SPE 163925; Optimal Sand Cleaning Algorithms for Planning and Executing Electric Wireline Suction Tool Operations

Background

Conventional methods used for downhole sand or debris cleaning have considerable investments in economic costs and time. Scheduling these activities to minimize production interuptions and to maximize unrestricted flow is a constant challenge faced by production and flow assurance engineers in the Oil & Gas industry. A sand and fill cleaning decision can be expressed as an optimization problem since its purpose is to maximize the production of the well to accomplish production goals at the end of a time period.

This paper presents optimization algorithms applied in dynamic programming to help in scheduling cleaning interventions for wells in order to maximize the continuous production under physical, technical and economic considerations while minimizing the investment cost of the total operation. A strategy will be presented which takes into account mainly the fixed costs but also the variable costs and sensitivity analysis that allow the model to better approximate reality. The strategy considers three options based on the costs of cleaning with either electric line or coiled tubing technology; 1) to completely remove a given volume of sand, 2) to clean to a minimum acceptable level or, 3) not to clean and allow sand volume increases to continue.

The cost to perform sand cleaning with a certain technology based on a mathematical function, considers the following requirements:
 the relationship between volume of sand produced per unit time, Vt (flow),
 (Vmax-Vmin) or the interval between Vmax in which production of oil is minimized and the Vmin for maximum production,
 the time horizon to perfom the sand cleanings, and
 Points that make the operation unfeasible such as physical restrictions in the well or operator time/cost constraints.

This paper will also present two cases demonstrating the strategy to schedule cleaning interventions that achieves a set production goal.

Find the full paper at onepetro.org

Read more SPE papers