Dinesh Krishnamoorthy

Modelling and model predictive control of oil wells with Electric Submersible Pumps

In this paper we describe control challenges related to operation of oil wells equipped with Electric Submersible Pumps (ESP) and formalize them in a control problem setting in the language of control system engineers. Then we present a simple dynamic model of an oil well equipped with ESP. This model can be used for controller development. To solve this problem, we propose a Model Predictive Control (MPC) strategy and present experimental results of an MPC controller successfully tested in a large scale test facility with a full scale ESP, live crude oil in an emulated oil well.

Gas Lift Optimization under Uncertainty

Abstract In this paper, we consider the problem of production optimization under uncertainty applied to gas lifted well networks. Worst-case and scenario optimization methods are presented to explicitly handle the uncertainty. We also compare the performance and computation time of the presented methods with nominal and ideal cases using Monte Carlo simulations. We show that the scenario optimization method is able to reduce the conservativeness, however at the cost of computation time. We also show that the performance can be improved by parameter adaptation using an extended Kalman filter for combined state and parameter estimation.

Control of the Steady-State Gradient of an Ammonia Reactor using Transient Measurements

Abstract This paper presents the application of a steady-state real-time optimization strategy using transient measurements to an ammonia synthesis reactor case. We apply a new method for estimating the steady-state gradient of the cost function based on linearizing a dynamic model at the present operating point. The gradient is controlled to zero using a standard feedback controller, for example, a PI-controller. The applied method is able to adjust fast to the new optimal operation in case of disturbances. The advantage compared to standard steady-state real-time optimization is that it reaches the optimum much faster and without the need to wait for steady-state to update the model. It is significantly faster than classical extremum-seeking control and does not require the measurement of the cost function and additional process excitation. Compared to self-optimizing control, it allows the process to achieve the true optimum.

Changing between Active Constraint Regions for Optimal Operation: Classical Advanced Control versus Model Predictive Control

Abstract Control structures must be properly designed and implemented to maintain optimality. The two options for the supervisory control layer are Advanced Control Structures (ACS) and Model Predictive Control (MPC). To systematically design the supervisory layer to maintain optimal operation, the constraints that can be given up when switching active constraint regions should be prioritized. We analyze a case study in which we control the temperature and the flow in a cooler with two degrees of freedom (DOF) represented by two valves, one for each of the two streams. Either valve can saturate and make a constraint active, forcing other constraints to be given-up, and thus changing the set of active constraints. We show that optimal or near-optimal operation can be reached with both ACS and MPC. We do a fair comparison of ACS and MPC as candidates for the supervisory layer, and provide some guidelines to help steer the choice.

Steady-state real-time optimization using transient measurements

Abstract Real-time optimization (RTO) is an established technology, where the process economics are optimized using rigourous steady-state models. However, a fundamental limiting factor of current static RTO implementation is the steady-state wait time. We propose a “hybrid” approach where the model adaptation is done using dynamic models and transient measurements and the optimization is performed using static models. Using an oil production network optimization as case study, we show that the Hybrid RTO can provide similar performance to dynamic optimization in terms of convergence rate to the optimal point, at computation times similar to static RTO. The paper also provides some discussions on static versus dynamic optimization problem formulations.