This paper presents horizontal well test design and interpretation methods. New analytical solutions are developed, which can be easily handled by a desktop computer, to carry out design as well as interpretation using semi-log and log-log analysis. These analytical solutions point out a distinctive behavior of horizontal wells which is :- At early time, a circular radial flow in avertical plane perpendicular to the well.- At late time, a horizontal pseudo-radial flow. Each type of flow is associated with a semi-log straight line to which semi-log analysis has the adapted. The horizontal pseudo-radial flow takes into account a pseudo-skin depending on system geometry which is a priori defined and estimate Practical time criteria are proposed to determine Practical time criteria are proposed to determine the beginning and the end of each type of flow and provide a guide to semi-log analysis and well test design. We study the behavior of uniform flux or infinite conductivity horizontal wells, with wellbore storage and skin. The homogeneous reservoir is infinite or limited by impermeable or constant pressure boundaries. pressure boundaries. A method is also outlined to transform all our solutions for homogeneous reservoirs into corresponding solutions for double porosity reservoirs. 1 - Introduction The first horizontal producing wells yielded extremely positive results These results made it necessary to study flow around a horizontal well with a view to practical applications : test design and interpretation. The purpose of this study is to provide, with a horizontal well configuration a way :- to decide upon whether a test has to be made ornot : will the test help to obtain the requiredinformation ?- to optimize test time, if necessary : test mustbe long enough to be interpreted.- to interpret the test performed usingapplicable methods. The simplest and most significant analytical approach is first described in this paper. It relates to the transient behavior of a well with no wellbore storage C and no skin S, with a uniform flow (flow per unit length is the same everywhere in the well). Then the wellbore storage, the skin and the effect of the boundaries of the reservoir are taken into account, leading to a practical approach applicable to most real cases. The transposition of these solutions to infinite conductivity wells is described.

In a channelized environment, the overbank facies (levees, crevasse-splays) have poorer reservoir qualities than the channel itself. In this type of environment, the geometry of this two-feature complex, together with its permeability contrast, influences the pressure transient response. The pressure behavior of this kind of reservoir has been computed for well-test analysis purposes.Our geometric model comprises a main channel bounded laterally with finite or infinite width levees. It is conceptualized as a composite linear strip reservoir with different mobilities and diffusivities. The length of the whole system also can be either finite or infinite. The solution uses a time-space Laplace transform and a spatial Fourier transform already used by Ambastha, 1 Butler, 2 and Kuchuk. 3 Moreover, our model adds lateral limits by no-flux boundary conditions and longitudinal limits by superposition of image wells or by using a finite Fourier transform.This paper exhibits specific features of the type curves generated for the model. The influence of the levees and of the permeability ratio between channel and levees on the derivative signature is discussed. Hints are given for diagnostic purposes and parameter estimation. Fieldcase buildup analyses show the applicability of the approach.

This paper compares the vertical permeability of several field cases, as obtained using different techniques. The vertical permeability of a reservoir is often of major importance: –where water or gas coning is likely to be a problem,–when pressure maintenance is carried out in multi-layer reservoirs,–in many reservoir simulation studies,–if horizontal wells are envisaged. The following methods were used to determine the vertical permeability: –well test analysis: partial penetration, multilayer and horizontal well tests in both conventional and fractured reservoirs,–coning analysis,–RFT vertical gradient and build-up analysis. A comparaison of the vertical permeability values obtained, using the various methods listed above, makes it possible to establish the interpretation consistency. A comparison with core measurements makes it possible to evaluate reservoir homogeneity. When vertical permeability is measured from cores it is generally higher in comparison to measurements obtained using dynamic methods. This is because core measurements do not allow for middle scale fluid barriers. Dynamic measurements of vertical permeability combined with statistical methods help determine the magnitude of fluid barriers causing permeability anisotropy in the reservoir. In fractured reservoirs, vertical permeability is often low as a result of discontinuity of the fractures at the bedding planes. Well tests and coning analysis are the best methods for determining vertical permeability and evaluating fracture continuity in the reservoir.