Estimation of micromixing parameters from tracer concentration fluctuation measurements

Abstract

A method of micromixing parameter estimation is proposed which is convenient and applicable to a wide variety of models. One conducts a typical tracer reponse test, except that in addition to the (time-averaged) mean tracer response, the variance of the tracer response is also recorded as a function of time. Micromixing parameters for a given model are determined by equating the predicted and measured values of the time response of the tracer concentration variance. The predicted tracer concentration fluctuation response has been determined for both step and pulse tracer test, for the IEM model, the coalescence--redispersion model, and the two-, three- and four-environment models. Parameter estimates were calculated by minimizing the mismatch between predicted and measured concentration fluctuation responses. Preliminary numerical results indicate that the method provides satisfactory parameters estimates, even from moderately noisy measurements. These results also show that statistical analysis of the fit to measured response data can provide discrimination between competing mixing models.