Reservoir and fluid characterization of a tight gas condensate well in the Montney Formation using recombination of separator samples and black oil history matching

• Published in: Journal of natural gas science and engineering Vol. 49; pp. 227 - 240
• Main Authors: Behmanesh, Hamid, Hamdi, Hamidreza, Clarkson, Christopher R.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.01.2018
• Subjects: Black oil simulation; Differential evolution; History matching; Sensitivity indices; Tight gas reservoir; Sensitivity indices; Differential evolution; History matching; Tight gas reservoir; Black oil simulation
• ISSN: 1875-5100
• DOI: 10.1016/j.jngse.2017.10.015

Liquid-rich shale (LRS) reservoirs are economically attractive but operationally challenging particularly for cases where multi-phase flow occurs within the reservoir. Proper treatment of PVT and rock properties, as well as rock-fluid interaction, in these unconventional reservoirs is central to providing improved short- and long-term oil and gas production forecasts. In the presence of limited surface sampling, the available analytical models often do not provide satisfactory results due to the uncertainty in the initial in-situ fluid system. In such situations, and particularly when there are many unknown parameters, numerical models are ideally suited, using a history matching framework, to assist with reservoir and fluid characterization. In this paper, production data from a multi-fractured horizontal well completed in a tight gas condensate reservoir in the Montney Formation in western Alberta, Canada is presented and analyzed using black oil numerical simulation. An assisted history-matching routine (i.e. Differential Evolution (DE) algorithm) is used in combination with black oil numerical simulations to characterize reservoir fluids and estimate reservoir and hydraulic fracture properties. The applicability of black oil numerical simulation for accurate prediction of the fluid model and well performance using numerous compositional numerical simulations and various fluid systems is first verified. The in-situ fluid is assumed to be a mixture of recombined separator samples with unknown oil-gas recombination ratio. The effect of time of sampling on the produced well stream composition is considered. Our results show that recombined separator sample mixtures collected early on during production of wells subjected to limited drawdown can successfully represent reservoir fluid properties with the exception of saturation pressure. In other words, the reservoir fluid PVT behavior can be predicted by recombination of an early initial separator fluid sample only by varying the saturation pressure. Hence, the assisted history matching can be performed using black oil numerical simulations with a reduced number of unknowns for the fluid system. The well/reservoir properties and unknown reservoir fluid are characterized in terms of a 12-parameter system. The history-matched results using DE could satisfactorily reproduce the water and hydrocarbon surface flow rates, and flowing bottomhole pressure curves. The quality of results are comparable to those obtained previously (Hamdi et al., 2015) using a fully-compositional numerical model. The new workflow using assisted history-matching combined with black oil simulation provides a practical yet accurate method for characterizing fluid, reservoir and fracture properties in unconventional gas condensate systems. •Real production data of a tight gas/condensate well is analyzed by black oil numerical simulations.•Differential evolution is used for the history matching.•Global sensitivity analysis is performed to evaluate Total Sobol' indices.•In situ fluid model and well/reservoir parameters are estimated.