Abstract This paper presents a method to match the production data of commingled tight gas reservoirs. A simple computer program is used to history match production data, estimate the individual OGIP and productivity of each layer, and forecast the total performance of the well. The method was verified against both simulation and actual field data. The method is based on a layered model that couples the material balance equation for gas reservoirs with the stabilized gas flow equation for each layer in the commingled system. The analysis is done through history-matching the total well production data using an optimization routine to help determine the two parameters (OGIP and flow coefficient) of each layer that best match the production history. The method showed excellent results for cases that are predominantly in the stabilized flow period. This method can be used to analyze commingled tight gas reservoirs and predict their performance. It can be also used to obtain the OGIP value for each layer and assess the relative importance of each layer in the commingled system. Introduction It is a common problem in gas reservoir engineering that volumetric methods are not adequate for estimating gas reserves. Also, build-up tests may be impractical in tight (low permeability) reservoirs because of the long time required for the pressure to stabilize. Consequently, material balance (M.B.) methods cannot be used to predict reservoir performance. The problem is even more complicated in multilayer no-crossflow reservoirs, where little information is available. In these situations, decline curve analysis seems to be the most practical tool for reserve estimation and performance prediction. Decline curve analysis started as early as 1945 with empirical equations that give the rate performance as a function of time. Fetkovitch showed that some of the empirical decline equations can be based on both the material balance equation (M.B.) and the deliverability equation for gas reservoirs. Later investigators tried to develop techniques to aid in the matching process for one-layer radial reservoirs. Others developed other sets of decline curves. Different approaches of combining M.B equation with the gas flow equation for single-layer reservoirs were presented by different authors. The current paper describes a method which is sometimes more accurate in matching and forecasting production data. It also gives estimates of OGIP and flow coefficients, Jg, for each layer in commingled reservoirs. Background Commingled Reservoirs. Commingled reservoirs are reservoirs connected only through the wellbore. These reservoirs do not exhibit crossflow within reservoir boundaries. Each layer in the commingled system can be defined by its OGIP and its flow coefficient, Jg, if stabilized flow is reached. If these parameters are different for each layer, the layers' performance will be also different. Fig. 1 shows typical production performance for a two-layer commingled system producing against constant bottom hole flowing pressure, Pwf. The data for this figure were obtained from simulation of a two-layer reservoir where the two layers have equal OGIP and differ only in their productivity. Decline Curve Analysis Limitations. For multi-layer reservoirs, investigators have observed that the hyperbolic decline exponent, b, can reach values above the limiting value for single layer reservoirs (b>0.5). However, matching the rate-time data for these high values of b is strictly empirical and only based on fitting the hyperbolic decline equation to the production data. Also, matching the data is often difficult if the history of production data is not large enough or if transient data prevail. P. 545