This paper discusses results from the first successful deployment of a predictive modelling technology that informs pressure optimization procedures to help minimize sand production and increase hydrocarbon production efficiency in sand prone oil wells.
The technique takes variabilities in sand production observed through time across the reservoir section, inferred from downhole sand entry logs, alongside real-time sand transportation logs that monitor sand deposition in pipe as key inputs (both of which computed using a fiber optic Distributed Acoustic Sensor (DAS) based Downhole sand monitoring system). This data is then combined with other time series sensor inputs, like choke position, Down Hole Pressure (DHP) and surface flowline acoustic measurement (sand detector) to predict drawdown pressure envelopes to improve production efficiency.
This paper details observations and initial field results from the first deployment of the capability in a highly deviated sand prone oil well completed with an open hole gravel pack (OHGP) completion in the BP-operated Azeri- Chirag- Gunashli (ACG) field located in the Azerbaijan sector of the Caspian Sea. The paper will detail observations and procedures used to increase oil production by over 25% and eliminate sanding risks using the technology. The proposed workflow is part of a comprehensive suite of downhole sand surveillance and management tools fueled by streaming analytics capabilities run on DAS data that have played a key role in managing sand production challenges in the ACG field.
The technology has been applied numerous times for base protection, drawdown optimization and targeted remediation. In this instance, we discuss the use of the technology to (1) identify and inform the source of sand detected at surface e.g., formation or completion accumulation, (2) identify formation intervals at risk of sanding, and (3) design advisory operational procedures for production optimization.
