Theoretische und praktische Untersuchung einer mehrstufigen solarthermischen Kleinanlage zur Meer- und Brackwasserentsalzung

In many coastal regions of developing countries the employment of industrial desalination plants is not possible due to the high investment and operating costs. For these purposes there is a need of small and low-priced desalination plants. Among a multitude of existing desalination methods the multi-stage heat recovery seems to offer a large potential for energy conservation. However, this process has not yet reached efficencies as high as those of the MEH process or membrane destillation. Since there is no published, detailed investigation of this process, it is examined theoretically and experimentally in this work with the aim to lower energy consumption to such a degree that the process can be powered economically by solar supply. For optimization, the heat- and mass-transfer between the stages must be well understood. The mass transfer was first examined in detail in a single distillation chamber. It was found to be affected by different factors, such as evaporation and condensation temperature, salinity, surface properties of the condenser and the geometry of the distillation chamber. Losses by drip back from the condenser reduce the distillate yield. The mass transfer resistance can be described as a serial connection of at least four single resistances in each phase boundary and in the gas and liquid bulk. The mass transfer model described in this work includes all known resistances and permits the variation of the surface areas of evaporator and condenser. The thermal resistance of the water layers is determined by the convective resistance of the liquid phase. This resistance involves high production losses. Using seawater the production was reduced by 25%. Since constantly water of less concentration flows into the system, the resulting density gradient weakens free convection and reduces the heat flow. Except for the decrease of water and gas layer height no further factors were determined to optimize mass transfer. A thermodynamic model of destillation was implemented in the simulation environment MATLAB/SIMULINK taking into account all heat- and mass-flows in the system as well as thermal inertia and salt concentrations. The validation of the model with results of a multi-stage laboratory plant shows an error of less than 1% in stationary conditions. This is regarded as sufficiently exact for the aim of optimizing the process regarding constructional and thermodynamic aspects. While thermodynamic optimization covers the number of stages, refilling factor and recovery of sensitive heat, the constructional optimization takes concerns the back dripping losses and the production rate. An optimized design was found, allowing a simple construction of the plants. A desalination module with 8 stages, a surface area of 2 m x 0.5 m and a height of 0,5 m achieves a production rate of 8 kg/(m²h) and an energy recovery of GOR= 3,5. That means, that the power requirement is only 28,5% compared to a single stage desalination unit. With the recovery of sensitive heat a GOR value of 4,6 is achieved. Compared with the best results published up to now, this represents an efficiency increase of about 100%. For a field test in Pozo/Gran Canaria A pilot plant with 4 desalination modules (1 m² each), powered by solar flat plate and vacuum tube collectors was set up, achieving a daily output of 33 litres or 40 litres with sensitive heat recovery respectivly. This corresponds to 9 litres daily per square meter of collector area (at 8 kWh/m²d). With the validated, thermodynamic simulation models including the solar collectors and the thermosiphon circuit it is now possible to predict the distillate production at any location on the basis of meteorological data. The produced water quality exceeds drinking water standards. However pitting corrosion was problematic. According to the validated simulation model further improvements are possible. Reduced back dripping losses, the use of the lowest stage for distillate production, sensitive heat recovery and improved insulation can raise the daily output to 19,5 litres per square meter collector area when using low inertia heatpipe vacuum tubes. Partial improvements were already achieved in practice with a capillary desalination module, using mainly plastics. This way corrosion is avoided and drip-back losses are reduced. With this new technology larger surfaces can achieve higher production rates at lower costs. A water price of 10 €/m³ seems to be possible. In review it can be stated that drip-back losses and the formation of stable density gradients in seawater operation represent systems-inherent and weak points which are difficult to avoid. Nevertheless, with further optimization measures similar water costs as with MEH or membrane distillation could be achieved with the advantage of simple construction and maintenance and no auxiliary energy for pumps is needed