TO STUDY THE FACTORS WHICH PROMOTE SILICA DEPOSITION IN GEOTHERMAL PIPELINES, AND TO DEVELOP A MODEL FOR PREDICTING DEPOSITION RATES FROM FLUIDS OF KNOWN COMPOSITION.
An experimental study of horizontal air and water flow has been carried out in a new 50 mm internal diameter pipe loop. Continuous measurements of liquid film thickness were made using parallel wire conductance probes.
At high gas flow rates, the liquid distribution tends to have axial symmetry. However, at relatively low and moderate gas flow rates, the influence of gravity is very significant. This results in a thick liquid layer at the pipe bottom which has disturbance or roll waves travelling on it. Characteristic roll wave frequencies can be determined from the spectra of film thickness fluctuations at the pipe bottom.
A relatively simple theoretical model based on turbulent diffusion has been developed to predict the droplet concentration distribution in the pipe cross section and droplet deposition on the pipe wall. Comparisons with experimental data show that this model is reasonably accurate.

The formation of hard and tenacious scale in pipes and in equipment of geothermal plants constitutes the main obstacle in the economic exploitation of many geothermal fields. The scaling problem is encountered in almost every geothermal installation, but it can be more acute in plants handling high enthalpy brines.

Analyses of samples from the Milos 2MWe Geothermal Plant have been used to characterize the scale forming in various parts of the plant. Heavy metal sulphides are the dominant compounds of scale close to the point of primary fluid flashing, whereas silicates and silica tend to dominate at the other end of the plant (point of fliud reinjection). Laboratory experiments with a typical sulphide (lead sulphide) show a very strong effect of pH and concentration on the rate of scale formation. Similar results have been obtained in silica polymerization experiments.

The lead sulphide deposition at ambient conditions was strongly dependent on pH and total concentration. The same trend regarding pH and concentration was obtained from the silica polymerization experiments.
For a fixed lead sulphide concentration, appreciable deposition occurs in a limited range of pH values of about 2 pH units. The maximum deposition rate is observed at the pH value of complete dissolution of lead sulphide.

A practical implication of this work, provided that the same deposition features and trends in the results hold at high temperatures and high salinity fluids, is as follows: to prevent sulphide scale formation it seems necessary to reduce the pH by about 1 unit below the pH value at which practically all the scale forming heavy metal ions are in solution. The concentration of heavy metals in the Milos geothermal brine are compared with the solubility of the corresponding sulphides, at pH values 5.3 (the brine pH) and 4.3. The solubilities are calculated for a fluid with 2M salinity, at 230 C. The pH reduction is expected to have a beneficial effect on silica deposition as well.

A second implementation is that by increasing the pH of the brine, and possible adding a coagulant, the sulphides will tend to agglomerate thus facilitating their separation by mechancial means. It is very likely that silica will behave similarly in the same pH range (5.5 to 7.5). Obviously more work is necessary to confirm the above conclusions and to arrive at economically acceptable solutions for the scaling problem.
A HORIZONTAL TWO-PHASE FLOW FACILITY WILL BE CONSTRUCTED. THIS WILL BE USED INITIALLY TO DEFINE THE CONDITIONS UNDER WHICH ANNULAR FLOW IS OBSERVED. SUBSEQUENTLY, THE RATES OF SILICA DEPOSITION FROM SIMULATED GEOTHERMAL BRINES WILL BE STUDIED, AND TECHNIQUES DEVISED FOR CONTINUOUS MEASUREMENT OF DEPOSITION RATE. IF POSSIBLE, THESE OBSERVATIONS WILL BE CHECKED IN THE MILOS GEOTHERMAL FIELD. FINALLY, THE DATA WILL BE USED TO CONSTRUCT A MATHEMATICAL DESCRIPTION OF THE DEPOSITION PROCESS AS A FUNCTION OF FLOW CONDITIONS.