The effect of pressure and rock type on optimal dual energy levels and minimalisation of saturation uncertainties for threephase core experiments was investigated. X-ray scanning of a composite core comprising of two chalk plugs and one Berea sandstone core was done. Reference scans using CO2, natural seawater, n-CO[10] and aCO2/n-C[10] mixture at four pressure stages and five different pre-defined dual energy levels were acquired. Two unique typical given saturations were input together with data from the reference scans in a routine for back-calculating three-phase intensity data. A detailed uncertainty analysis using these synthetic three-phase data and combining low and high energies was performed. The resulting combination of low and high energy level giving the lowest uncertainty was selected as a given setting for all dynamic scans. Saturated with live oil, the core was waterflooded while dynamic X-ray scans were acquired. After the waterflood, the core was depleted below the bubble point to atmospheric pressure while scans were acquired continuously. The results show that the rock type largely effects the selection of energy levels for both two- and three-phase X-ray scans. To minimise the uncertainty, the chalks generally required a higher low energy level independent of pressure than the sandstone. Also, at high pressures and for all rocks the optimal high energy level was found in the same range. At lower pressures, a lower high energy level was favourable. Introduction After X-ray computerised tomography (CT) was introduced to medical radiology, X-ray measurements of rock material has gradually become an important tool within many branches of petroleum engineering.[1–6] Today, many of the worlds reservoirs are declining, and the need for accurately assessing a reservoir is therefore even more important than before.[7] Despite this, for laboratory work reported in the petroleum literature, measurement uncertainties are often neglected.[8] Independent of technique, measurements of saturations inside rock are in practise associated with considerable uncertainties. Identifying and quantifying uncertainties are valuable when interpretingand applying laboratory data to a field production scenario. Adding this information can contribute to improved reservoir assessment, economical analysis and, thus, improved oil recovery. In practise, to calculate three-phase saturations from X-ray scanning of rock material, at least one of the liquid fluids are doped, i.e. a chemical that increases the density of the fluid substantially is added.[9] Altering the density will also alter the fluid viscosity and the interfacial tension between fluids.[10] Thus, important fluid/rock interactions will be described unrealistically. In future, one should preferably conduct experiments using reservoir fluids, even when the saturation calculation is based on in-situ measurements of attenuation of a photon beam. To achieve this goal, the need for a well-designed experimental setup is decisive. Although it should be of interest to the experimental reservoir engineer, guidelines for selecting appropriate dual energy levels for rock scanning are, to our knowledge, not established.[11] However, from any experimental setup to the next there exist substantial differences: Carbonate rock is composed of different species than sandstone, heavy oil consist of bigger molecules than simple hydrocarbons not to mention that one coreholder will scatter radiation more or less than others. Appropriate energy levels are not trivial to find and we assume them to be dependent on the particular set of fluid, rock and experimental apparatus used. This paper provides a scope for designing saturation measurements in terms of finding suitable dual energies for different rock type at different saturations and pressures, by using an X-ray scanner.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractThe effect of pressure and rock type on optimal dual energy levels and minimalisation of saturation uncertainties for threephase core experiments was investigated.X-ray scanning of a composite core comprising of two chalk plugs and one Berea sandstone core was done. Reference scans using CO , natural seawater, n-C¡ £ ¢ and a CO /n-C¡ ¤ ¢ mixture at four pressure stages and five different pre-defined dual energy levels were acquired. Two unique typical given saturations were input together with data from the reference scans in a routine for back-calculating three-phase intensity data. A detailed uncertainty analysis using these synthetic three-phase data and combining low and high energies was performed. The resulting combination of low and high energy level giving the lowest uncertainty was selected as a given setting for all dynamic scans. Saturated with live oil, the core was waterflooded while dynamic X-ray scans were acquired. After the waterflood, the core was depleted below the bubble point to atmospheric pressure while scans were acquired continuously.The results show that the rock type largely effects the selection of energy levels for both two-and three-phase X-ray scans. To minimise the uncertainty, the chalks generally required a higher low energy level independent of pressure than the sandstone. Also, at high pressures and for all rocks the optimal high energy level was found in the same range. At lower pressures, a lower high energy level was favourable.