Summary Partitioning tracer tests are used to determine residual oil saturation (ROS). At Koninklijke/Shell Laboratorium, Amsterdam, laboratory equipment based on the flow injection analysis (FIA) method has been constructed for rapid (within 2 hours) and accurate (better than 8%) determination of the equilibrium partition coefficient of tracers between crude oil and brine under simulated reservoir conditions. The apparatus makes possible investigation of the influence of all relevant parameters on the partition coefficient: tracer type and concentration, crude oil type, pressure (up to 34 MPa [4,930 psi]), GOR'S, temperature (up to 150 deg. C [302 deg. F]), and brine psi]), GOR'S, temperature (up to 150 deg. C [302 deg. F]), and brine salinity. The FIA equipment has been critically evaluated with model compounds. Its versatility has been verified with experimental results obtained on dead and live crudes. Introduction Economic incentives for optimizing reservoir production have focused attention on the development of methods to obtain information on the internal structure and associated fluid saturations of the reservoir. The application of tracers, in combination with advanced reservoir analysis techniques, provides a successful method of generating this information at every stage (waterdrive and tertiary recovery) of the reservoir development without costly interruption of normal operations. In a special application of tracer techniques, partitioning of the tracer between the different liquids present in partitioning of the tracer between the different liquids present in the formation can be used to obtain information on the reservoir. In particular, equilibrium partitioning of a tracer between the oil and aqueous brine phases may yield information on the amount of residual oil present in the watered-out reservoir. In the interwell tracer test (IWTT), a solution of partitioning tracer is pumped downhole, displaced with the brine, and detected by sampling a producing well. The tracer zone migrates in the reservoir at a velocity less than that of the brine because the tracer molecules dissolve partially into the stationary phase. The retardation of the tracer zone is directly connected with the amount of oil present; i.e., measurement of the residence time yields the required present; i.e., measurement of the residence time yields the required information on the ROS, Sor, as given by (1) where the partition coefficient K= C /C, C and C, =concentrations at equilibrium of the tracer in oil and water phases, and = retardation or capacity factor given by the ratio of the amounts of tracer in the oil and brine phases. In practice, can be obtained experimentally as the relative differ-ence in residence times in the reservoir for the retarded partitioning tracer, t, and a nonpartitioning tracer, t, flowing along partitioning tracer, t, and a nonpartitioning tracer, t, flowing along the same paths: (2) It can be seen from Eq. 1 that accurate determination of the partition coefficient, K, of the tracers involved in the tests is a keep partition coefficient, K, of the tracers involved in the tests is a keep step for accurate determination of ROS in watered-out reservoirs. K-Value Determination A useful source of K values obtained by manual procedures in model solvent/water systems is Ref. 4. Data on real crude/brine systems are scarce. A well-known method for determining the K value is the shake-flask technique, in which oil and brine are brought into contact by mechanical shaking and stirring. The technique is not very suitable, however, because equilibrium is reached only after several days and oil and brine are less likely to form (stable) emulsions. In the dynamic method applied to evaluate the popular singlewell tracer test (SWTT), brine with a known initial concentration of tracer is continuously circulated through a cell filled with oil until equilibrium is established. Reliable K values can be obtained within a reasonable time. At the relatively high concentration of chemical tracers applied in the SWTT, the partition coefficient is dependent on the concentration. The K-value/concentration relation is measured beforehand and incorporated into the interpretation of the SWTT. To circumvent this complicating factor, the partition tracer test should be carried out at a very low tracer concentration (e.g., with radioactively labeled compounds) where is no longer concentration dependent. To extend the measurements of K values to very low tracer concentrations, we selected FIA as the instrumental approach. Witthe FIA apparatus described, all relevant parameters influencing the partition coefficient in either dead or live crude oil systems can be readily investigated, as shown in this paper.

No abstract