)

Double basin

E

Flat plate eoJlector

Drainage ~ Solar slill

F

1 1 Pipes with

Drainage for Drainage for excess water distilled water

Fig. 1. Solar desalination systems (adapted from Fath [5]). A. Single-effect basin still. B. Single-sloped still with passive condenser. C. Cooling of glass cover by (a) feed back flow, and (b) counter flow. D. Double-basin solar stills: (a) schematic of single and double-basin stills and (b) stationary double-basin still with flowing water over upper-basin. E. Directly heated still coupled with flat plate collector: (a) forced circulation and (b) natural circulation. F. Typical multi-effect multi-wick solar still.

66 M.F.A. Goosen et al. / Desalination 129 (2000) 63-89

A variety of techniques is currently being used to purify saline and brackish water. The most common are thermal methods such as multi-stage flash, and membrane techniques, as in the case of reverse osmosis. Alternative techniques such as solar desalination are also being considered. Solar methods are well-suited for the arid and sunny regions of the world as in the Arabian peninsula, the southwestern region of the US, and the southwest coast of South America [2,3].

Numerous solar desalination devices have been developed over the years. Examples include basin-type stills, soaked-cloth inclined type stills, and multiple-effect stills (Fig. 1). Single-effect stills can be classified into two main categories: passive and active (Fig. 2). Each category can be further sub-divided. Among these~ the passive basin-type solar stills are particularly suitable for developing countries due to their low cost and the fact that they require very little technically skilled maintenance. On the other hand, the overall efficiency (i.e., the actual amount of fresh water produced divided by the maximum amount of fresh water that can be produced based upon incident solar radiation multiplied by 100) of such stills is 30% or lower. Many studies

discussed ways to enhance the distillate output and maximize the efficiency of basin-type solar stills [4]. In such studies the enhancement of the distillate output was attempted either by increasing the saltwater temperature or by decreasing the temperature of the glass cover since the distillate yield is governed mainly by the difference in saturation vapor pressure at the water and glass surfaces.

Solar energy has its challenges. When considering the use of solar energy, it is necessary to know the energy falling on a unit area of the earth's surface per year. Seasonal variations due to atmospheric conditions exist along with variations due to geographic location. Solar energy is divided into two components: direct radiation which is received straight from the sun and diffuse radiation which has been scattered by clouds and dust arriving from all directions. The relative proportions of direct to diffuse radiation will depend on the site, season, and time of day. Diffuse radiation will be around 10% to 20% of the total radiation (the combined diffuse and direct) on a clear day, but will amount to 10% of a much smaller total radiation on a cloudy day [2].

Basin

Passive Stills

Diffusion Others

Single Effect Solar Stills

Active Stills

Wick Integrated With Solar Collecting Systems

Solar Solar Heaters Concentrators

Fig. 2. Classification of single-effect solar distillation systems (adapted from Fath [5]).

Waste Heat Recovery

M.F.A, Goosen et al. /Desalination 129 (2000) 63-89 67

The problem of low daily productivity of solar stills has encouraged scientists to investigate various means of improving thermal efficiency in order to minimize water production costs [5]. This can be achieved by recovering wasted latent heat of condensation so as to increase production of distillate water and improve system efficiency. It may be carried out in two or more stages, generally referred to as multi-effect solar distillation. It has become apparent that a key feature in improving overall thermal efficiency is the need to gain a better understanding of the thermodynamics behind the multiple use of the latent heat of condensation within a multi-effect humidification-dehumidification solar still [6].

While a system may be technically very efficient, it may not be economic (i.e., the cost of water production may be too high). Fath [5] noted that there is a shortage of research on economic analysis to determine the ultimate cost of the water product. Materials of construction, plant size and location, and operating costs (i.e., energy and labor) must all be taken into account. Therefore, both efficiency and economics need to be considered when choosing a solar desalination system.

In this paper the thermodynamic efficiency of solar water desalination technology was reviewed with special emphasis on humidification- dehumidification systems. A secondary aim was to assess how the economics of a system affect the final cost of the produced water. The challenges and opportunities of solar energy were also briefly discussed.

2. Solar desalination technology

The distillation of sea or brackish water to obtain fresh water can be accomplished by exposing thin layers of salt water to solar radiation and condensing the water vapor produced in such a way that it can be collected in receiving troughs. Since many uncontrollable

conditions exist in solar distillation, quantitative analyses are often difficult to perform. Therefore, the unit design has been largely empirical. Mahdi and Smith [7], for instance, combined a V-trough solar concentrator or reflector with an inclined flat-plate wick-type solar still. The equipment was used to investigate the enhancement of the outdoor performance of the wick-type solar still by the solar reflector. The authors concluded that the use of the solar concentrator with the inclined wick-type solar still can lead to a greater fractional increase in still efficiency and productivity on clear days in winter than on clear days in summer. In a similar study, Varol and Yazar [4] enhanced the distillate output of a single-basin solar still by coupling it with a flat-plate solar collector and coating a thin layer of SnO2 on one side of the transparent cover plate. The heat transfer fluid was circulated between the still and the collector through a heat exchanger and storage tank by thermosyphonically-induced flow. It was observed that good insulation around the storage tank considerably increased the yield at night due to the decrease of the ambient temperature. Thermosyphonically-induced flow eliminated the need for pumps and control units, thus reducing capital costs. A layer of SnO2 on the transparent cover lowered thermal radiation loss, thus increasing the source of heat energy in the solar still. The distillate yield was measured to be 6.7L-m 2"d-I for a September solar radiation of 17,820 kJ. The overall efficiency of Varol and Yazar's improved still (i.e., coating cover plate with SnO2) was found to be 3.26 times greater than that of a conventional still.

The operating efficiency of basin-type solar stills is low (around 35%), mainly due to the rejection to the atmosphere of the latent heat of condensation. Consequently, the production is also low, less than 5 L.m-Z'd ~ for a good insulation climate [8]. Furthermore, many devices have operational problems such as condensation on the inner side of the glass and algae growth, both

68 A~F.A. Goosen et al. / Desalination 129 (2000) 63-89

decreasing the radiation absorption of basins. Several semi-empirical attempts have been made to increase production based on the utilization of several evaporation stages after using in each stage the latent heat of condensation rejected by the preceding stage. An alternative method to increase the efficiency is to increase the evaporation area. For example, Joyce et al. [9] increased the evaporation area by spraying the salt solution in the air and then condensing the humid air. They generated experimental results that were used to tune a computation model of the chamber to obtain the performance (daily production and energy consumption) of the system when connected to a solar collector.

Fath [5] presented an excellent overall review and technical assessment of recent developments in various single- and multi-effect solar stills. He concluded that solar distillation is particularly appropriate for rural areas when weather conditions are suitable and the demand is not so great (i.e., less than 300m3"d-1). He noted that while many researchers were concerned with increasing still efficiency and productivity, there was a shortage of research on economic analysis to assess the ultimate cost of the water product. Based on the discussion presented in Fath's paper, it was recommended that a combination of the following design and operational parameters should be considered in future developments in solar distillation systems: • higher basin temperature (lower water level,

use of wick, adding black dyes, additional external heating-collector, concentrator, and waste heat recovery)

• lower cover temperature (cover cooling, multi-effect, overnight with basin energy storage, and additional condenser)

• large evaporation and condensation surface areas,

• re-utilization of the latent heat of condensation (multi-effect)

• minimize heat losses (good side and bottom insulation) and

• utilization of the shaded area (additional condenser and combined stills).

Given that the design of many solar desalination systems is largely semi-empirical in nature, there is the need to increase our fundamental understanding of both the thermodynamic as well as the economics behind such systems. It is also important to remember that water is a strategic commodity. An article by AI-Hajri and A1-Misned [10], for example, highlights the dimension of water problems in the Arabian Gulf and suggests a strategic option. Presently, the scarcity of water supplies in the Arabian Gulf countries are being overcome by desalination plants using depletable fossil fuels. The region will face major challenges to secure water supplies in the next century and beyond. Coupling seawater and solar energy is one viable option to secure water needs for future generations.

3. Thermodynamic considerations in solar desalination systems

3.1. Single-basin solar still productivi ty

A good understanding of the thermodynamics behind a particular desalination system is required in order to optimize the water production rate. This may be accomplished, for example, by using mathematical modeling. Yeh et al. [ 11 ] employed both theoretical and standard empirical equations for designing basin-type solar distillers with reduced operating pressures. The equations were derived from energy and material balances coupled with experimental data. They reported that considerable improvement in productivity was obtained by reducing the operating pressure in the solar distillers. In a similar study by Yeh and Chen [12], the energy balances for basin-type solar distillation with air flowing through the still were derived. It involved an experimental investigation of the effects of dry

IVL F.A. Goosen et al. /Desalination 129 (2000) 63-89 69

airflow rate and temperature difference between the salt water and atmosphere on the ratio of absolute humidity of the flowing air at the outlet to the absolute humidity of air saturated with water vapor. A correlation equation was derived. Considerable improvement in productivity was obtained if the water vapor was carried away directly by the flowing air. Presumably, this was due to the fact that the absence of condensed water vapor on the transparent plate enhanced the direct radiation reaching the saline water, thus improving the evaporation rate.

Singh [13] reported that a considerable effort was put into trying to understand the manner in which the productivity of solar stills is affected by design and operating variables. The effects of the design changes on the productivity were reported in this study. Numerical computations based on energy balances of different components of this system confirmed an improvement in productivity between 24% to 50% depending on the water flow rate from the vertical water column on the multi-wicks. This design incorporated a multi-wick solar still and a conventional basin-type solar still. The key to the success of this design was that hot water at considerably higher temperature than the ambient temperature was obtained.

Solar still productivity was enhanced by AI-Hussaini and Smith [14] using vacuum technology. The purpose of a vacuum inside the still was to avoid any heat transfer due to convection. Therefore, the heat loss from water in an insulated still would be mainly due to evaporation and radiation, thereby eliminating the need for including air mass transfer in the analysis. In the presence of vacuum, the effect of the non-condensable gas, which reduces the rate of condensation, was also avoided. It was shown that applying vacuum inside the solar still increased the water productivity by about 100% from 3.79L'd I to 8.07L-d -1. In an alternative method, bromine and iodine were utilized by AI-Abbasi et al. [ 15] as heat-generating media for

the enhancement of the overall efficiency of basin-type solar stills. Results of their experimental work showed that the solar still of bromine medium had a higher efficiency than that of the iodine medium. However, both stills had significantly higher efficiencies when compared to that of a conventional one. The results also showed that the productivity of the still of the bromine medium was about 65% greater than that of a conventional still. For an iodine-medium still, the productivity was about 42% greater than a conventional one. A good overview of solar desalination schemes is given by Hamed et al. [2] where nearly 50 scientific papers and technical reports were reviewed.

Quantitative assessment of the solar radiation incident on a tilted plane is very important to engineers designing solar energy-collecting devices, to architects designing buildings, and to agronomists studying insolation on vegetation on a mountain slope. Although solar radiation (including direct and diffuse components) has been measured on horizontal surfaces at many locations, the measurement on sloping surfaces has rarely been carried out [16]. For locations where no measurements exist, radiation data can be estimated from theoretical models. Several methods are available to compute global radiation on tilted surfaces from measured daily direct and diffuse radiation on a horizontal surface. Measured solar radiation data were used by Gopinathan [16] to test the applicability of two theoretical models for computing total solar radiation on inclined surfaces. Those selected included the isotropic and anisotropic models. The two models were compared and tested for their applicability to locations in southern Africa on the basis of statistical error tests. A comparative study showed that both models were equally accurate. The isotropic model was used to estimate total radiation for five other locations on north-facing surfaces of various tilt angles. Similarly, Chiou and E1-Naggar [17] developed a simple equation for the explicit determination

70 A/LF.A. Goosen et al. / Desalination 129 (2000) 63-89

of the optimum slope for solar insolation on a flat surface tilted towards the equator during the heating season. This equation was valid for both the northern and southern hemispheres.

The thermal performance of a single deep- basin still was investigated experimentally and theoretically by Aboul-Enein et al. [18]. Energy balance equations were written to include the basin liner, basin water, and glass cover. They assumed that the heat capacities of the basin liner, glass cover, and insulation were negligible compared to that of the basin water. Also, additional assumptions included the system being vapor-tight without the existence of a temperature gradient across the thickness of the glass and basin water. The hourly productivity of the still, Ph, was given by

3 6 0 0 h e ~ g ( T - T , ) = ( 1 ) L

where Lw is the latent heat of vaporization of water, hewg is the evaporative heat transfer coefficient, and T,~ and Tg are the temperatures of the water and glass, respectively. The daily efficiency of the still, rla , was given by

1 O0 Pd L ~d - A At Y, I (2)

p g

where At is the time interval during which the solar radiation was measured, L~ is the average latent heat of vaporization of water, Pd is the daily productivity, Ap is the surface area of the basin liner, and Ig is the solar radiation intensity incidental on the glass cover. Good agreement between the experimental and theoretical results was observed. The influence of cover slope on the daily productivity of the still was also investigated as well as the effect of heat capacity of basin water on the daylight and overnight productivities. It was inferred that the

productivity of the still decreased with an increase in heat capacity of basin water during daylight while the reverse was true overnight. The optimum tilt angles of the still cover were found to be 10 ° and 50 ° during the summer and winter seasons, respectively. The daily efficiency of the still, though, was relatively low at about 27%.

In a similar modeling study, Kumar and Tiwari [3] performed a thermal analysis of a single solar still coupled to a flat plate collector in order to optimize water yield. They obtained the same hourly yield or productivity equation as Aboul-Enein et ai. [18] [refer to Eq. (1)]. How- ever, their overall daily thermal efficiency (rid) equation was slightly different since it included a term to account for the effect of the solar collector. It had the form

lqa = [~I(t)A + ~I ' (t)Ac]3600 (3)

where mew [a term similar to P~ in Eq. (1)] is the instantaneous yield or productivity of the still per unit area (kg.m-2); L is the latent heat of vaporization (J/kg); I(t) is the same as I x in Eq. (2) (W.m 2); A~ is the same as Ap (m2); Ac is the area of the solar collector (m2); and I'(t) is the total radiation available in the plane of absorber of the collector (W.m-2). It was found that the yield increased with the number of collectors going from 5 to 20 kg ofwater'm-Z'd - ~ as the number of collectors increased from 2 to 8. The overall efficiency, however, was the highest with four collectors at 18%. It should be noted that this efficiency is even lower than that obtained by Aboul-Enein et al. [18] for a single deep-basin still. They also observed that the optimization process was a strong function of water depth.

Two conclusions can be drawn from this section on the productivity of single-basin solar stills. The first is that the theoretical analysis

M.F.A. Goosen et al. / Desalination 129 (2000) 63-89 71

(i.e., modeling) of solar desalination systems is an effective tool for understanding and predicting system performance. The second is that the efficiency of single basin solar stills, even under optimum conditions, is very low.

3.2. Efficiency o f multiple-effect solar stills

In conventional single-effect solar stills, the collected solar heat is used only once since the heat of condensation of the vapor is transmitted through the cover and discharged into the atmosphere by convection and radiation. A double-effect solar still using the solar energy directly may be developed if the upper side of the enclosure is made of transparent material which transmits solar radiation. Thus, the upper side of the bottom plate serves as a solar collector, thereby making operation of the still directly responsive to the solar energy input and taking the place of a separate solar collector and storage tank. In order to collect more insulation, the equipment can be tilted from the horizontal, with the cover surface inclined towards the sun. Several weirs can be installed in parallel on the upper surface of the glass to hold the saline water, while a blackened wet jute cloth can be placed on the upper surface of the absorbing plate to form the liquid surface. Except for the glass cover, all parts of the enclosure need to be carefully insulated to make the heat loss as low as possible. The energy balances for such an upward-type double-effect solar still were outlined in papers by Yeh and Ma [19] and Yeh et al. [20]. They employed the energy balances to determine the theoretical performance of the still under various operating and design conditions.

During operation of an upward-type double- effect solar still, the transparent glass cover and glass plate transmit solar radiation. The absorbing plate is then heated directly by solar radiation. The saline water feeds are introduced on the upper surfaces of both the absorbing and glass plates where some water evaporates, while

the remainder is collected at the bottom and discarded as concentrated brine. The vapor produced from the saline water rises from the absorbing plate and is condensed on the lower side of the glass plate. The latent heat of condensation given up by the condensing vapor on the lower side of the glass plate is conducted through this plate and provides heat to evaporate an equal amount of water from the saline brine sliding down the upper side of the glass plate. Finally, the heat of condensation given up by the condensing vapor on the lower side of the glass cover is transmitted though this cover to the ambient air above the still (i.e., the heat is lost from the system). The condensates produced from the first and second effects are collected as fresh water in a trough from the enclosure. In some designs, the saline feed water is preheated by passing it over the glass water in order to capture the remaining latent heat of condensation (Fig. 1).

The main difference of a downward-type, double-effect solar still is that the water vapor must be transferred from the lower side of the upper plate downward to the upper side of the lower plate for condensation and to give up its latest heat for reuse. A downward concentration gradient exists between the plates, both in water vapor and temperature, thereby suppressing free convection and mass transfer of water vapor. This has a negative effect on solar still productivity or efficiency. Thus an upward-type double-effect solar still is preferred over a downward-type system.

A multiple-effect distiller is generally more effective thermodynamically than a single-effect distiller since the former uses the available heat energy more than once. Yeh and Chen [21] experimentally investigated the application of the multiple-effect concept to the direct utilization of solar energy in upward-type distillers. As mentioned above [19,20], an upward-type double-effect solar still was more efficient than the downward-type still since it creates

72 M.F.A. Goosen et al. /Desalination 129 (2000) 63-89

free-convection currents while reusing energy. Yeh and Chen [22] assessed the effect of airflow, through the second effect unit of wicks downward-type double-effect solar distillers, on productivity. When the airflow was conducted at an optimal rate, considerable improvement in productivity (i.e., kg of water.m 2.h-I) was obtained, especially at the low end of solar radiation. During operation, the suction by capillary action of the inclined cloth fibers provided the liquid surface at which some of the saline water feed was evaporated. The remainder was collected at the bottom and discarded as concentrated brine. The vapor produced from the saline water moved down the lower side of the first plate and was condensed on the upper side of the second plate. The heat of condensation given up by the condensing vapor on the upper side of the second plate was conducted through this plate and furnished heat to evaporate an equivalent amount of water from the saline brine sliding down the lower side of the second plate. The flowing air, which caused the improvement in productivity, was steadily introduced by using a blower through the second effect of the distiller device to carry away most of the water vapor.

Fernandez and Chargoy [23] designed and modeled a desalination system with the thermocline traveling upwards. Specifically, their solar still was built based on the principle of a stacked tray array for tandem distillation and heat recovery. Heat was supplied to the lower-most tray, containing seawater like all the rest, and a diffusion distillation occurred. Evaporation from the hottest tray caused condensation onto the upper, colder one, thus producing distilled water and a flow of heat upwards. As this exchange proceeded, the thermocline traveled upwards. This procedure was found to be dependent on the partial pressure difference of water vapor just as in greenhouse solar stills. A simple mathematical model was evolved and calibrated with field data to make it adequately fit experimental results gathered from 14 months of continuous

operation. They noted that conditions could be identified with this model for which the added cost of particular design features were adequately compensated for by an increased return of distilled water from the sea.

Yeh and Ma [19] developed a relation for the fraction of incident energyf(z) remaining after passing through a thickness z as follows:

f ( z ) = a - b In (z +c) (4)

where the constants a, b, and c may be determined from experimental data. Good agreement was obtained between calculated and experimental data. They confirmed that an upward-type double-effect solar still is more effective than a downward-type unit because it may create a free-convection effect while preserving the effect of reused energy. On the economic side, the cost for constructing an upward-type solar still will be less than a downward-type solar still because the former needs only two working chambers while the latter needs three.

Single- and double-basin solar stills can be coupled with a flat-plate collector to enhance the distillate output. Such solar stills can be operated in two modes. The circulation of water between the collector and still through a pump is referred to as the forced-circulation mode. The natural circulation of water between them is known as the thermosiphon mode (Fig. 1). The forced- circulation mode needs electrical power to run the circulating pump, which is not required in thermosiphon mode. In a study by Yadav and Jha [24], a transient analysis was presented of a double-basin solar still coupled with a flat-plate solar collector in the thermosiphon mode. Explicit expressions were derived for the temperatures of the upper glass-cover, the water in the upper basin and lower glass-cover, the water in the lower basin, the distillate output and system efficiency. The performance of the proposed system operated with forced circulation


was found to be slightly better than that of the system using the thermosiphon mode. The slight improvement in thermodynamic efficiency (i.e., production of fresh water), therefore, has to be balanced with the increased capital and operating costs associated with installing a circulation pump.

A paper by Kumar et al. [25] was devoted to the development of a transient model to study the performance of a double-basin solar still. The still was integrated with a heat exchanger in order to enhance its distillate output per unit area. This integration was achieved by flowing hot fluid through the still. For example, the use of waste hot water (available from power or chemical plants) is of great importance since it can significantly enhance the daily distillate output of a still. A detailed analysis was carried out by Kumar et al. [25] for the integrated system, incorporating the effects of various system parameters such as water mass in the upper and lower basins, heat-exchanger length, inlet temperature, mass-flow rate of heat-exchanger fluid, several heat-transfer coefficients, and meteorological parameters. It was found that the use of a heat exchanger significantly increased the system efficiency. The effect of the heat exchange on the system capital and operating costs was not mentioned.

In a recent paper, Bouchekima et al. [26] reported on the performance of a capillary-film solar distiller. The thermodynamic efficiency and system economics were assessed in this study. The device was essentially a double-effect, downward-type, wick solar distiller. The wick in this case was a very thin layer of tissue with a fine mesh saturated with brine water that maintained itself in contact on the underside of a metal plate due to the surface tension which was much greater than the gravitational force. The unit included three cells in series. The temperature difference between plates in each cell was in the range of 5 to 10°C. It was found that the efficiency of the distiller increased with an

increase in the temperature of the brackish water at the inlet and with increasing intensity of solar radiation. It depended also on the heat loss and the type of fabric used. The efficiency could be increased and the cost decreased if the plant was operated all day long.

Several conclusions can be drawn from this section. Multiple-effect solar desalination systems are more productive or efficient than single- basin stills in producing fresh water due to the reuse of latent heat of condensation. The increase in efficiency must be compared to (i.e., balanced against) the increase in capital and operating costs of a multi-effect solar still compared to a simpler single-basin solar still. The efficiency of a multiple-effect solar still can be increased by inclining the glass cover surface towards the sun and installing weirs (i.e., groves) on the upper surface of the glass to hold and warm the saline water before it enters the still. The efficiency of the system can also be improved by running it in an upward-type mode, with the saline water evaporating from the upper surface of a glass plate and then having the vapor condense on the lower side of the glass plate immediately above the first plate. At low solar radiation, employing flowing air to carry away most of the vapor between the plates will improve the total solar radiation reaching the saline water. Finally, the efficiency of the system can be improved by employing mathematical modeling to optimize still efficiency. The addition of flat-plate collectors and heat exchangers to transfer waste heat from local industry/plants to the solar still provides an additional way of enhancing the productivity of the system.

3.3. Nove l still designs

Various designs have been proposed in an attempt to improve the performance of solar stills. Hence, there is a need for the characteri- zation of such designs in order to compare their relative performance. Several expressions were

7 4 M.b.A. Goosen et al. / Desalination 129 (2000) 63-89

given by various authors for the internal convective heat transfer coefficient. The study of internal free convection in a triangular enclosure filled with humid air with the hot surface facing upwards continues to be a challenging area in engineering applications. Tiwari and Noor [27], for example, attempted to derive suitable parameters to assist in characterizing various designs of solar stills by analyzing the conventional-passive single-slope solar still with a triangular (trapezoidal) cavity.

The concept of instantaneous thermal efficiency was introduced by Tiwari and Noor [27] to characterize and standardize various designs of solar stills including the trapezoidal cavity system. The instantaneous thermal efficiency is a relatively new concept in solar distillation studies and is valid for solar stills with minimum basin-water heat capacity. It can be expressed as

r]'x = F'[ r'eff +Ur T~°-T] I(t) ] (5)

rh

T w - T,

l(t)

Fig. 3. Characteristic curve of a solar still (adapted from Tiwari and Noor [27]). I, solar still with negligible heat capacity and large time intervals (the ideal solar still); II, flat-plate collector; IlI, solar still with large heat capacity and small time interval; IV, solar still with medium heat capacity under normal operation condition. Area covered between curve III and rl,-axis, respectively.

where rli s is the characteristic curve of a solar still, F' is a solar still efficiency factor, ~:eff is the effective absorptance, Ur is the total heat loss coefficient, I(t) is the incident solar radiation, T,~ o is the initial water temperature, and Ta is the ambient temperature. The slope of the characteristic equation is positive, unlike the slope of the characteristic curve of a flat-plate collector (curve II, Fig. 3).

The ideal solar still shall have a characteristic curve similar to curve I of Fig. 3. However, a practical solar still will deviate from ideal performance due to the heat capacity of the water mass and the sensitivity of evaporation heat transfer with temperature (curve IV). On the other hand, curve II! in Fig. 3 indicates that the solar radiation is not effective for distillation.

In another case, an inverted trickle solar still was developed by Badran and Hamdan [8]. The concept behind their device was based on the

flow of a thin layer (trickle) of water on the back of an absorber plate. This layer was maintained attached to the plate by means of a wire screen welded to the plate, together with the effect of surface tension force. As the water moved downward along the plate at a very low flow rate, it was heated up and evaporated. The vapor was transferred to an adjacent compartment, similar to the flowing air system of Yeh and Chen [22] where it was condensed and collected. The main feature of this concept was the elimination of condensation on the glazing thereby increasing the solar radiation reaching the absorber plate, and increasing the temperature difference between the evaporator and the condenser. The productivity of their system reached a value of 3.83 L-m 2"d-1. This represents one-half the value achieved for the inclined V-trough/flat-plate wick-type still reported by Mahdi and Smith [7].


In an interesting study by Minasian et al. [28], a simple solar earth-water still of a single-sloped type designed for producing drinking water in remote areas was investigated. The still was installed using three formats: over an insulated hole, a hole with an insulated base, and a hole with insulated walls. Multiple linear regression equations were developed to estimate the productivity of these stills. These equations relate to ambient air temperature, wind speed, and solar radiation. The study showed that the average wall's contribution in supplying fresh water was about 56%, whereas the base contribution was about 31%. It was concluded that setting many stills on a number of small holes will give higher output compared to setting a still on one large hole of the same volume since the former setup gives a greater wall area.

To enhance the yield, a simple solar still coupled to an external condenser was modeled and analytically investigated by Fatani and Zaki [29]. Using actual meteorological data, their results showed that the yield improvement depended upon the brine level and surface area provided for heat removal. For an ideal still with negligible thermal inertia (i.e., water depth approaching zero), an auxiliary condensing surface at a temperature below the glass temperature decreased the daily yield. The still yield was improved for brine levels up to 10 cm above this level. Additional heat removal from the still did not significantly improve the daily productivity.

In a unique technique for increasing the evaporation rate in distillation, Armenta-Deu [30] added a non-volatile surfactant to the water to be distilled. Through the use of a surfactant, the investigator was able to increase production of drinking water using the same energy. This was possible since there is linear dependence between the latent heat of vaporization and the surface tension of water.

The thermodynamic information that has been gained from an analysis of solar water distillation has also been used for other applications such as

in ethyl alcohol distillation in a basin solar still [31 ], for a solar powered absorption/refrigeration system, and for a roof-spray cooling system [32]. All three reported on the use of thermodynamic models in their investigations.

In summary, the section has shown that novel still designs can be relatively simple but effective in improving water production. For example, a triangular cavity solar still with the hot surface facing upward shows promise as well as the removal of the condensation on the glazing to increase solar radiation. For solar earth-water stills, a larger number of small holes is more effective than one large hole due to the greater wall area of the former (the wall contributes the most in supplying fresh water). Finally, it is clear that surfactants can work to improve water production efficiency by reducing the surface tension of the saline water.

3.4. Case s tudies

A solar desalination plant in Abu Dhabi, built in 1984, had a capacity of 120m3"d I and used seawater as the feed to a multi-effect-stack distillation unit [33]. The thermal requirement of this distiller was provided by a bank of evacu- ated-tube flat-plate collectors having a total collector area of 1862m z. The plant also con- tained a heat-accumulator system, which allowed the solar energy collected during the day to be utilized at nighttime, thus allowing the plant to run continuously 24h/d. The heat accumulator was thermally stratified and used distilled water as the heat-storage medium. It was found that the temperature distribution inside the heat accumulator significantly affected the performance of the evaporator and its water production. It is important, however, to be able to predict accurately how this temperature distribution varies throughout the day. E1-Nashar and Qamhiyeh [33] developed a one-dimensional, unsteady-state heat-transfer computer model and compared the results with the actual temperature


measurements taken at the solar plant. The model was developed in such a way as to take into consideration the different modes of operation of the heat accumulator which include charging, discharging, combined charging/discharging, and free cooling. A comparison between the calculated and measured temperature distributions indicated that the accuracy of the model was reasonably good.

In a study by Adhikari et al. [34] using a similar multi-effect-stack system as that employed in the Abu Dhabi plant, a computer simulation model was presented for the steady-state performance of the still. The model was validated by the results of simulated experiments on a three-stage unit having an immersion-type electric heater as the heating source. The results obtained from the model using the modified heat and mass transfer relationships proposed in the paper were in good agreement with those obtained from the experiments. Numerical results were also presented in the paper to compare the relative performance of a multi-stage stacked-tray solar still with a diffusion-type multi-stage still.

In another case study, a double-basin solar still was designed and tested at the Indian Institute of Technology for the climatic conditions of Delhi [35]. The overall efficiency of the system decreased during operation at elevated temperatures. The decrease in the efficiency of the system was explained by the fact that heat was not lost from the glass cover by evaporation because of stagnant water in the upper basin, In addition, there was a maintenance problem that made the original design unacceptable from a commercial point of view, particularly since frequent cleaning was required of the upper basin to remove the salt deposition associated with the structural design of the basin. In a revised design there was rapid heat transfer caused by water flow in the upper basin, and as such was different from the previous design. The collector panels were integrated into the lower

basin of the system though a tube-in-tube heat-exchanger for high-temperature distillation. Based on energy balances for different components of the double-effect distillation unit under an active mode of operation (i.e., feeding of thermal energy into the basin from an external source similar to the heat exchange system used by Kumar et al. [25]), an analytical expression was derived by Tiwari and Sharma [35] for the daily yield as a function of the system and climatic parameters. An increase of about 30% in the daily efficiency was observed by using double-effect distillation when the flow rate was small.

The highlighted examples indicate that solar desalination plants can be effectively modeled and that the models can be employed to improve the efficiency of the desalination process. Still designs that consist of humidification~lehumidification systems will be considered next, in some instances combined with greenhouses for growing crops.

4. Humidification-dehumidification systems

The earliest humidification-dehumidification solar distillation plant on record was designed and built in 1872 by Charles Wilson in Chile [36]. Sixty-four separate water basins were constructed of wood, with sloping glass covers. The total water surface area was 4459m 2 (48,000ft2). The still produced 22.7m3.d l (6000US GPD) of fresh water for a mining operation. The still was still in operation 36 years later. The concept of humidification of air, followed by dehumidification to collect fresh water, is therefore not new. It was further developed at the University of Arizona in 1961 in cooperation with the Georgia Institute of Technology and the University of Sonora, Mexico, at Puerto Pefiasco, New Mexico. The group built a pilot plant called "humidification cycle distillation". Specifically, solar energy was

M.FA. Goosen et al. / Desalination 129 (2000) 63-89 77

used to preheat the saline water for the humidification-dehumidification cycle. In this system the basic functions of heating the saline water, evaporating a portion of it, and condensing the vapor were performed in three separate components. Therefore, it was not a conventional simple solar still.

In the early experimental plants built at the University of Arizona and in a pilot plant constructed at Puerto Pefiasco, solar energy was used to warm the saline water to a final temperature of 65.6°C (150°F) before it was circulated through an evaporation chamber. The five solar absorbers at Puerto Pefiasco were shallow troughs (2" deep) lined with plastic or butyl rubber and covered with two layers of clear plastic film of PVC or PVF. The inside plastic film rested on the surface of the water, and the outer film was inflated to provide an insulating effect. The total area of the solar absorbers was 966 m 2 (10,400 ft2). The pilot plant produced over 18.9m3"d 1 (5000 GPD). In 1965, the Puerto Pefiasco plant was modified so that it could operate from the waste heat of a 60kW diesel- electric set rather than from solar collectors. Both jacket-water and exhaust-gas heat were utilized. The overall plant production increased to 22.7 m<d - 1 (6000 GPD) using only waste heat.

A more recent example of a humidification- dehumidification system is a pilot plant built at Kuwait University [37]. The system consisted of a salt gradient solar pond, 1 7 0 0 m 2 in surface area, used to load the air with humidity. Fresh water was then collected by cooling the air in a dehumidifying column, producing 9.8m3.d -~ of distillate. Another system, a solar rain tower desalination plant, was installed in Tunisia. Brackish feed water was employed to produce 2.4-3.6 m<d - 1 fresh water at 60-80 °C. The fresh water was used for drip irrigation to grow crops and for greenhouse heating. The overall system consisted of a spray and a cooling tower. It had an 81m 2 active surface area in the flat-plate collectors.

An air dehumidification method suitable for coastal regions was also described by Khalid [38]. Moist air passed over the cooling coils of an air conditioner with an effective coil temperature lower than the dew point of the humid air. The fresh water produced was suitable for human consumption. Khalid noted that the method was economic if the fresh water were considered as an air-conditioning by-product. He performed mass and energy balances on the control volume of the cooling coil. The following balance equations were obtained. These include the dry air balance

: m =m (6) m " 2 a

where m a is the dry air flow rate; the moisture mass balance

~ l m , = mw +w,2m, (7)

where m w is the water condensate mass flow rate and % is the specific humidity of air (humidity ratio); and the energy balance

q + m h = m h +m h (8) a a I a a 2 w w

where Cp,, is the mean specific heat for air water vapor mixture (Cpm = Cry+ cOCpv), Cp, is the specific heat of dry air, Cpv is the specific heat of water vapor, q is the rate of heat transfer, ho is the air enthalpy, hw is the enthalpy of the condensate, t~ is the air dry-bulb temperature, and hfg is the latent heat of vaporization. Khalid [38] showed that for an ideal process, the end temperature

78 h/LF.A. Goosen et al. /Desalination 129 (2000) 63-89

should be on the saturation line. However, since the cooling coil surface is finite, only a portion of the air stream reaches the surface temperature while the rest is bypassed with the initial condition 1 resulting in a mixture at condition 2. The enthalpy moisture ratio, Ah/A(co), is constant along the condition line. The actual coil condition reveals that the coil surface temperature changes along the heat transfer path. The curve becomes steeper as the air progresses through the coil. One interesting observation was that if the coil surface at the air inlet was dry, condensation did not occur until further in the coil. Condensation occurred where the surface temperature of the coil dropped to the dew point of the entering air at state 1.

Paton and Davis [39] used the humidification- dehumidification method in a greenhouse-type structure for desalination and for crop growth (Fig. 4). Their seawater-greenhouse produced fresh water and crop cultivation in one unit. It was suitable for arid regions that have seawater nearby. The temperature differences between the solid surface heated by the sun and cold water drawn from below the sea surface was the driving force in the system. The greenhouse acted as a solar still providing a controlled environment inside the greenhouse. A thermodynamic model was employed to optimize the greenhouse design (i.e., dimensions, evaporator condenser areas, water flow rates, and temperatures).

Different multi-effect humidification- dehumidification units have been built and tested by German firms in Indonesia, Portugal, and the Canary Islands [40]. The productivity of these units was as high as 20L.m-Z-d ~, which was of an order five times higher than that of the single-basin still. Farid and A1-Hajaj [40] stated that multi-effect humidification-dehumidification units with forced circulation, while giving better conditions for heat and mass transfer, were only practical (i.e., economically viable) if wind energy were used to power the fan in place of electrical power. They built a humidification-

dehumidification unit based on a closed-air, open-water cycle. It consisted of a humidifier, a flat- plate collector and a cooling coil condenser. The unit was provided with an air blower, water pump, and feed and preheating tanks as shown in Fig. 5. The humidifier and condenser were made out of hard PVC pipes connected to form a loop with the blower fixed at the bottom. The condenser was made of a copper pipe mechanically bent to form a helical coil fixed in the PVC pipe. The preheated feed water was further heated in a flat plate collector. The hot water leaving the collector was uniformly distributed over wooden-shaving packing in the humidifier.

Farid and AI-Hajaj [40] defined the system efficiency with a performance factor (PF) which was the ratio of the energy utilized for water evaporation to the incident solar energy. This definition is somewhat similar to the overall efficiency defined earlier (i.e., the actual amount of fresh water produced divided by the maximum amount of fresh water that can be produced based on the incident solar radiation). They found that decreasing water flow rates increased the values of PF due to the increase in the temperature of the water leaving the collector, causing more efficient evaporation and condensation. However, decreasing water flow rates below 70kg.h -t reduced the PF as a result of the decrease in the efficiency of the collector at elevated temperatures. The effect of the air was complicated by its combined effects on the heat and mass transfer coefficients and on the driving force in both the humidifier and the condenser. Finally, a daily productivity of 12 L'm-Z'd - 1 was achieved which was over three times that of a single-basin still.

In a similar study, a closed air cycle humidification-dehumidification process was used by AI-Hallaj et al. [6] for water desalination. The circulated air by natural or forced convection was heated and humidified by the hot water obtained either from a flat-plate solar collector or from an electrical heater. The latent heat of

,.7

/

.ak .

°'

/

Cro

ps g

row

n i

n co

oled

sh

ade

ten

t ar

ea

Dee

p $e

,a W

,ite

r In

t,lk

e ~

- ,

~t~

PM

#~ ~

t~r

Sea

Wat

t* R

etur

n "

" S

tora

ge

... "

"~,

20m

3

Fig.

4.

Seaw

ater

gre

enho

use

[39]

. 1.

Sur

face

sea

wat

er t

rick

les

dow

n th

e fr

ont

wal

l ev

apor

ator

thro

ugh

whi

ch a

ir is

dra

wn

into

the

gre

enho

use.

Dus

t, sa

lt sp

ray,

pol

len

and

inse

cts

are

trap

ped

and

filt

ered

out

leav

ing

the

air

pure

, hum

idif

ied,

and

coo

l. 2.

Sun

light

is

sele

ctiv

ely

filt

ered

by

the

roof

ele

men

ts t

o re

mov

e ra

diat

ion

that

doe

s no

t co

ntri

bute

to

phot

osyn

thes

is.

Thi

s he

lps

to k

eep

the

gree

nhou

se c

ool

whi

lst

allo

win

g th

e cr

ops

to g

row

in

high

lig

ht

cond

itio

ns.

3. A

ir p

asse

s th

roug

h a

seco

nd s

eaw

ater

eva

pora

tor

and

is f

urth

er h

umid

ifie

d to

the

sat

urat

ion

poin

t. 4.

Sat

urat

ed a

ir p

asse

s th

roug

h th

e co

nden

ser,

whi

ch i

s co

oled

usi

ng c

old

deep

sea

wat

er.

Pure

dis

tille

d w

ater

con

dens

es a

nd is

pip

ed t

o st

orag

e. 5

. Fan

s dr

aw th

e ai

r th

roug

h th

e gr

eenh

ouse

an

d in

to t

he s

hade

hou

se a

rea.

5"

t.,a I ',o

80 ~V~F.A. Goosen et al. / Desalination 129 (2000) 63 89

v4

v3 ' [ t :~ % ,2

@

PRODUCT ~ ] (distilled waler)

t - - F . . . . . . . . . . . . . . . . . . . . .

fi)

f Y @ £

Fig. 5. Schematic of a humidification-dehumidification desalination system. 1 preheating tank, 2 feed tank, 3 condenser, 4 collector, 5 humidifier, 6 rotameter, and 7 air blower (adapted from [40]).

condensation was recovered in the condenser to preheat the saline feed water (Fig. 6). The productivity of the units tested was found to be much higher than those of the single-basin stills. Moreover, these units were capable of producing a large quantity of saline warm water for domestic uses other than drinking. No significant improvement in the performance of the desalination units was achieved using forced air circulation at high temperatures. At lower temperatures, a larger effect was noticed. This was related to the low heat and mass transfer coefficients at low temperatures and to the non-linear increase in the water vapor pressure with temperature.

Two units similar in design to that shown in Fig. 6 but different in size and materials of construction were fabricated by Al-Hallaj et al. [6]. A bench unit was made from Piexiglas and a

pilot unit from steel plates. The arrangement shown in Fig. 6 provided a closed air cycle and an open water cycle. The pilot condenser consisted of a rectangular duct. Within this duct was a galvanized steel plate with a copper tube in a helical shape, welded to one side of the plate. The humidifier, on the other hand, was a typical cooling tower built of a wooden structure and fixed in a second duct. In normal outdoor operation, a tubeless flat-plate solar collector was used to supply the pilot desalination unit with the required heat. The hot water leaving the solar collector was sprayed over the packing in the humidifier (i.e., evaporator) section.

To check out the accuracy of the measurements, an overall energy balance was made. In the case of the pilot unit, the heat input from the solar collector must be equal to the summation of the heat rejected with the brine leaving the

M.FA. Goosen et al. / Desalination 129 (2000) 63 89 81

3

Evaporator4 v]

BRINE to t h e s e a

Flat plate solar collector

I Heat input tl-,~ I

6

DRY .

2 v

Condenser

1 from t h e s e a

Distillate

Fig. 6. Schematic of a closed-air and open-water humidification~tehumidification unit (adapted from [6]).

humidifier, the heat rejected with the condensate leaving the condenser, heat losses from the unit to the ambient through the walls of the unit, and the heat stored in the walls of the unit. The first two quantities were calculated easily by knowing the feed-water temperature, 7"1. Evaluation of the third term, heat losses through the walls, required an accurate estimate of the heat loss coefficient, Utos~. This was theoretically calculated by AI-Hallaj et al. [6] from the summation of all the thermal resistances using standard heat transfer textbooks. Heat loss through the walls was measured experimentally using the following equation:

where T 5 and T6 are the air temperature at the top

and the bottom of the unit, respectively (Fig. 6). The measured and the theoretically calculated Uloss were 1.0 and 0.6 W'm-Z-K l, respectively, for the pilot unit and 1.2 and 0.8 W.m-2.K ~, respectively, for the bench unit. The heat stored in the walls of the unit was very significant in the outdoor operation due to the transient nature of the solar energy. It was positive in the morning, indicating heat storage, and negative in the afternoon, indicating heat release. The energy balance was found to be correct within 10%. The results suggest that the production increased with night operation using the rejected hot water from the humidifier.

A computer program was developed [6,41 ] to study the effect of different operating parameters on the performance of the unit. Simulation results revealed that doubling the condenser area could increase production by 50%. However, productivity could only be increased slightly by increasing the surface area of the humidifier. The effect of feed-water temperature on the humidifier was only noticed at temperatures below 50°C. This was related to the low heat and mass transfer coefficients at low temperatures and to the non-linear increase in the water vapor pressure with temperature.

In summary, four main conclusions can be drawn from the reviewed literature. First, the concept of humidification of air, followed by dehumidification to collect fresh water, dates back more than 100 years. Second, the concept of using a humidification-dehumidification system for the growth of crops in a controlled environment of a greenhouse is relatively new. The next conclusion is that mathematical models (i.e., thermodynamic models) may be employed to optimize the desalination system performance. Finally, several humidification<lehumidification pilot plants have been built around the world. Lower operating costs in the form of alternative energy sources (e.g., waste heat or wind energy) were found to be key factors in their economic viability.


5. Economics of solar desalination

A review of cost estimation of solar desalination systems was reported by Delyannis and Belessiotis [37]. They noted that solar energy conversion plants are capital-intensive enter- prises. According to the authors, a general economic analysis is not easy to accomplish since only a few studies report on predicting the price of water or focus on economics. For solar distillation plants, the problem is compounded with the fact that most of them are constructed from inexpensive local materials using local personnel. In such a situation, prices differ considerably from one location to another. Hoffman [42] presented a detailed theoretical cost analysis based on real operational data for solar-driven desalination plants. He concluded that the price of water from solar-powered desalination plants ranges from 0.52 to 1.595.m -3. The figures apply to capacities over 100,000 m 3.d- 1.

The effect of site, technological and economic parameters, and the comparative cost of partial or fully solar-powered desalting systems were described by Glueckstern [43] for large installa- tions with capacities of 20,000-200,000m3"d -1. He presented investment prices that ranged from 15 to 20 $.m -2 of the surface area of the pond and 800-12005"m-3"d -1 of desalinated water produced. A similar analysis was given by Ophir and Nadar [44].

A solar distillation plant has a mean lifetime of about 20 y. The total annual cost of providing fresh water by solar distillation depends on the total annual interest and amortization rate, IA (percent of investment); the total cost of supplying salt water to the stills in a year, C~; and the total average annual maintenance and repair cost, M R (percent of investment). The main components of the annual average cost of distilled water is given by the following equation [37,45]:

C = I O I ( I A + M R + T I ) + IO00(OC) + C

where A is the still area on which yields are based (m2), c is the cost of operating labor ($'man-h-l), C is the cost of distilled water ($'m-3), C~ is the fixed and operating costs of salt water supply ($'m-3), I is the total capital investment ($), O is the annual operating labor (man-h), Yc is the annual unit yield of collected rain water (m3"m-2), Ya is the annual unit yield of distilled water (m 3. m z), and TI is the average annual taxes and insurance charges (percent of investment). The average annual interest and amortization rate, IA, is a function of the annual interest rate (%), r. It is defined as

1 Im =r 1 + (1 +r/100)n-1 (12)

where n is the payment period in years. For a solar-powered multi-effect distillation

plant operated by solar collectors or a solar pond, and for feed water of salinity 5000-35,000 ppm, the following equation can also be used [in place of Eq. (11)] to estimate the cost of distilled water C (in $'m -3) [46]:

]

_ N( IA+I ) 1.246 (2 +N) C e +1] C

J 868(1 +N) h ( A T - a N )

0 .199(IA+1)Cc 0.266 (IA +1) 7.54 + + +

/ (1 +N) 1 +N 1 +N

(13)

where ~ is the boiling point elevation of the brine (K), Cc is the cost of collectors ($'m-2), Ce is the cost of the heat transfer area of the evaporator ($'m-2), h is the overall heat transfer coefficient of evaporator (kW.m-2"K-1), L is the insolation (MJ'm 2"d-l), Nis the number of effects, and AT is the temperature difference between two consecutive effects (K).

In a related area, Voivontas et al. [47] analyzed water management strategies based on advanced desalination schemes (such as reverse


Technology Database

Estimate Desalination Plant

Investment Cost

Estimate Desalination Plant

O &M Cost

I Estimate RES I Investment Cost

EstimateO&MRES ]

Esthnate income t from water sales

Estimate income from energy sales

Set Inflation Rate Investment Life Discount Rate

t Estimate Annual ~_ Cost

Estimate Annual Income

Estimate

Water Production Cost Internal Rate of Return

Net Present value Specific Investment Cost

Fig. 7. Financial analysis of a renewable energy sources/desalination plant (adapted from Voivontas et al. [47]).

osmosis and electrodialysis) powered by renewable energy sources (such as wind and solar energy). A framework was presented for developing a decision procedure that monitors water shortage problems and identifies the availability of renewable energy resources to power desalination plants. The cost of alternative solutions, taking into account energy costs or profits by energy selling to a grid, was estimated. Emphasis was given to the market forces and the relationships among technology prices and market potential. The algorithm of the financial analysis (i.e., decision procedure) employed by Voivontas et al. [47] to estimate the water

production cost is shown in Fig. 7. While their study is based on the application of wind and solar energy to run a RO desalination plant rather than direct solar distillation, it does give a good description of the factors and decision steps that need to be taken into account (e.g., inflation rate and investment life) when estimating water production costs.

In a similar type of study, Hasnain and Alajlan [48] proposed the coupling of a photovoltaic (PV) powered water desalination reverse osmosis (RO) plant with a solar still plant. The aim was to reduce the overall water production costs by utilizing most of the reject brine from the PV-RO


plant in the solar still. They performed a cost analysis o f the solar still plant in order to determine the water production cost under various conditions. The estimated costs o f material required for the fabrication and installation o f their solar still are given in Table 1. In order to make an assessment o f the

cost effectiveness o f the proposed plant o f single-effect solar stills, a simple cost analysis was conducted. I f P is the capital cost o f the system and CRF is the capital recovery factor, the first annual cost o f the system (M) can be determined by the formula

M : P . C R F (14)

with

CRF - r (1 +r) n (15) (l+r)n-I

where r is the interest rate (%) o f lending banks and n the life o f the system (y). The salvage value o f the system was considered as 50% of the cost o f usable material (with the exception o f glass sheets where 25% of the cost was considered) which were saved even after the system life is over. I f S is the salvage value o f the system, the first annual salvage value (N) was determined as

N = S ' S F F (16)

with

S F F - (1 +r)n-1

(17)

Table 1 Material cost breakdown of a proposed solar still with a capacity of 5.8 m3"ff ~ of desalinated water (adapted from Hasnain and Alajlan [48])

Items Dimensions Estimated cost (in $US)"

Aluminum channels Aluminum angles Aluminum Z sections Aluminum T sections Aluminum sheet Glass sheets

PVC pipes (Schedule 80) Mild steel pipes (GI) Pipe fittings

35 mmx35×mmxl.7 mm

35 mmx35 mmxl.7 mm

35 mmx35 mmxl.7 mm

25 minx25 mmxl.7 mm

0.71 mmx0.23 mmx2 mm

1.43 mm×0.74 mmx2 mm 0.74 mmx0.23 mmx4 mm 2.54 cm diameter 3.8 cm diameter 2.54 cm diameter 3.8 cm diameter PVC/GI of different diameters

Silicon sealant - - Concrete foundation Storage tank - - Miscellaneous - - Total system - - cost Salvage value of the system

(50% cost of usable items)

1,314

1,795

133

2,267

2,432

15,090 4,907

587 1,229

125 745

2,667

267 2,667

3,467 13,335 53,029

12,344

"Assumes an exchange rate of 3.75 SR per US $1.

Since every system requires some maintenance, the annual maintenance cost was also considered. For their system, Hasnain and Alajlan [48] stated that maintenance is required frequently due to the following reasons:

continuous water supply into the stills, replacement o f broken glass during maintenance, and cleaning o f solar stills. Keeping this in mind, the annual maintenance cost (AMC) was taken at

M.F.A. Goosen et al. /Desalination I29 (2000) 63-89 85

10% of the first annual cost (M). Therefore, the actual annual cost (AC) of the system was given by

A C = M + A M C - N (18)

I f the annual yield of the system is taken as Y, then the product cost per liter (PC) is given by

P C - A C y (19)

The yield per dollar (y) can be established as follows:

Y Y - A C (20)

Table 2 predicts the effect of different parameters such as the life o f the system, interest rate, and annual maintenance cost on the cost of distilled water obtained from the proposed distillation plant. The total estimated cost of the pilot solar still was US $890,784 including 12%

labor charges without land values. I f one assumes that the solar still has a lifetime of about 10y (with a 5% interest rate), the overall cost for the case in Table 2 would be about 5.75 $'m 3 of fresh water or 0.084 $.L 1. I f the still 's lifetime is 20 y, then the overall cost of fresh water would be 2.99 $.m -3 at an interest rate of 5%.

Singh and Tiwari [49] used an annualized life-cycle costing method for the economic evaluation of various designs of solar stills. The cost of distilled water and the payback period were evaluated on the basis of monthly performance. It was observed that the cost of distilled water per unit area was most economical for a multi-wick double-effect distillation unit due to a low water depth in the basin and the re-utilization of latent heat of vaporization. Furthermore, the payback period was also reduced. The economic viability of various solar stills studied included a conventional single-basin solar still (CSBSS), a conventional double-basin solar still (CDBSS), a multi-wick single-effect solar still (MSESS), a multi-wick double-effect solar still (MDESS), an active single-basin solar still (ASBSS), and an active double-basin solar still (ADBSS) (Table 3).

Table 2 Effect of different parameters on the cost" of distilled water obtained from a proposed solar still plant coupled to a photovoltaic powered RO desalination plant (adapted from Hasnain and Alajlan [48])

No. Year Rate of Annual maintenance Capital recovery Annual Product water Annual cost, Yield interest cost, US$ factor cost, US$ cost, US$/m 3 US$/m 2 L/US$

l 10 0.05 7601 0.128 9336 5.60 6.44 174 2 10 0.08 8849 0.149 8882 5.33 6.13 183 3 15 0.03 5701 0.090 5703 3.47 3.94 285 4 15 0.08 6948 0.117 7186 4.27 4.96 226 5 20 0.05 4751 0.08 4856 2.93 3.35 334 6 20 0.08 6057 0.120 6391 4.00 4.41 254

aExchange rate 3.75 SR per US $1 (1999).

86 M.F.A. Goosen et al. / Desalination 129 (2000) 63 89

Table 3 Cost of various solar stills and other parameters (adapted from Singh and Yiwari [49])

Still no. Solar still type Initial investment, I n t e r e s t Maintenance, Life, y Increase in US$ rate, % US$ maintenance, %

! CSBSS 800 ! 3 345 19 I 0 2 CDBSS 1067 13 345 ! 9 10 3 MSESS 800 13 345 19 l0 4 MDESS 1067 13 345 19 10 5 ASBSS 2800 13 345 19 ! 0 6 ADBSS 3067 13 345 19 10

The annualized life cycle costing method was used for the economic analysis of all the stills mentioned. This method turns the difference of present value into a series of equal payments throughout the life of the system. The various parameters including subsidy, salvage value, initial maintenance cost, increase in maintenance cost, and inflation rate were incorporated into the economic analysis. The annualized life-cycle costing was calculated by the following relationship:

C,.:C,.f,.+MZ.S,.F~+CjZ.SS,.Z. (21)

v : c R (1 +p)n~_(l +~)n' (p-r) 0 +r).2

(24)

where r is the rate of interest (%), R is the initial maintenance cost (i.e., maintenance cost at the end of the first year in Rs), and p is the rate of increase in maintenance cost (%). The optimum life of replacement (C,) in years is determined using the following iterative method:

r(n_l) < ( I / R ) + [1 - ( rv ) (~ I)/(1 -rv)] <r ~

(1-tv ~)/[t (1-v)] (25)

where CT is the annualized life cycle cost (Rs), C, is the initial investment (Rs), f is the capital recovery factor, M~ is the maintenance cost (Rs), S, is the scrap value (Rs), F, is the sinking fund factor, C s is the fixed cost including installation and electricity (Rs), and sS~ is the subsidy (Rs). The parameters f , F~, and M,. are determined as follows:

Jr(1 +r)"] f" = [(1 +r)"-l]

where r = l+p, v = ( l+r ) - 1, and t is the number of times the capital cost (C,) is being increased in n years. The parameter C, should be less than the average cost in the years (n+l) and (n- 1). The payback time was considered to be the time at which the first cost and annual expenses with compounded interest equal the total savings of energy costs with compounded interest. It was defined as

F =f,. - r (23)

22, ,nf <l lnI lC n ' = L (r-e) J [ - ~ (26)

In[ (l +r~] (1 +p)j


where E is the energy saved per year (Rs), e is the energy inflation rate (%), and n' is the payback period (y).

The optimum life for replacement of these solar stills came out to be 19y using Eq. (25) and the data given in Table 3. All the stills were considered to be operational for 11 months in a year. On the basis of numerical computations, it was concluded that the multi-wick double-effect solar still is the most economical design, irre- spective of extreme cold/hot weather, due to a high output for the least heat capacity of water in the wicks.

To summarize this section, solar still plants have a mean lifetime of about 20y while the cost of fresh water produced by solar plants ranges from 0.52 to 2.99 $US'm -3, depending on the plant and the cost analysis method. Various equations can be employed to determine the cost of distilled water, and renewable energy sources can be coupled to desalination systems to make the process more economical. Also, it appears that a decision-making algorithm can be used to estimate water production costs as well as for monitoring water shortages. Finally, it is important to realize that maximum output does not mean that a solar still is the most economical. Current research shows that a multi-wick double-effect solar still is the most economical design.

6. Challenges and opportunities

In order to more effectively design and build a solar desalination plant, it is important to know the challenges and opportunities in using solar energy. By understanding the source of these challenges, more intelligent decisions can be made. The challenges fall into four categories: climate, solar technology, cost, and scale. The climatic factors, for example, include the low- intensity or low-flux density of the surface of the earth, the intermittent nature of this energy, and

the necessity for back-up systems during periods of extreme weather. Large collection areas are necessary for large-scale plants. The larger the scale, the more expensive and the more difficult it becomes to integrate the system into the desalination plant. The amounts of solar radiation also vary geographically with maximum radiation around the equator between latitudes 25°N and 25°S. The presence of clouds and air pollution will increase the proportion of diffused radiation. The intermittent nature of solar energy occurs due to the rotation of the earth and cloud. For example, the daily total solar energy for the month of June in Salt Lake City, Utah, is 621 langleys (1 langley = 1 cal.cm-2"d -1) while it is as low as 146 langleys in December. This difference of a factor of four will affect the design and economics of a solar water purification plant. Extreme weather conditions, which are difficult to predict, must also be taken into account. All of these factors may tend to diminish the return on the investment of a solar desalination system and present challenges as well as opportunities in the design and building of more efficient solar desalination plants.

7. Conclusions

The design of many solar desalination systems is largely semi-empirical in nature. For this reason, we need to increase our fundamental understanding of the thermodynamic and the economic principles behind such systems. Theoretical analysis (i.e., modeling) of solar desalination systems is an effective tool for predicting system performance. The efficiency or productivity of single-basin solar stills, for instance, is very low even under optimum conditions. Multiple-effect solar desalination systems are more productive due to the reuse of latent heat of condensation. The increase in efficiency, though, must be balanced against the increase in capital and operating costs compared

88 M.F.A. Goosen et al. /Desalination 129 (2000) 63-89

to the simple single-basin still. The efficiency of multiple-effect solar stills can be increased, for example, by inclining the glass cover surface towards the sun, running the system in an upward-type mode, employing flat-plate collectors, and using a heat exchanger to transfer waste heat from local industry to the solar still. Relatively simple but effective changes can be made in solar still designs for improving water production. Removal of condensation on the glazing to increase solar radiation is effective as well as the addition of surfactants to reduce the surface tension of the saline water are but two examples. Case studies have shown that solar desalination plants can be effectively modeled and that these models can be employed to improve the efficiency of the desalination process.

Humidification-dehumidification solar distillation plants have been around for over a century. These systems use solar energy either directly or indirectly to produce fresh water by dehumidification of air. However, the concept of using the humidification-dehumidification process in combination with the growth of crops in a greenhouse system is relatively new. Lower operating costs in the form of alternative energy sources were found to be key factors in their economic viability.

Solar still plants have lifetime of about 20y, with the cost of fresh water production ranging from 0.52 to 2.99 US$'m 3. A multi-wick double- effect solar still plant was found to be the overall most economical design when both water production and costs were factored into the same equation. Maximum output for instance does not mean that a solar still is the most economic.

To be able to produce commercially viable solar desalination systems, both efficiency and economic criteria need to be satisfied. Optimum water production must be balanced against any increase in the capital and operating costs of the plant. Moreover, it is important to know the challenges of solar energy in order to more

effectively design and build a solar desalination plant. Intelligent decisions can be made by understanding the source of these challenges that include climate, solar technology, cost, and scale. The final cost of the water produced must, at worst, not be higher than the cost of fresh water produced by conventional means such as multi- stage flash or multiple-effect evaporators. In the case of humidification-dehumidification systems, the energy costs associated with the condensers and pump operation as well as the energy saving associated with coupling the system to waste heat energy sources may end up being crucial in developing a commercially viable system. All of this in turn is dependent on a thorough understanding of the thermodynamic efficiency, as well as the economics of building a plant at a specific location. Having a commercially viable solar desalination plant in North Africa does not necessarily mean that the same plant can be built in southern California. Improved mathematical models and user-friendly software may pave the way for the optimization and development of economically viable humidification~tehumidification systems. Finally, with the increase in pressure on the world's fresh water resources due to expanding industrialization and population growth, a niche can be found for solar desalination technology, particularly in arid climates. This is an opportunity that should not be missed.

Aeknowledgements

This study has been made possible partially through funding from the Middle East Desalina- tion Research Center (MEDRC) through contract number 97-A-005b.

References

U] M.F.A. Goosen and W.H. Shayya, in: Water Manage- ment, Purification and Conservation in Arid Cli- mates, Vol. 1, Water Management, M.F.A. Goosen

A/LF.A. Goosen et aL /Desalination 129 (2000) 63-89 89

and W.H. Shayya, eds., Technomic, Lancaster, PA, USA, 1999, pp.l~5.

[2] O.A. Hamed, E.I. Eisa and W.E. Abdalla, Desalination, 93 (1993) 563.

[3] A.M. Kumar and G.N. Tiwari, Desalination, 116 (1998) 1.

[4] H.S. Varol and A. Yazar, Inter. J. Energy Res., 20 (1996) 541.

[5] M.E.S. Fath, Desalination, 116 (1998)45. [6] S. Al-Hallaj, M.M. Farid and A.R. Tamimi, Desali-

nation, 120 (1998) 273. [7] J.T. Mahdi and B.E. Smith, Renewable Energy, 5(1)

(1994) 520. [8] A.A. Badran and M.A. Hamdan, Int. J. Solar Energy,

17(1995) 51. [9] A. Joyce, D. Loureiro, M. Collares Pereira and M.

Moreira, Renewable Energy, 5(1) (1994) 517. [10] R.R. AI-Hajri and L.P. A1-Misned, Renewable

Energy, 5(1) (1994) 524. [11] H.-M. Yeb, L.-W. Ten and L.-C. Chen, Energy, 10(6)

(1985) 683. [12] H.-M. Yeh and L.-C. Chert, Energy, 10(11) (1985)

1237. [13] A.K. Singh, Int. J. Solar Energy, 17 (1995) 101. [14] H. AI-Hussaini and I.K. Smith, Renewable Energy,

5(1) (1994) 532. [15] M.A. A1-Abbasi, A.A. AI-Karaghouli and A.N.

Minasian, Desalination, 86 (1992) 317. [16] K.K. Gopinathan, Solar Energy, 45(1) (1990) 19. [17] J.P. Chiou and M.M. E1-Naggar, Solar Energy, 35(5)

(1986) 471. [18] S. Aboul-Enein, A. EI-Sebaii and E. EI-Bialy,

Renewable Energy, 14(1-4) (1998) 299. [19] H.-M. Yeh and N.-T. Ma, Energy, 15(12) (1990)

1161. [20] H.-M. Yeh, S.-W. Tsai and N.-T. Ma, Energy, 13(2)

(1988) 115. [21] H.M. Yeh and L.C. Chen, Energy, 15(2) (1990) 123. [22] HM. Yeh and Z.F. Chen, Energy, 17(3) (1992) 269. [23] J.L. Fernandez. and Chargoy, Solar Energy, 44(4)

(1990) 215. [24] Y.P. Yadav and L.K. Jha, Energy, 14(10) (1989) 653. [25] A. Kumar, M. Singh and J.D. Anand, Energy, 14(10)

(1989) 643. [26] B. Bouchekima, B. Gros, R. Ouahefand M. Diboun,

Desalination, 116 (1998) 185. [27] G.N. Tiwari and M.A. Noor, Int. J. Solar Energy, 18

(1996) 147.

[28] A.N. Minasian, A.A. AI-Karaghouli, M. Hasan and A. Shakir, Solar Energy, 49(2) (1992) 107.

[29] A.A. Fatani and G.M. Zaki, Int. J. Solar Energy, 17 (1995) 27.

[30] C. Armenta-Deu, Renewable Energy, l 1(2) (1997) 197.

[31] P. Namprakai, J. Hirunlabh and Y. Kiantsiriroat, Renewable Energy, 1 I(2) (1997) 169.

[32] P. Gandhidasan, Appl. Energy, 34 (1989) 69. [33] A.M. EI-Nashar and A.A. Qamhiyeh, Solar Energy,

44(4) (1990) 183. [34] R.S. Adhikari, A. Kumar and G. D. Sootha, Solar

Energy, 54(5) (1995) 317. [35] G.N. Tiwari, and S.B. Sharma, Energy, 16(6) (1991)

951. [36] S.G. Talbert, J.A. Eiblingl G.O.G. Lof, C.-M. Wong

and E.N. Sieder, Manual on solar distillation of saline water; Research and Development Progress Report No. 546, US Department of the lnterior Contract

! No. 14-01-0001-1695; 1970. [37] E. Delyannis and V. Belessiotis, in: Water Manage-

ment, Purification and Conservation in Arid Cli- mates, Vol. 2, Water Purification, M.F.A. Goosen and W.H. Shayya, eds., Technomic Lancaster, PA, USA, 1999, pp. 277-296.

[38] A. Khalid, Desalination, 93 (1993) 587. [39] A.C. Paton and Davis, The seawater greenhouse for

arid lands. Proc., Mediterranean Conf. on Renewable Energy Sources for Water Production, Santorini, 1996.

[40] A. Farid and A.W. AI-Hajaj, Desalination, 106 (1996) 427.

[41] N.K. Nawayaseh, M.M. Farid, A. Omar, S. AI-Hallaj and A. Yamimi, Desalination, 109 (1997) 277.

[42] D. Hoffman, Desalination, 89 (1992) 115. [43] P. Glueckstern, Desalination, 101 (1995) 11. [44] A. Ophir andN. Nadar, Desalination, 40 (1982) 103. [45] E. Delyannis and A. Delyannis, Desalination, 52

(1985) 167. [46] B.W. Tleimat, Desalination, 44 (1983) 153. [47] D. Voivontas, K. Yannopoulos, K. Rados, A. Zervose

and D. Assimacopoulos, Desalination, 121 (1999) 159.

[48] S.M. Hasnain and S.A. Alajlan, Desalination, 116 (1998) 57.

[49] A.K. Singh and G.N. Tiwari, Int. J. Ambient Energy, 16(2) (1995) 89.

Documents

Thermodynamic and economic considerations in solar desalination