The intent is to make better use of production data to estimate gas reserves and long term well deliverability. Downhole pressures are calculated from surface pressures via multiphase flow calculations using wellbore parameters. The reservoir is simulated based on a simplified reservoir model to determine the resulting pressure profile after each monthly production withdrawal. Reservoir inflow coefficients are then calculated based on production rates and calculated sandface and reservoir pressures. The reservoir, wellbore and inflow coefficient values arrived at are tested by history matching production data, and then used to generate production forecasts for estimating long term well deliverability and recoverable reserves.
The purpose of this paper is to present a novel technique for estimating original gas in place, recoverable reserves and well productivity for deep >750 m total vertical depth) gas wells using production data. The technique is applicable to wells producing under gas expansion drive from single well pools.
Current industry practice for estimating reserves and deliverability for producing gas wells is to rely on three separate techniques. These are: absolute open flow (AOF) testing, material balance using P/z plots, and production decline analysis using rate versus time or cumulative production. Each technique is based on numerical analysis performed on a subset of the data available for the well. For both material balance and decline analysis plots, the raw data is generally just "eye-balled". Data points are selected where they fit the expected trend, and a straight line is fitted to the "good" data points. Production analysis software packages make this simplified approach very quick and easy, and discourage the more detailed calculations required on a point by point basis to ensure that the bottomhole shut-in pressures (BHSIP) are built up, or that the well has produced against constant back pressure for long enough to truly exhibit pseudo steady state behavior. Where discrepancies arise between the different analysis techniques, rather than balancing off the answers received from each analysis, often the analysis techniques are themselves ranked in order of perceived reliability and the "best" technique is assumed to be providing the "right" answer.
Where the data available is known to be reliable, the above is not a concern. This includes better quality reservoirs with short build-up and stabilization times, as well as older pools that have a well-established production history.
For tighter gas wells, especially those that must be hydraulically fractured to achieve commercial production rates, the quality of data available to establish reserves and deliverability is greatly diminished. Longer shut-in times, often in excess of a year, are required to get reliable built up pressure data for material balance. Similarly, long periods are required to reach stabilized flow before production decline analysis is appropriate and these are often never achieved if the well is regularly shut-in due to lack of demand. When a well has been hydraulically fractured, radial flow is usually not reached during an AOF test. Curve fitting is then required to estimate fracture length, skin and reservoir quality.
The analytical approach used merges as much data as possible within a single composite wellbore/reservoir model. The data incorporated into the model includes BHSIP data, production rates and surface pressures, expected reservoir geometry and original reservoir conditions;