Recent step changes in downhole video technology and image analysis have coincided with a growing understanding of how perforation erosion can be used to measure the effectiveness of limited entry hydraulic fracturing. This has led to rapid growth in video-based perforation imaging and produced a substantial database of measured and statistically analysed perforations. A review of recurring patterns and common trends identified in the database provides useful insights on proppant placement and distribution.
An analysis was undertaken on a dataset that includes detailed individual perforation dimensions from more than 6,000 clusters and 600 stages. With a focus on understanding the uniformity of proppant distribution, the initial phase was to identify significant recurring patterns in cluster level proppant placement derived from perforation erosion measurements. Multiple treatment design parameters were then analysed to understand their influence on proppant distribution. Among those considered were stage length, number and spacing of clusters per stage, the number of perforations shot per cluster, perforation charge type and phasing.
While every well produces a unique set of results, several recurring trends were identified across the database. These often indicated sub-optimal proppant placement with undesirable consequences for production and ultimate recovery. Results demonstrate that
Proppant placement is often significantly non-uniform across a stage
A strong tendency for greater heel-side perforation erosion is typically observed for ‘geometric’ stage and cluster designs
A similar strong preference for proppant to be placed in perforations located towards the low-side of the wellbore is also apparent
More uniform proppant distribution can be obtained using an engineered design approach
Although they are often inter-related several treatment parameters can be engineered with relative ease to produce more uniform proppant placement
Analysis methods, results, treatment parameter considerations, primary conclusions and other relevant findings will be discussed in detail.
The majority of research on treatment design parameters that influence proppant placement has mainly used CFD-DEM models. The approach presented in this paper, however has used empirical, in-situ data. The size of the dataset and the frequency at which certain tendencies are observed provide some confidence that the approach and results are valid and can help improve treatment design. It is hoped that the results of the study will provide hydraulic fracture specialists with further evidence-based guidelines that ultimately help increase production and enhance ultimate recovery.