Core flood experiments are an integral part of the methodology for chemical selection and the optimisation of scale inhibitor treatments, providing information on formation damage, inhibitor return profile and the dynamic retention isotherm under post flush conditions (as a function of SI, brine, pH, rock matrix, T etc.). In the absence of formation damage, the comparative inhibitor returns are often used to select the chemical prior to field treatment. However, it is recognised that there can be significant discrepancies between the core data and field data, depending on how the tests are conducted and how the data are interpreted.
In this paper, we describe the factors causing differences between the field and the core flood return. This includes inherent petrophysical differences between the small core plug and the larger near wellbore formation and differences in the saturation levels (e.g. core flood return typically at Sor) etc.. The paper also demonstrates that when considered appropriately, these differences should not impact the relative return profiles for different products.
Other, more significant aspects however, often relate to artifacts of the core test design, e.g. oversaturation with pre-flush and main treatment solutions of which multiple pore volumes are commonly applied in core tests. This work will demonstrate that such aspects can have significant consequences for the comparative inhibitor returns and moreso that these can have considerably different impacts (order of magnitude) for different chemical treatments, especially when highly retentive products are being assessed. The result being that for ill suited test protocols, selection of less effective chemicals and poor initial treatment design can result.
This paper presents examples of a series of core flood data in which the return profiles and derived isotherms show excellent agreement with field returns together with other examples were this is far from the case. We identify certain artifacts in commonly conducted core flood designs that should be avoided and recommend approaches and test protocols which allow more appropriate product ranking as well as modified modelling approaches which allow improved simulation from core to field..