Refrac Implementation in Legacy Shale Play Re-Development: A Multivariable Approach

Sharma, Akash (Enverus) | Mathukutty, Shibu (Enverus) | Siegel, Samantha (Enverus)

OnePetro 

Abstract

Improved completion design and field development strategies have provided commodity price resilience by sustained efficiency gains across most major US Shale plays. This rapid evolution in completion practices, however, has created behind pipe opportunities. Refracturing offers a viable solution to maximize on these opportunities, however, its effectiveness is dependent on a variety of factors. The present paper explores the implementation of refracturing as a re-development strategy in legacy shale plays and evaluates it as a truly multivariable problem.

The paper takes into consideration petrophysical parameters, initial completion design, chemical composition, formation quality, time from original completion, refrac completion design and production performance to quantify impact on refrac KPIs such as IP ratio, EUR ratio, decline trend impact, amongst others. The paper does this by using an ACE (alternating conditional expectation) non-linear regression model that incorporates the KPI’s as response variables and utilizes the transforms of a wide range of input variables to identify cause and effect relationships. By running this analysis across multiple legacy shale plays, including the Haynesville, and Barnett, the paper provides best-practices to maximize refracturing success.

While refrac can offer a viable solution in obtaining incremental production, depending on the basin, a refrac can be a tenth of the expense of a new well and can beneficially impact the production from the existing well. In most cases, the analysis found EUR predictions improved by 30% - 200%. While correlations varied across basins and completion design, an inverse correlation was found between refrac KPIs and initial frac intensity.

Although, refracturing in horizontal shale wells is a well-established practice, a significant amount of analysis on their performance is focused on one or two key variables. The present paper adds to the existing body of literature by using data analytics and machine learning to evaluate this strategy from a truly multivariable standpoint. The paper also provides best practices to evaluate and predict refrac performance to de-risk refrac as a field re-development strategy.