To optimize the development of low permeability gas reservoirs, long hydraulic fractures are normally required. In such reservoirs, rapid decline rates during the first few months of production are to be expected. Even when a well has been satisfactorily stimulated, many operators with limited experience in tight gas reservoirs may incorrectly conclude that the stimulation treatment was unsuccessful, due to the rapid decline rates.
In some cases, however, an extremely rapid decline rate is a sign of an unsuccessful stimulation treatment. Poor well performance can be caused by (1) an extremely low formation permeability - .001 md or less, (2) insufficient fracture length - normally caused by fracturing out of zone or from poor sand transport, (3) insufficient fracture conductivity normally caused by crushing or embedment of the propping agent or (4) insufficient gas in place. propping agent or (4) insufficient gas in place. If an operator decides to continue the development of a low permeability formation, it is imperative that the cause of poor well performance be isolated. When the problem is extremely low formation permeability, the only solution is to create longer permeability, the only solution is to create longer fractures by pumping larger treatment volumes. If the problem is short fractures due to poor sand transport, the operator can increase the gel concentration or, perhaps, use a smaller mesh proppant. When low fracture conductivity is the problem, future fracture treatments should be designed using a higher sand concentration or a stronger proppant, such as bauxite. These solutions are straightforward after the problem is diagnosed; however, it is often difficult to determine the cause of poor well performance.
Numerous well test analysis techniques for fractured wells have been proposed in the literature. Each technique is limited by various assumptions; therefore, it is usually necessary to use several techniques and determine a "consensus" answer.
In this paper field examples are presented which illustrate the use of the different analysis techniques for determining the in situ reservoir and fracture parameters. In addition, a single phase, two dimensional, finite difference reservoir model has been used to history match all of the well tests. With the reservoir model it is possible to simulate variations in fracture conductivity, non-Darcy flow, flow rate changes, wellbore storage, gas compressibility gradients, gas viscosity gradients and other parameters which can complicate the analysis of field parameters which can complicate the analysis of field data. The finite difference history match analysis normally provides a more realistic reservoir description which can be used to evaluate the various "hand" calculation analysis techniques.
Computer model history matching has been used in the petroleum industry for many years. A majority of the work has been directed towards matching the production from large fields; however, some work has been production from large fields; however, some work has been published concerning individual well problems. Dogru published concerning individual well problems. Dogru et al. presented a review of previous history matching literature and investigated the single well, radial flow case. In their work, only porosity, permeability and drainage radius were varied. It was permeability and drainage radius were varied. It was concluded that if both porosity and permeability are allowed to vary, a number of -k combinations can be used to match short term production data. To determine a unique value for permeability, long term production data are required so that the value of porosity production data are required so that the value of porosity can be calculated based upon the decline rate.
Aydelotte also studied the radial flow problem including the effects of skin upon the history matching results. He concluded that if porosity is fixed, reservoir permeability and skin can be accurately history matched from transient well test data.
When analyzing pressure buildup tests, several performance parameters are normally considered as performance parameters are normally considered as fixed for a specific problem. These parameters are (1) producing time prior to shut-in, (2) volume of gas produced prior to shut-in, (3) producing rates during the production period, and (4) flowing bottom hole pressure at the time of shut-in.