Interpretation of Skin Effect from Pressure Transient Tests in Horizontal Wells

Abstract

Abstract Most horizontal wells have non-uniform distribution of skin along their lengths and this creates a challenging problem in the interpretation of their pressure-transient responses. The theory indicates that the rigorous incorporation of non-uniform skin distribution into horizontal well pressure-transient models requires the knowledge of not only the skin distribution but also the flow rate distribution into the horizontal well from the reservoir. Because this information is not normally available to the analyst, standard pressure interpretation techniques and tools assume uniform distribution of skin with the expectation that the estimates would correspond to some average of the skin distribution. The question that has not been adequately addressed in the literature is the physical meaning of the skin estimates from different pressure-transient analysis tools in common use. Because this question has not been adequately addressed, purely geometrical interpretations of the skin estimates have been proposed to calculate horizontal well productivities and develop flow models. In this paper, we generate synthetic pressure-transient responses for different non-uniform skin distributions along a horizontal well and analyze these responses by using the conventional tools that assume uniform distribution of skin. Skin estimates from well-test interpretation are then compared with the known skin distributions. The findings of this study are practical and important. First, the pressure drop caused by skin depends on the flow regimes if the skin distribution is non-uniform. Because the models used in commercial software assume the same additional pressure drop due to skin, the regression analysis can only match one of the flow periods for a constant skin value. To interpret the meaning of this skin estimate, we used the semi-log analysis techniques and demonstrated what type of average the estimated skin represents for different flow regimes and different skin distributions. For most cases, the estimates of skin from early-time radial flow analysis represent the arithmetic average of the skin distribution which may be useful for stimulation decisions. The skin estimate from the pseudo-radial flow period corresponds to the skin pressure drop at the heel of the horizontal well, which represents the additional pressure drop to be considered in the productivity calculations. We demonstrate that the geometric interpretation of the non-uniform skin effect proposed in the literature is inaccurate and leads to significant errors in the calculation of horizontal well productivity. Horizontal Well Skin Factor Van Everdingen[1] and Hurst[2] defined the concept of skin for vertical wells which quantifies the severity of the formation damage. The relation between the skin factor and the permeability and radius of the damaged zone is given by Hawkins' relation.[3] In the literature, many researchers have demonstrated that formation damage varies from the heel to the toe in especially long horizontal wells[4–14]. In this section, we present the derivations of horizontal well skin factor following Ozkan and Raghavan.[15,16] The horizontal well model shown in Fig. 1 was used in the derivations of the horizontal well skin factor. The horizontal well is located at an elevation, Zw, with respect to the bottom boundary and has a length, Lh, along the x-axis. The radial distance in the vertical, y-z plane is defined as -. The skin zone is assumed to be concentric with the horizontal well and has a radius, - , that is a function of the location along the horizontal well. The effective permeability in the y-z plane is defined as kr while the effective permeability of the skin zone is defined as ks. The flow within the skin region as defined above is assumed to be normal to the well axis (i.e. in the r-direction). This is because the radius of the skin zone is small (thin skin). If the storage capacity of the skin zone is negligibly small as implied by the thin skin assumption, then the fluxes entering and leaving the skin zone are identical (that is seeformula). Since the flux is a function of the properties of the skin zone, qh represent the flux when there is no skin and qhs represents the flux when there is skin.