Effective Electrical Submersible Pump Management

Synthetic carry plays an critical role in petroleum enterprise to preserve manufacturing flowrate and to extend Electrical the lifespan of oil wells. one of the maximum popular artificial elevate strategies is electric Submersible Pumps (ESP) due to the fact it may produce excessive flowrate even for wells with exceptional intensity. despite the fact that ESPs are designed to work below excessive situations along with corrosion, high temperatures and high stress, their lifespan is much shorter than anticipated. ESP screw ups result in production loss and increase the value of substitute, due to the fact the price of intervention work for ESP is a lot better than for different artificial Electrical raise techniques, specially for offshore wells. therefore, the prediction of ESP failures is especially treasured in oil manufacturing and contributes a lot to the layout, creation and operation of oil wells. The contribution of this take a look at is to use three machine gaining knowledge of algorithms, which might be choice Tree, Random woodland and Gradient Boosting gadget, to build predictive models for ESP lifespan whilst using each dynamic and static ESP parameters. The consequences of those models have been in comparison to discover the most Electrical appropriate model for the prediction of ESP lifestyles cycle. similarly, this examine additionally evaluated the affect factor of various running parameters to forecast the maximum effect parameters on the length of ESP. The outcomes of this study can provide a higher information of ESP conduct in order that early moves can be found out to prevent capability ESP disasters.
artificial lifts are widely utilized in production wells to optimize manufacturing flowrate [1]. the electric Submersible Pump is a famous technique applied to approximately 15 to 20 percent of a million wells international thanks to its wonderful characteristics as it produces a high fee in depth properly [2]. however, the screw ups of ESP typically arise all of sudden and they’re tough to predict because of the complicated nature of viable causes. The disasters lead clearly to the manufacturing disruption and a large amount of money for substitute. for this reason, the estimation of ESP lifespan is vital to have an early substitute plan to be able to avoid the production loss. In Electrical parallel, being capable of become aware of the important thing parameters that have the most impact on ESP screw ups can make a contribution to enhancing working performance.
In recent years, oil and gasoline specialists have attempted to identify the primary causes of ESP disasters and to expect the life cycle of ESP by using unique methods along with the usage of the harmonic styles in the electric powered supply [3], or real-time ESP tracking gadget [4], or thinking about ESPs screw ups pattern evaluation [5]. Guo, et al. (2015) constructed a assist Vector device model that used electric and frequency data to detect anomalies in ESP in the course of operation [6]. Gupta, et al. (2016) presented an analytical framework for early fitness tracking of ESP based totally on facts-driven modeling [7]. The framework can robotically perceive actual-time popularity and verify the health of ESP continuously, for this reason any detection of bizarre hassle may be signaled to operators earlier than it occurs. Sherif, et al. (2019) [8] used decision Tree approach mixed with principal element evaluation (PCA) to decide the solid vicinity for ESP operation with parameters taken as consumption pressure and temperature, vibrations, system cutting-edge and frequency.
using PCA turned into additionally noted inside the paintings of Abdelaziz, et al. (2017) to are expecting failure of ESP [9]. The input parameters had been also mentioned in the research of Popaleny, et al. (2018) wherein the authors presented how ESP mechanical and electrical malfunctions were contemplated inside the dynamic contemporary spectrum using Motor modern Signature evaluation [10]. In quick, system studying has been widely used in petroleum enterprise to predict ESP lifespan in latest years [11]. various methods had been proposed and used separately including choice Tree, Linear Regression, Random wooded area [12].
This examine gives a one-of-a-kind manner to technique this hassle with the aid of constructing a predictive version the usage of exclusive system learnings algorithms, which can be decision Tree, Random woodland and Gradient Boosting device, to expect ESP lifespan with both dynamic and static parameters. a complete of thirteen working parameters had been collected from ninety seven ESP. moreover, the version additionally classifies the effect of these parameters at the ESP lifespan. The results can be used to enhance the ESP performance by way of as it should be adjusting the most influential parameters on ESP lifespan.