Big Data Analytics Maximizes Value from Smart Well Completions

Abstract

This paper presents a workflow based on big data analytics to model the reliability of downhole Inflow Control Valves (ICVs) and predict their failures. The paper also offers economic analysis of optimum ICV stroking frequency to maintain valves functionality at the lowest possible cost to the oilfield operator. Installing an ICV in a petroleum well is a costly process and is done by a drilling or workover rig. As such, maintaining a fully functional ICV throughout the lifecycle of a well is important to ensure proper return on investment. ICVs are known to malfunction if not periodically stroked/cycled. The action of stroking ensures that each valve opening is free from obstructing material that would prevent the ICV from operating between one valve opening step to another. When an ICV malfunctions, a costly functionality restoration operation is sometime required without guaranteed results. In other cases, the valve is declared no longer useful and the asset cannot be further utilized due to malfunction. In this paper, an analytical decision making model to predict failures of ICVs is presented that is based on rigorous big data analytics. The model factors in the frequency of stroking before a valve fails. Then, an economic analysis accounting for the CAPEX & OPEX of an ICV is included to optimize the stroking frequency. The utilized techniques include ICV failure and stroking records and classifying the data into pre-defined criteria. Cumulative probability distribution functions are defined for each data set and used to generate failure probability functions. The probability equations are factored into an asset management cost scheme to minimize expected maintenance costs and probability of ICV failure. The results of applying this novel methodology to any smart well clearly showed maximized ICV service life and proper return of investment. The results demonstrate that ICVs lifecycle was prolonged with low maintenance cycling cost. Methodologies similar to the one presented in this paper are true manifestation of the fruitful impact IR4.0 technologies have on oilfields day-to-day operations.