Abstract The production of silica in thermal petroleum recovery projects is a well-known phenomenon, and considerable efforts for its control are a common feature of facilities engineering in such projects. Recent work related to the generation of silica in SAGD projects has shown that the largest effect on silica production in SAGD is the steam zone pressure. Silica levels in produced water tend to increase with temperature. However, a significant suppressive effect of silica dissolution from quartz sands is provided by the presence of carbon dioxide dissolved in steam condensate. Lesser effects of silica concentration are due to ionic strength of the condensate. The carbon dioxide concentration in the steam condensate is amenable to theoretical prediction, available from recent progress in gas dissolution thermodynamics, and this allows the estimation of one of the major factors responsible for the suppression of produced water silica concentration. Other effects may be due to pH and fluid alkalinity, which are not independent of CO2 concentration. This paper is intended to provide an overview of the progress in obtaining a predictive capability, and to highlight the issues related to appropriate sampling and analytical methods that hinder simple correlations to date. Introduction The effect of temperature on silica dissolution from quartz has been known for a long time. Cowan and Weintritt(1) in their well-known treatise on water-formed scale deposits, cite Kennedy's work(2) and provide convenient tabulations and graphs of silica solubility and temperature. The experimental data for the solubility of silica from quartz are shown in Figure 1. By comparison, the data generated by the SOLMINEQ program of the Alberta Research Council are quite similar (Figure 2), with slightly increased solubilities at lower temperatures, and a slight reduction at higher temperatures. The SOLMINEQ program is a program based on the solution thermodynamic database of the U.S. National Bureau of Standards (now NIST), and operates by a free energy minimization algorithm. Data such as these usually make it mandatory that silica concentrations be reduced in produced water treatment in SAGD schemes, whether the water is recycled for steam generation or injected into a disposal well. Common methods in use are warm or hot lime softeners, MagOx units or acidification schemes. One would expect that apart from temperature, some aspects of produced water chemistry, such as dissolved acid gas content, pH, alkalinity and ionic strength, play a part in controlling the silica content of produced water from SAGD operations, so that the simple solubility curves published by Cowan and Weintritt cannot be used to correctly predict silica concentrations in any given scheme. Temperature Effect The magnitude of the temperature effect has already been shown. The use of the SOLMINEQ program produces results consistent with original experimental data. The use of this program to FIGURE 1: Experimental quartz solubility(2). Available in Full Paper. FIGURE 2: Equilibrium silica, mg/L, as function of steam zone temperature. Available in Full Paper. generate silica solubilities for any given produced water composition is a simple trial and error procedure, as follows: