Summary
Perforating-laboratory experiments can be a useful element of field-perforating-job design. In some instances, the goal is to qualitatively compare multiple candidate perforating techniques. In others, the goal is to obtain quantitative insight into likely flow performance in the field. Although the laboratory will never perfectly replicate the downhole environment, it can yield useful results, which—if properly interpreted—can enable informed prediction of downhole-flow performance.
A traditional flow-laboratory experiment (API RP 19B 2006) yields numerous key results, four of which are required inputs to downhole-inflow simulators. These are perforation-tunnel length and diameter, crushed-zone thickness, and permeability. In the case of natural perforated completions as opposed to stimulated completions (e.g., hydraulically fractured or sand control), these parameters (in addition to other system and wellbore parameters) dictate the skin and ultimate flow performance of the completion.
Crushed-zone permeability is typically inferred from core-flow efficiency (CFE) and an assumed crushed-zone thickness. Traditionally applied, this technique can yield values that are inaccurate, and can produce misleading predictions of downhole performance more significantly.
To address this, we have developed new methods for both measuring and interpreting CFE. The new measurement technique yields CFE values that we show to be more meaningful and relevant. The new interpretation technique provides a consistent method of translating CFE to crushed-zone permeability, and is capable of accounting for the effect of partially plugged tunnels. This work clarifies and improves the link between laboratory and field performance of perforators, with the ultimate goal of increasing the value of downhole-inflow-performance predictions.
While other work is ongoing to challenge the framework of the conventional skin models, the present paper accepts these models as a premise. This work simply presents a coherent methodology of interpreting laboratory data, with the intent of generating the required inputs for skin models as they currently exist. Furthermore, it is recommended that this workflow be considered for inclusion in any revisions to the API Section 4 (API RP 19B 2006) testing protocol.
