Electric Submersible Pump Startup Using Model-Predictive Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for controlling electric submersible pumps (ESPs) lack efficiency in optimizing startup schedules and real-time adjustments, leading to suboptimal operation and potential damage to reservoirs, pumps, and processing facilities.
Innovation Solution
The implementation of model-based methods for developing offline startup schedules and real-time adjustments using model-predictive control (MPC), which utilizes ESP speed and choke opening as actuators to optimize startup sequences while adhering to operational constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional startup methods are used for ESPs, then the startup process is simple to implement, but the startup schedule is suboptimal and may cause damage to reservoirs, pumps, and processing facilities
Solution Approach 1:
The system performs preliminary analysis and optimization of the startup schedule before actual startup execution. Model-based methods predict optimal ESP speed and choke opening sequences in advance, allowing the system to prepare optimal startup parameters before committing to the actual startup process, thereby ensuring safe operation while maintaining manageable complexity
Solution Approach 2:
The system implements model-based control that continuously monitors actual startup performance and compares it with predicted behavior. This feedback mechanism allows real-time adjustments to ESP speed and choke opening to maintain optimal startup sequences, ensuring reliability while the automated feedback loop manages the complexity of control
2Productivity
If model-based offline startup schedules are implemented, then optimal startup sequences are achieved, but computational complexity and processing requirements increase
Solution Approach 1:
The model-based optimization is performed offline before actual startup operations. By pre-calculating optimal ESP speed and choke opening sequences using reservoir models and system constraints, the computationally intensive work is done in advance rather than in real-time during startup, achieving productivity improvement while keeping online computational requirements manageable
Solution Approach 2:
The startup control is divided into discrete sequences of ESP speed changes and choke opening adjustments. This segmentation allows the complex optimization problem to be broken into manageable steps that can be pre-calculated offline and then executed sequentially during actual startup, reducing real-time computational burden while maintaining optimization benefits
3Reliability
If real-time adjustments are made during startup, then operational constraints are satisfied, but response time and control complexity increase
Solution Approach 1:
The system pre-calculates the sequence of control actions needed to satisfy operational constraints during startup. By determining optimal ESP speed and choke opening trajectories in advance based on predicted system behavior, the system can execute pre-planned adjustments that satisfy constraints without requiring time-consuming real-time calculations during the actual startup process
Solution Approach 2:
The system implements dynamic adjustment of ESP speed and choke opening during startup based on real-time system state. This dynamic control allows the system to respond to actual conditions while maintaining constraint satisfaction, and the automated dynamic adjustments occur efficiently without significantly extending startup time
Data Source
AI summary
A method of using an electric submersible pump startup using model-predictive control includes defining an objective of the control algorithm comprising an intake pressure to achieve by an end of the startup schedule. The method also includes translating the objective into a cost function that mathematically describes the objective to develop a model-based offline startup schedule based on startup operational parameters, constraints, and a physical model and entering the startup operational parameters, the constraints for the startup operational parameters, and the physical model into a processor. The method also includes simulating system responses with the processor. The method also includes determining one or more optimal control actions by optimizing the cost function. The method also includes controlling the electric submersible pump based on the optimal control actions determined.