Maximizing fracture productivity in unconventional fields; analysis of post hydraulic fracturing flowback cleanup

Abstract

Hydraulic fracturing, is a promising stimulation technique which is also known as hydrofracturing, hydrofracking and fracking. During the hydraulic fracturing (HF), the rock is cracked, i.e., fractured, by a high pressure injection of a fluid which is known as fracturing fluid (FF). The FF is mainly water, carrying suspended sand or another type of proppants into the well to initiate fractures in the reservoir rock, and consequently, hydrocarbon and FF will move towards the well more easily through fractures.
Hydro-fracturing is extensively used to increase the well productivity index, particularly in unconventional, tight and ultra-tight reservoirs. This expensive procedure, though, sometimes fails to meet expectations regarding the production enhancement. The leading explanations for this reduced performance is fracture clean-up inefficiency of the fracturing fluid (FF) that was primarily injected.
In this study, a parametric investigation of FF clean-up effectiveness of fractures was performed with 143360 simulations (in 35 different sets) including injection, shut-in and production stages. Because of the vast number of simulation runs which was required to be implemented by a reservoir simulator, a computer code was developed and utilised to routinely read input data, implement the simulation runs and produce output data. In each set (which consists of 4096 runs), instantaneous impacts of twelve different parameters (fracture and matrix permeability (i.e., Kf and Km) and capillary pressure (Pc), end points and exponents of gas and FF in the Brooks-Corey relative permeability correlation in both fracture and matrix) were investigated. To sample the domain of variables and to study the results, full factorial experimental design (two-level FFS) and linear surface methodology explaining the dependency of the loss in gas production, compared to the case there is no loss (i.e., 100% clean-up) to the related parameters at different production stages were investigated through he tornado charts of the response surface models, frequency of simulation runs with obtained Gas Production Loss, GPL, and saturation distribution maps of FF.
Results pointed out that in general, factors that control the mobility of FF inside the fracture had the most significant impact on cleanup efficiency. Conversely, in tight and ultratight sets, particularly when the applied pressure drawdown for the duration of production stage was small, the impact of fluid mobility within the matrix on gas production loss was more noticeable, i.e., it is crucial how fluids flow inside the matrix rather than how fast fracture is cleaned. In lower permeability matrix, in general, more gas production loss was detected and clean-up was slower. The impact of Pc on GPL minimisation was stronger when pressure drawdown was small and/or shut-in time was prolonged. As the formation becomes tighter, this observation was more pronounced, in other words, for such formations, the impact of a change in pressure drawdown and/or shut-in time on Pc and GPL was more noticeable.
Additionally, the results showed that as the length of the fracture reduced the impact of fracture pertinent parameters (i.e., fracture permeability and fluid (gas and FF) mobility pertinent parameters of Corey correlation in the fracture) on GPL reduced and the impact of those pertinent parameters in the matrix on GPL increased. The impact of Pc on minimising GPL is less noticeable in shorter fractures and vice versa. As the length of fracture reduced, quicker fracture clean-up was detected compared to those for longer fracture.
These discoveries help us to better understand the hydraulic fracturing process and can be used to settle issues regarding the performance of hydraulic fracturing and to improve the design of hydro-fracturing operations, which is an expensive but popular stimulation method for tight and ultra-tight reservoirs.