Applied Process Simulation-Driven Oil and Gas Separation Plant Optimization Using Surrogate Modeling and Evolutionary Algorithms

Abstract

In this article, the optimization of a realistic oil and gas separation plant has been studied. Using Latin Hypercube Sampling (LHS) and rigorous process simulations, surrogate models using Kriging have been established for selected model responses. The surrogate models are used in combination with an evolutionary algorithm for optimizing the operating profit, mainly by maximizing the recoverable oil production. A total of 10 variables representing pressure and temperature at various key places in the separation plant are optimized to maximize the operational profit. The optimization is bounded in the variables and a constraint function is included to ensure that the optimal solution allows export of oil with a Reid Vapor Pressure (RVP) < 12 psia. The main finding is that, while a high pressure is preferred in the first separation stage, apparently a unique optimal setting for the pressure in downstream separators does not appear to exist. In the second stage separator, apparently different, yet more or less equally optimal, settings are revealed. In the third and final separation stage a correlation between the separator pressure and the applied inlet temperature exists, where different combinations of pressure and temperature yields equally optimal results.

Optimisation of carbon capture from flue gas from a Waste-to-Energy plant using surrogate modelling and global optimisation

“…Others suggest that for up to 10 variables a sampling size of 10-15 times the number of variables should suffice [29,30]. In the present study a sampling size 20 times the number of variables has been applied, which has previously been used by the author [31] with good experience. An automated process of running all the computer experiments defined by the sampling plan is made combining the process simulator with Python (programming language) via COM (Microsoft Component Object Model) [32].…”

“…A black-box wrapper is made in Python exposing the process simulation as an object which can be called like a regular function, taking the seven factors/variables as input, and returning the desired output when the simulation has converged. A similar black-box approach has been used by others [31,[33][34][35][36] using either VBA, python or Matlab/Octave as programming layer. For each sample in the sampling plan a corresponding simulation is made and the results recorded.…”

“…The sampling plan and recorded simulation model output is used to train a Kriging model [37][38][39] using the pyKriging package [27,40,41]. See also [31,34,42,43] for other examples and more information about Kriging in chemical engineering applications.…”

“…For the present application the optimisation problem is a global optimisation problem and analytical derivatives are not available directly from the process simulation. Numerical derivatives can be estimated, however, it is often experienced that process simulators based on sequential solvers which include tear-streams/recycles can generate noise in the numerical derivatives due to the applied tolerances [31,33,34]. Noise can be reduced by setting extremely low tolerances for the solving of tear-streams, but this comes at the expense of a significantly increased number of iterations for solving the flow sheet and may not be desired.…”

“…For that purpose derivative-free optimisation routines are required. On the other hand when applying surrogate models numerical derivatives obtained from the surrogate can be smooth and this can also enable choosing derivative based algorithms, with numerical estimates of the derivatives being used [31,44].…”