Multi-Phase Flow Well Test Analysis in Multi-Layer Reservoirs

Abstract

Abstract This paper investigates the performance of an oil well under multi-phase flow lest conditions when the reservoir pressure falls below the bubble point pressure and is correspond with the performance of dissolved gas reservoirs. The model reservoir comprises two commingled layer, where a well test is conducted on a fully perforated interval. The water phase is assumed immobile. The main objective of this work is to interpret the flowing well pressure response and to predict reservoir characteristics based on its performance. The work presented is based on a constant terminal rate analysis, but it can also applied to constant bottomhole pressure and can be used to predict the Inflow Performance Relationship (IPR). Introduction Over the past several decades pressure transient testing has been recognised in the industry as a viable tool for reservoir characterisation and to estimate in-situ average reservoir parameters to identify large scale features such as sealing faults and boundaries; and to assess near wellbore conditions. Such analysis inform field development and adjustment of parameters on reservoir simulation models, and may influence well completion and workover decisions. Stratification is a particular kind of permeability heterogeneity, which can have major effect on reservoir performance. Information on individual layer characteristics and the impact of pressure and saturation distribution in layers tinder multi-phase flow conditions can be helpful in the operational management of reservoirs. Transient testing offers the possibility of obtaining such in-situ information about the layers. Proper design and analysis of these tests can give estimates of pressure and saturation distributions and fluid mobility in the layers, and may also monitor movement of fluid banks. Although, in principle, the layers can be isolated by downhole packers and tested separately, the applicability of this approach is limited by economical or operational constraints. Thus, an alternate testing method and interpretation model for layered reservoirs under multi-phase flow conditions can add value. The behaviour of multi-layer systems for slightly compressible fluid flow has been studied extensively in the literature. In the past, these studies have been concerned with average transport property estimation. Lefkovits, et. al, with their classical approach, show that for a two-layer single phase case, the average reservoir flow capacity, kh equals the arithmetic average of the layer's reservoir flow capacities. The individual layer contribution to total flow, (qi/qT) during the transient period depends on the fractional flow capacity, kihi/ kihi. The study of commingled reservoir systems has been extended by including other well and reservoirs parameters such as different layer initial pressures, wellbore storage and skin. Kucuk et al., used the convolution integral method to determine the individual layer properties and skin factors from a multi-layer test. A multi-layer test is a series of simultaneous measurements of layer sandface flowrates and pressures. Superposition principles were used to estimate average permeability and skin factors for layers below the measurement point. The individual layer properties are then determined after the bottom layer permeability and skin factor have been estimated. Ehlig Economides studied individual layer properties in multi-layer reservoirs using simulated models. She used flow-rate transient type curve analysis to understand the flow mechanism in a layered reservoir system. She compared the general characteristics of two different multi-layer reservoir models. The first model consisted of two commingled zones. Each zone was divided into several crossflow layers. The second model was a five-layer commingled reservoir. She found that in the early time, the pressure behaviour of the first model behaves as if it were a commingled reservoir, and concluded that commingled reservoir approaches are adequate to analyse it. P. 709