Modeling Two-Phase Flowback From Multi-Fractured Horizontal Tight Gas Wells Stimulated with Nitrogen Energized Frac Fluid

Abstract

In recent years it has become common practice to energize water-based fracs placed in tight gas formations with nitrogen to enhance well clean-up and maximize future reservoir performance. Despite the operational benefits, the addition of a second gas phase in the fractures complicates the quantitative analysis of flowback data.

In this work we expand on the model presented by Clarkson and Williams-Kovacs (2013) for the quantitative analysis of two-phase (water+gas) flowback data using analytical history-matching for the estimation of hydraulic fracture properties. Similar to the previous work, our interpretation is that the early flowback data corresponds to wellbore + fracture volume depletion (storage), under the assumption that fracture storage volume is much greater than wellbore storage. In cases where multi-phase matrix flow exists, this flow-regime is followed by multi-phase transient matrix linear flow. From these two flow-regimes bulk permeability (dominated by fracture permeability), effective fracture half-length and other parameters can be estimated.

To account for the addition of a nitrogen gas phase we make the assumption of perfect mixing between the hydrocarbon gas phase and nitrogen gas phase where gas composition varies as a function of time. Using this method, nitrogen mole fraction declines from an elevated value and approaches hydrocarbon gas phase nitrogen content (from standard gas analysis) once all injected nitrogen has been depleted (commonly after the flowback period). Further, all wells analyzed previously using the developed methodology have been single-phase dry shale gas wells, where we have assumed that gas is primarily sourced from the adsorbed state during flowback and fracture depletion is immediately followed by single-phase matrix linear flow. In these cases there has also been no condensate production. In the tight gas case presented in this paper, all gas is sourced from free gas and significant condensate production is seen, therefore further changing the dynamics of the problem. For the analysis, we assume that the well is being flowed above the dew point and therefore there is no free condensate flowing in the reservoir. In this example the well has also been shut-in during the flowback period to install production tubing and therefore we will examine the impact of shut-in on effective hydraulic fracture half-length. With this example we show that shutting in the well during flowback decreases the effective half-length and therefore hinders long-term production potential.