Subscribe to be notified for updates: RSS Feed

Luca Cadei

Home
Process Safety
Luca Cadei

Luca Cadei

Production Engineer, Operations, ENI SpA

Luca Cadei holds a BSc in Energy Engineering and an MSc in Petroleum Engineering cum laude from Politecnico of Milan. He completed his background with two Exchange Programs in TU Delft and Norwegian University of Science and Technology (NTNU) specializing in Computational Fluid Dynamics and Reservoir and Production Engineering awarded with ENI scholarship. After graduating, he joined Exxon Mobil as a process engineer, working directly on field with a focus on optimization and troubleshooting activities in the downstream division. He moved ENI Exploration & Production as a production engineer in 2015. His activities are mainly focused on production optimization, troubleshooting, debottlenecking and development of field exploitation strategy on ENI worldwide assets. His activities involves the use of process and thermo/fluid dynamics simulations, integrated models, advanced algorithms and analytics, combined with operations performed directly on field. He is currently SPE YP Italian Section president.

Definition of a Dynamic Operating Envelope of a Complete Production System

Process Safety

Abstract: This paper reports the development and application of an Advanced Process Control (APC) methodology based on genetic algorithms (GA) that provides plantwide optimized values for the PID controller parameters applied to a process dynamic simulation of a real operated oil field.
The objective is to guarantee a fast control action for critical user-defined scenarios, such as disturbances or setpoint changes. The novelty is represented by the use of GA as a scheduler to control the tuning process of the controller parameters. Moreover, the possibility to develop and test this approach on a dynamic process simulation before the direct application on field, allows to maximize the efficiency and the accuracy of the APC itself.
The GA creates a series of solutions composed by a set of controller parameter values and then keeps combining them with methods inspired by natural selection, such as crossover, mutation and selection. The solutions are tested in a process dynamic simulator until the time required to bring the system back to steady state conditions after an upset reaches a minimum value.
The optimized set of parameters is able to guarantee a faster control action compared to the one tuned by the plant simulation software alone and is particularly efficient in managing the process disturbances, upsets and setpoint changes scenarios.
The GA, thanks to its better convergence rate with a high number of parameters and constraints compared to other optimization methods, allows an accurate tuning of the controllers of larger plant sections with a limited computing power.

Copyright 2016 DMS Global - Design by DMS Cybernation