Single Well Chemical Tracer Tests (SWCTT) are widely used to estimate the remaining oil saturation after waterflooding (Sorw) and enhanced oil recovery including low salinity waterflooding. SWCTT have the advantage of being a more direct measure of Sorw than logs and unlike cores do not have the uncertainty of maintaining or restoring wettability and samples more reservoir rock. Interpretation of the test is often superficial, lacking a full description of physical processes involved and proper uncertainty analysis. This paper analyses SWCTTs from high and low salinity waterflooding, explains what is necessary to model the tests, demonstrates rigorous uncertainty analysis, and shows the value of designing tests with simulation.
Simulation procedures involve finding multiple history matches to the observed tracer production. Parameters affecting tracer production such as Sorw, mechanical dispersivity, hydrolysis reaction rate, number of layers and the amount of fluid injected and produced for each layer are adjusted to get the best fit. Nonstandard aspects of SWCTT must also be modeled. Because Sorw is sensitive to the ester partitioning coefficient which in turn is sensitive to salinity and temperature, injection of water with a lower temperature than the reservoir impacts interpretation. In one example, if the temperature variation is not accounted for the error in Sorw is 3 su.
Each SWCTT has its own characteristics which often appear to be different even in repeat tests on the same well. Apart from the common variables, other variables often must be introduced to get a good history match. In most cases, the shape of the tracer profiles cannot be satisfactorily matched with a single layer and layers with different levels of irreversibility must be added to match the shape. The well bore volume is sometimes varied. In repeat tests on the same well, constraining the interpretations to have the same layering and not increasing Sorw can reduce uncertainty and improve interpretation. In one case, a one-layer analysis shows no benefit but more detailed analysis with 3 layers indicates an incremental recovery of 12% PV, depending on the assumptions about well bore volume; the best fit overall indicated 2% PV.