SMALL-SCALE SYSTEMS FOR SOLAR-THERMAL DESALINATION
OF SEA AND BRACKISH WATER
Markus Spinnler,
Jürgen Blumenberg,
Wolfgang Moik,
Hendrik Müller-Holst,
Hans-Ulrich Krispler,
EUVT, Technical University of Munich1
EUVT, Technical University of Munich
Moik Industrial Components2
Division 4: Solar Thermal and Biomass, ZAE Bayern3
CASE GmbH4
ABSTRACT
Clean drinking water is one of the most important international health issues today: In
many places of the world, only brackish or polluted water is available. In W EST BENGAL and
BANGLADESH, villagers have been severely afflicted by drinking water from deep wells containing
arsenic. This fact leads to an increasing interest in new water purification technologies (desalination technologies respectively). The standard methods are reliable in the fresh water capacity
range of some 100 to 200.000 m3 per day. They are not used in decentralized regions due to their
permanent need of qualified maintenance, infrastructure and high costs. Here, the use of smallscale solar desalination systems is desirable and makes economic sense.
For this purpose, two systems were developed in Munich: (1) Together with CASE GmbH,
MUNICH, and the Austrian company MOIK, HALLEIN, the TECHNICAL UNIVERSITY OF MUNICH developed the AQUASOL single-stage-flash-evaporation unit. (2) The GOR multi-effect humidification
unit has been developed and optimized by the ZAE BAYERN and T.A.S. GmbH. The investigated
water purification plants are described in detail, operation principles, energy consumption ratios,
amount of maintenance and financial aspects are compared and summarized. Both working
groups started a cooperation to solve the problems in removing arsenic from contaminated well
water. Experiments on this topic showing satisfactory results are presented and a possible benefit
for the people of W EST BENGAL and BANGLADESH is discussed.
1
TU MÜNCHEN, LTD, Boltzmannstr. 15, D-85747 Garching, Germany, Tel.: + 49 – 89 – 289 – 16223
W OLFGANG MOIK INDUSTRIEPRODUKTE, Untersbergweg 14, A-5400 Hallein, Austria, Tel.: + 43 – 6245 – 7617 – 4
3
ZAE BAYERN, ABTEILUNG 4, Domagkstraße 11, D-80807 München, Germany, Tel.: + 49 – 89 – 35 62 50 – 17
4
CASE GMBH, Westendstraße 125, D-80339 München, Germany, Tel.: + 49 – 89 – 50 03 59 – 0
2
OVERVIEW ON DESALINATION PROCESSES
Mostly applied desalination processes are the Multi-Stage-Flash (MSF), the MultieffectDistillation (ME) with Thermal Vapor Compression (ME-TVC) or with Mechanical Vapor Compression (ME-MVC) and the reverse Osmosis Method (RO). By coupling them with solar collectors,
distillation, MSF and ME-plants can be driven by solar energy. Their specific energy consumption
varies between 40 and 70 kWh/m3distillate thermal and up to 4 kWh/m3distillate electrical energy. The
choice of any process depends on the salt concentration as well as on the process economics. As
main problems, high investment costs as well as an high amount of maintenance in these hightech plants can be mentioned. [1][2]
In the following, two small scale systems for thermal desalination of sea and brackish water
are described in detail. A major skill of the two plants is their extremely low need of maintenance
and their ability of autonomous decentralized stand-alone operation.
THE AQUASOL FLASH-EVAPORATION PLANT
The AQUASOL-project was realized by the companies CASE GmbH AND MOIK INDUSTRIAL
COMPONENTS in close cooperation with the TECHNICAL UNIVERSITY OF MUNICH (TUM), where optimization and testing is carried out. It was planned to develop a low-cost solar desalination plant for
decentralized use with a daily fresh water output of 100 l for application in remote areas with lack
of clear water. It should be possible to build up and maintain the plant easily on site.
Application possibilities are the desalination of brackish or sea water. Main goal is the supply of families and small farms in remote areas. The daily production should be enough to cover
their fresh water demand. Basic idea of the AQUASOL unit is the use of a single flash stage at ambient pressure. Further measures aim at enhancing the air humidification rate. The AQUASOL project was realized in three steps, including a feasibility study for the process in 1996, investigation
of a laboratory-scale pilot plant in 1997 and a prototype on the solar experimentation field of TUM
in 1998. For developing pilot plant and prototype, a numerical simulation program was developed.
Extensive experimental investigations for testing the plants and for validating the simulation program were carried out.
Seawater Desalination by flash evaporation
Figure 1 shows the operation principle of the AQUASOL desalination plant. The AQUASOL
plant is consisting of two parts: a pressurized section and an ambient pressure section. The pressurized section of the plant is fed by a sea-water pump, which brings feed water to a required total
pressure of 2 bars in order to avoid boiling while temperatures exceed 100°C. In the following,
feed water is preheated by two heat exchangers, brine heat exchanger and condenser, respectively. In a final step, feed water is heated up to 120°C by the solar collector system.
Passing through an expansion valve, the overheated water now flashes to the ambient
pressure section, where two physical effects support evaporation: flash evaporation as well as air
humidification. The two phases (1) steam with saturated air and (2) brine pass the separator,
where steam and saturated air are condensed by flowing in contact with the condenser and the
brine is fed back for preheating the feed water.
The pressurized section of the two pilot plants was realized with an excenter pump and a
pipe-assembly heat exchanger as condenser which was built in stainless steel 1.4539. In the pilot
plant, the feed water was heated up to a temperature of 120°C by means of an adjustable electrical heating system (total power: 18 kW), which was used for simulating different thermal loads.
Flash temperature is kept constant by varying the feed flow via the speed of pump.
Figure 1: Principle of the AQUASOL -Process
Due to the high material and manufacturing costs of sea-water resistant heat exchangers,
the brine heat exchanger was replaced by a mixing basin, where 50% of the heated brine are remixed and thus preheat the feed water. Expansion of heated sea or brackish water is realized by
special PTFE pressure relief valves of 2 bar tripping pressure.
In the flash chamber, made of ceramic-coated steel, a dashboard is separating the steam
from the liquid brine. The steam is rising to the condenser, whereas the brine is collected in the
brine chamber. Piping and containers in the ambient pressure section are made of polymers. A
schematic view of the laboratory plant can be seen in figure 2.
In 1998, experiences with the pilot plant were transferred to a prototype working with 6 m2
vacuum-tube solar collectors (STIEBEL ELTRON SOL 200 A), which were installed on a 1-axis
tracking device for seasonal tracking and were equipped with specially developed sea-water resistant heat collector heads. As back-up, an additional electrical heating system was installed.
Furthermore, thermal insulation of both flash chamber and piping was improved. Figure 3 shows a
photo of the prototype. Three collector modules with 2 m2 absorber area each can be seen as
well as the flash chamber and the excenter pump at the left hand side.
Fig. 2:
AQUASOL Pilot Plant
Fig. 3:
AQUASOL prototype at TUM
Experiments
On both plants, steady state experiments were carried out in order to achieve comparable
results for distillate output as a function of thermal input and flash temperature. Besides, day
simulations were carried out for getting information about the possible daily distillate output. As a
reference for day simulations, the 1st July at ALMERÍA (ANDALUCÍA, SPAIN) was fixed together with a
vacuum-tube collector area of 6 m2 (STIEBEL ELTRON SOL 200 A).
Figure 4 shows the results for steady state experiments. Efficiency is rising together
with higher solar irradiation and flash temperature. The specific thermal energy consumption
reaches approximately 1 kWh/ldistillate at high irradiation values of 1.000 W/m2. Testing the prototype, distillate output could be increased by 10%. Together with the new stainless heat collectors,
a total collector efficiency of 50% could be measured, what indicates a satisfying operation of the
modified collector system. Day simulations at ALMERÍA yield a daily production of 30 to 40 l/d.
Furthermore, theoretical results of numerical process simulation go in conformity with experimental data. Small deviations between theory and experiment are caused by steam and heat
losses, which are not being considered in the simulation. [2][3]
Fig. 4:
Distillation output at steady-state
Results and Conclusion
During research on the AQUASOL project, feasibility of the process was demonstrated, a pilot plant and an improved prototype were working successfully. A main problem of the AQUASOL
desalination process is its high energy consumption which could be lowered by improving air humidification processes in the flash chamber itself. Other crucial problems are the relatively complicated control mechanism and the need to use sea water resistant steel qualities. The price of
the plant has been fixed at 10.000 US$, where the majority of the costs result in using vacuum
tube collectors. Based on the technical protocols for solar collectors of TECHNIKUM RAPPERSWIL
[4], economic alternatives to these high-tech collectors have been developed and several commercial flat plate collectors are taken under consideration. An experimental validation in Munich is
planned. Another project aims at developing a low-tech sea-water resistant parabolic through
collector for temperatures of 120°C. A simulation program for collector design has been developed, the erection of first test collectors is planned for summer 2000.
THE GOR MULTIEFFECT HUMIDIFICATION PLANT
The GOR multi-effect humidification unit has been developed and optimized by ZAE
BAYERN and T.A.S. GmbH. It can be driven by low-grade thermal energy, either by use of solar
thermal energy, by wind or by supply of waste heat from generating sets or industrial plants. The
thermodynamic principle of multi effect humidification is based on the evaporation and condensation of water inside a thermally insulated box at ambient pressure. Vapor is transported from the
evaporator to the condenser, where the distilled water is collected. The process has been technically optimized, so that the air circulation is driven entirely by natural convection and high heat
recovery ratios can be realized. The GOR desalination system is also sized for small groups or
villages with a drinking water demand of up to several 1000 l/d.
A pilot system has been installed in FUERTEVENTURA (CANARY ISLANDS, SPAIN) and has
been measured and analyzed in detail by the ZAE BAYERN since 1992. In 1997, a desalination
system with 24 hour thermal storage system was erected in SFAX, TUNISIA.
Seawater Desalination by Multieffect Humidification
The principle of this desalination process is based on the evaporation of water and the
condensation of steam to and from humid air, that flows in a circuit driven by natural convection
between condenser and evaporator (clockwise in figure 5). Evaporator and condenser are located
in the same thermally insulated box.
In the evaporator, hot seawater is distributed onto vertically hanging fleeces made of polypropylene and is slowly trickling downwards. The air moves in countercurrent flow to the brine
through the evaporator and becomes saturated with humidity. Partial evaporation cools the brine
which leaves the evaporation unit at higher salt concentration and a temperature of approximately
45°C. The condenser unit is located opposite to the evaporator. Here, the saturated air condenses on a flat plate heat exchanger made of polypropylene. The distillate is flowing down the
plates and is trickling into a collecting basin. The heat of condensation is mainly transferred to the
cold sea water flowing upwards inside the flat plate heat exchanger. Thus, brine temperature in
the condenser rises from 40°C to approximately 75°C. In the next step, the brine is heated up to
the evaporator inlet temperature, which is between 80°C to 90°C by one of the heat sources
mentioned above.
In order to make the process competitive from economic point of view, 24 hours operation
of the desalination system is indispensable. Thus, if the system is driven solarthermally, a thermal
storage tank has to be implemented. The outstanding advantage of the natural air convection
loop inside the distillation unit is the fact that it is self-adjusting to the optimum operating state
regarding to the preset parameters. By changing these parameters – evaporator inlet temperature
and load volume flow – it is possible to influence and optimize the thermal efficiency of the system
for a determined solar input. A TRNSYS transient simulation tool for the distillation unit as well as
simulation tools for collector field and thermal storage have been developed by ZAE BAYERN. [5]
Experimental Investigation
The following considerations are based on actual laboratory investigations at ZAE BAYERN.
Here, a distillation module similar to the demonstration plant in FUERTEVENTURA, is installed.
Steady state operation modes with various load parameters have been investigated. The evaporator inlet temperature, the condenser inlet temperature and the volume flow of the brine were varied.
As can be seen in figure 6, an increasing evaporator inlet temperature at constant brine
load flow causes rising distillate volume flow, because the evaporation and condensation process
is much more efficient at high temperatures. It is obvious, that a higher distillate flow can be
achieved with increasing load flow, even when limited at flow rates higher than 600 l/h due to
overstraining the evaporation and condensation surfaces. The energetic process efficiency also
strongly depends on the load flow of the system. Higher load flow and distillate production need
higher temperature differences for constant transfer resistances. Hence, the heating demand
rises. The temperature dependence of the specific energy consumption is inverse proportional
due to an improvement of heat and mass transfer at higher temperatures. [5][6]
Figure 6 shows also, that the specific energy consumption of the process can reach about
100 kWh/m3distillate, what is one order of magnitude lower than at AQUASOL. The estimated distillate
costs which can be achieved with a load flow of 600 l/h at an evaporator inlet temperature of 85°C
are 20 US$/m3 using solar collectors as heat source. [5]
Figure 5:
Principle of the GOR-Process
Figure 6:
Specific energy demand for different operation
modes at 40°C condenser inlet temperature
ARSENIC CALAMITY IN BANGLADESH AND WEST BENGAL
In W EST BENGAL and BANGLADESH, tube-well groundwater has been found to be contaminated with arsenic, which occurs naturally in alluvial and deltaic sediments. It is containing high
arsenic levels, in some places several hundred times the level recommended by the W ORLD
HEALTH ORGANIZATION WHO for acceptable drinking water. The latest statistics available on the
arsenic contamination in groundwater indicates that around 85% of the total area of Bangladesh
and about 75 million people are at risk of severe health problems. It is estimated that at least 1.2
million people are exposed to arsenic poisoning with 24 millions potentially exposed. [7]
The introduction of deep-drilled tube wells some decades ago not only supplied farmers
with irrigation water throughout the dry season but also provided families and villages with a “reliable” source of drinking water. However, the massive and extended use of groundwater for agricultural irrigation led to a lowering of the watertable, resulting in arsenic dissolution in water. [8]
Today, as a maximum contaminant level for safe drinking water, WHO has agreed to lowering the recommended arsenic level from 0.05 to 0.01 mg/l or 10 PPB. In W EST BENGAL and
BANGLADESH, about a million people may be consuming water in excess of 0.5 mg/l or 500 PPB,
and wells have been found which contain as much as 10 mg/l or 10.000 PPB. This is 1.000 times
the recommended limit. In [7] and [8], detailed charts of arsenic pollution in Bangladeshi wells are
given.
With skilled staff, good monitoring equipment and a well-managed facility, arsenic removal
does not constitute a major problem in large-scale water treatment. Problems appear when trying
to accomplish facilities for small-scale or household purposes. Maintaining the facilities and
monitoring the results appears quite critical, even as the known treatment methods show a high
dependence on arsenic concentration. Due to these problems and due to a negligible steam
pressure of arsenic at ambient temperature, multi-effect humidification as a proven reliable desalination process needing low maintenance shows a good feasibility for removing arsenic from
contaminated water.
Experiments on Arsenic Removal
As can be seen in [8], a widespread arsenic pollution average of 1.000 PPB is reached.
This concentration was chosen for a feasibility study. For simulation of arsenic well water, pure
arsenic (III) oxide (99,996%) was dissolved in deionized water with a total electrical conductivity of
139 µS/cm.
The GOR ME 30 laboratory distillation plant at ZAE BAYERN is containing a total water volume of 150 l. In order to obtain a total feed water concentration of 1.000 PPB, 150 mg of arsenic
(III) oxide had to be dissolved. In order to improve arsenic solubility in pure water, the solution
was treated according to volumetric analysis standard methods [9] with NaOH and H2SO4. For
laboratory tests, feed water and the two resulting mass currents brine and distillate were recirculated while cooling the brine. Samples of 40 ml were drawn off the distillate flow.
Four different operating states were investigated by varying the evaporator inlet temperature and evaporator load flow. Feed water before and after the test series was analyzed. Arsenic
analysis was carried out by TECHNICAL UNIVERSITY OF MUNICH, LEHRSTUHL FÜR W ASSERGÜTE UND
ABFALLWIRTSCHAFT according to EN ISO 11, DEV D 18. Table 1 shows the experimental results:
Table 1: Results of Arsenic Analysis
Feed Water
Start
254 l/h
25°C → 80°C
256 l/h
25°C → 85°C
420 l/h
30°C → 85°C
Feed Water
End
Arsenic
Concentration
818 PPB
< 5 PPB
< 5 PPB
< 5 PPB
738 PPB
Electrical
Conductivity
160 µS/cm
1 µS/cm
1 µS/cm
1 µS/cm
160 µS/cm
Distillate Flow

15 l/h
18 l/h
26 l/h

The results for three different operation parameters show clearly, that the arsenic level can
be reduced to a level lower than the detection limit of 5 PPB arsenic. Within detection accuracy,
there seems to be no relation between feed water mass flow, evaporator inlet temperature and
arsenic concentration in the distillate. By this, the feasibility of arsenic removal with the GOR multieffect distillation process was proven. Further experiments investigating a possible dependence
of feed water and distillate arsenic concentration are on the way.
CONCLUSIONS
In the paper, a short overview on the worldwide water situation as well as on conventional
desalination technologies was given. Two small-scale, autonomous plants for stand-alone operation in developing countries were described in detail and experimental results were discussed.
The AQUASOL flash evaporation plant yields a daily production of 30-40 liters per day. The
specific energy consumption has been determined at 1 kWh/ldistillate. Two experimental plants
have been investigated and were working in a satisfactory way. One autonomous plant was
tested on the solar experimentation field of TUM in GARCHING. Main problems were the high energy consumption, the need of seawater resistant steel qualities due to high-temperatures and a
complicated controlling device. Further investigations have to aim at enhancing the air humidification rate for obtaining a lower specific energy consumption and at gaining mass manufacturing maturity.
With the GOR multieffect humidification plant, a remarkably lower specific energy consumption of 100 kWh/m3distillate can be realized. The GOR plant is nearly maintenance free and
has proven its reliability in two demonstration plants in FUERTEVENTURA (CANARY ISLANDS, SPAIN)
and with 24 hour thermal storage system in SFAX, TUNISIA. The estimated distillate costs using
solar collectors as heat source are 20 US$/m3.
Due to the W EST BENGAL and BANGLADESH tube-well groundwater contamination with arsenic, experiments on arsenic removal with the GOR desalination plant were carried out. Therefor,
the situation in W EST BENGAL and BANGLADESH was briefly discussed, and the experiments with a
GOR ME 30 laboratory plant were described in detail. In preliminary experiments, multieffect humidification was found to be a suitable process for removing arsenic from contaminated wellwater. The outstanding advantage of this process is the fact, that distillate arsenic concentration
is no function of feed water concentration. By this, monitoring of feed water and distillate is not
necessary and low maintenance of the plant is guaranteed.
REFERENCES
[1]
J. Ayoub et al., Water Requirements and Remote Arid Areas: The Need for Small-Scale
Desalination, Desalination, vol. 107, 1996
[2]
A. Walthes, Examination of a New Sea-Water Desalination System, Final Report
EUVT-TUM DA 97/12, 1997
[3]
L. Giagdzoglou, Examination of a Solar Sea-Water Desalination System, Final Report
EUVT-TUM DA 98/5, 1998
[4]
U. Frei, T. Häuselmann, F. Flückinger, R. Frey, Leistungsdaten thermischer Solarkollektoren, Technikum Rapperswil, Prüf- und Forschungsstelle Solarenergie, April 1994 xxx
[5]
H. Müller-Holst, M. Engelhardt, W. Schölkopf, Small Scale Thermal Sea Water Desalination Systems for Decentralized Use: Optimization of System Design by Simulation, Desalination vol. 122, pp. 255-262, 1999
[6]
H. Müller-Holst, M. Engelhardt, M. Herve and W. Schölkopf, Solarthermal Seawater Desalination Systems for Decentralised Use, 6th Arab International Solar Energy Conference,
Muscat, 1998
[7]
www.hvr.se
[8]
www.dainichi-consul.co.jp
[9]
G. Jander, K. Jahr, Maßanalyse, 14. Auflage, Walter de Gruyter, Berlin/New York, 1986