Dynamic Model Calibrations Using Short Flow Tests of Hydraulically Fractured Wells in Green Gas Condensate Fields

Abstract History matching of hydraulically fractured gas condensate green fields and proper forecast are very challenging because of significant uncertainty in formation permeability, lateral and vertical heterogeneity, complicated fluid behaviors, fracture properties and limited data duration. For green fields located in remote areas, surface facility is often unavailable, and restriction of releasing green-house gas makes flaring of produced hydrocarbon in an extended well test undesirable. Only short well tests after fracking are available for the dynamic model calibrations which can span from a few days to several weeks. History matching of these short flowback and well test data are not straightforward as significant amount of hydraulic fracturing fluids exist near the wellbore and will be produced during the early phase of flowback. The flowback of the injected foreign fluids distorts gas rate, condensate rate, tubing head pressure and bottom-hole pressure measurements. This paper discusses a systematic workflow/methodology that enables reliable and sufficiently accurate model calibrations from history matching of these short flowback data before the model is used for efficient and fast field performance forecast. The workflow includes: (a) using equivalent negative skin to represent hydraulic fractures in the respective flow units; (b) local grid refinements to better represent water saturation and pressure distribution near the wellbore during the fracturing fluid injection and flowback phases; (c) considering injected fracturing fluids to create more realistic water saturation and energy near the wellbore at the initial condition before flowback; (d) history match gas, condensate, water, tubing head pressure and bottom hole pressure data at the same time within the more representative duration in flowback data; (e) history match individual reservoir flowback data to calibrate the model for the respective flow unit; (f) only match the stable and more representative data during flowback; (g) proper adjustment of water relative permeability curves to match mobile formation water and injected fracturing water production; and (h) model verification through commingled production, which will enhance confidence of the calibrated dynamic models. Field histories show that the implemented workflow and methodology renders improved quality of dynamic model calibrations, which make field production forecast more reliable for green field gas condensate developments.