TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractA specific area under water injection in the Carmópolis field, Brazil, is been considered a candidate area for a polymer pilot project for mobility control. A reservoir characterization and an evaluation of the polymer performance in this high heterogeneous reservoir were required. For this purpose, radioactive, fluorescent and chemical tracers were applied associated with polymer in a reduced area. The tracer technology has an enormous potential use in Petrobras scenario and this Carmópolis field application was an opportunity to obtain know-how. This paper describes the basic steps from the laboratory tests to the final application including design and programming of field operation. The interpretation of the results using a new approach is also addressed.

In the design of routine coreflood tests, it arises a number of limitations for physical similarity (capillary number, end effect) and for laboratory equipment restrictions (accuracy of measurements, pump capacities, etc.), which restrict the test plan. In the current paper these restrictions are presented in terms of dimensionless parameters. These limitations result in a system of inequalities for core length and displacement velocity. For the cases of waterflood and EOR, the sets of significant parameters have been determined. The code written in a simple spreadsheet allows to determine minimum core length which provides fulfillment of above mentioned conditions and to calculate the optimal displacement velocity. Introduction Relative phase permeabilities are the basic functions which determine the efficiency of the waterflooding. Hence, the accurate laboratory determination of these functions from the routine coreflood procedure is an important problem of the waterflood design and planning. Well developed and widely spread routine unsteady state method for determination the relative permeability assumes that capillary forces are neglected, i.e. so called conditions for large scale approximation are fulfilled. This could be achieved for long cores. The major limitation for this study is the core length, which usually does not exceed 13–20 cm. The problem of designing the routine coreflood test is to find out whether the conditions of classical methods are fulfilled for a given core, and to find the displacement velocity which provides the minimum error. Number of works propose different methods for determination of the relative permeability honouring capillary pressure. The approximate methods are reliable for limited intervals of rock and fluid parameters. Optimisation methods lead to ill-posed inverse problems. Therefore, the JBN method is still used routinely. In the current paper we concentrate on conditions of validity for routine JBN method and on parameters for laboratory test which provide this validity. Criteria for physical similarity for waterflooding (limited capillary number, minimum capillary-viscous ratio and delay number) were used in the work of Bedrikovetsky et alii for planning the coreflood test based on Barenblatt's non-equilibrium theory. Alternate criteria for large scale approximation are minimum sizes of stabilized capillary zone and of end effect zone which also provide minimum core length and optimal velocity. Nevertheless, during routine tests arise operational limitations like accuracy of measurements, capacity of pumps, etc. which are to be considered with the test design. In the current work we concentrate on design of the coreflood honouring both physical similarity criteria and operational limitations. Simple software in a spreadsheet is developed to determine minimum core length and optimal displacement velocity for waterflood and EOR. ROUTINE WATERFLOOD TEST Formulation of the Problem of Lab Tests. The measurements during the routine waterflood test are (Fig. 1):total flux;pressure drop on the core;water cut at the outlet. The parameters to determine are:fractional flow function;relative permeabilities. Optimal coreflood test plan. We would like to determine an optimal displacement velocity U for a given core length L. Also, the analysis developed aims to find out whether conditions of physical similarity and of measurement accuracy are fulfilled under given conditions of the laboratory test, mainly for a given core length. So, it is not an optimisation problem in the classical formulation of minimisation, the work aims to determine admissible area on the plane (U,L) for a given core and oil.