Speakers:
Darin Foote, CHS Inc.
Chuck Johnston, Marathon Petroleum Corporation LP
Vidyashankar Kuppuraj, Marathon Petroleum Corporation
Johnny Serafin, Phillips 66
Nic Tognetti, CHS Inc.
Advanced process control (APC) has been employed in FCC applications for 40 years. Despite the similarities of FCC Technology and APC platforms the success of these applications varies widely. In this session we will discuss common threads that lead to poor outcomes and how to build a successful APC program on a modern FCC.
Artificial Intelligence, Advanced Process Controls
Facilitator:
Atique Malik, AIControl LLC
Speaker:
Yangdong Pan, Delek US
Toni Adetayo, Imubit, Inc.
Multi-unit optimization has long been a complex issue in the oil and gas industry. Despite efforts using first-principle models or empirical approaches, challenges persist. However, the emergence of machine learning and AI technologies offers an alternative solution. In particular, AI-based process control technology has shown promise for multi-unit optimization. This session will delve into an example using a distillate system optimizer to understand how these AI models address large-scale optimization challenges and how parent and child models collaborate.
Participants will learn:
The strength of the technology and its high flexibility of handling core issues
How the technology deals with the availability of individual units, and cooperates with unit controllers from other advanced control technologies
How to sustain the technology’s performance and benefits
In this session we will discuss in detail the continuous improvement in plant profitability achieved by having a holistic approach for advanced automation where applications such as procedural automation, Artificial Intelligence Plant Control (AIPC), control strategy in distributed control system and safety instrumented system all work together in harmony. AIPC, leveraging Reinforcement Learning (RL), offers a transformative approach to process plant optimization, addressing key limitations of traditional Model Predictive Control (MPC).
While MPC has been a cornerstone of advanced process control for decades, its reliance on linear models and manual tuning often struggles to effectively handle complex, nonlinear process dynamics, particularly in scenarios with frequent disturbances or changing operating conditions. AIPC overcomes these limitations by continuously learning and adapting to the plant's behavior. We will share our motivation to develop this product in-house and how easy it is to maintain the application and sustain the benefits by existing operations staff without the need of any AI or MPC expertise.
Participants will:
Understand the benefits of using AIPC over conventional MPC
Learn how Linde has successfully implemented AIPC on their Air Separation Units that include cryogenic distillation and very complex heat integration and recovery process.
Realize the importance of embracing this cutting-edge technology for improving profitability and safety of their plant and also for their personal and professional growth.
What is basic process control?
How is it done? Why is it important for the refinery engineer to be familiar with the basics?
In this foundational session, the discussion will focus on the workings of everyday loops and instrumentation that a refinery engineer is likely to encounter.
Moderator:
Sriram Ramaganesan, Phillips 66
Speakers:
Joe Boyce, Marathon Petroleum Corporation
Scott Flanagan, Marathon Petroleum Corporation
Tim Olsen, Emerson Automation Solutions
This session covers the complete spectrum of process optimization methods at a refinery or petrochemical plant. From models predictive control to real time optimization via process models and onto machine learning methods via neural networks. The attendee will gain an understanding of when to apply these methods in practice and the organizational support needed to succeed long term.
First principles models with adjustable parameters alongside parameterised neural networks are applied in a commercial package to a refinery product train. This method offers simplicity of implementation as compared to the more conventional real time first principles models and the entirely machine learning approaches. The implementation is at a commercial refinery and organizational, implementation and design details are discussed.
Moderator:
Rob Confair, Marathon Petroleum Corporation
Speakers:
Darin Foote, CHS Inc.
Chuck Johnston, Marathon Petroleum Corporation LP
Vidyashankar Kuppuraj, Marathon Petroleum Corporation
Johnny Serafin, Phillips 66
Nic Tognetti, CHS Inc.
Advanced process control (APC) has been employed in FCC applications for 40 years. Despite the similarities of FCC Technology and APC platforms the success of these applications varies widely. In this session we will discuss common threads that lead to poor outcomes and how to build a successful APC program on a modern FCC.