Generative AI is now widely available, often integrated into existing applications and tools in our facilities. While commonly used for writing, its potential to transform routine work is immense. GenAI can significantly enhance efficiency, effectiveness, and job satisfaction by reducing effort on repetitive low-value tasks. A panel will discuss various practical use cases to showcase how GenAI is being applied in everyday operations.
Take Aways:
Scope of capabilities represented by GenAI (Text, Graphics, Code, and more)
Creative and practical examples of how GenAI has been effectively applied
Insights into the attributes of a good opportunity for GenAI and implementation approaches
Moderator:
Bruce Taylor, FurtherTec
Speakers:
Ryan Byrne, Macedon Technologies
Brandon Taylor, Flatirons Digital Innovations
Andrew Vignes, HF Sinclair Corporation
Patrick Robinson, PBF Energy, Inc.
In this session, we will discuss in detail the definition of Normalization of Deviance and how it has contributed to major accidents, including both Space Shuttle disasters. Examples will be provided of Normalization of Deviance in operating and maintaining process plants, as well as in our daily lives.
Participants will:
Gain an understanding of how to stop Normalization of Deviance
Learn how to identify and correct this behavior
Take this concept home to apply with their families
Identifying, diagnosing and remedying catalyst losses are crucial to successful FCC operation. It is not a matter of if, but when, your FCC unit will encounter catalyst losses and whether it is small and manageable or large and unsustainable. This session will review catalyst loss fundamentals and troubleshooting followed by case studies from two units with very different outcomes.
Over the past 12 years, Ketjen FCC modeling engineers have simulated over 145 FCC units globally, identifying common challenges that prevent refiners from fully leveraging kinetic FCC models to maximize profitability. These challenges often stem from structural issues in model setup, poor-quality process measurements, or a lack of detailed knowledge about the models. This presentation will highlight 4 to 6 real-world examples, offering insights on how to: 1) avoid common modeling predicaments, 2) quickly identify if problems exist, and 3) implement actionable solutions.
Moderator:
Steve Shimoda, Technip Energies
Speaker:
Alan Kramer, Ketjen Corporation
Bridget Cadigan, Marathon Petroleum Corporation
This Lessons Learned session will focus on telling the stories of fatal incidents in Hydroprocessing units to facilitate an awareness and understanding of the different root causes contributing to them. Sharing the lessons learned will help prevent similar incidents from happening again. The presenters will cover a broad spectrum of incidents and their causes with a Q&A discussion at the end.
Moderator:
Thad Sauvain, Becht
Speakers:
Jeff Johns, Becht
Greg Mullek, Chevron U.S.A. Inc.
After more than 50 years of root cause analysis (RCA), you might expect problem solving to be a clear and straightforward process. It’s not. You can find yourself caught in unproductive debates. “Is that a contributing factor, or a causal factor?” “That’s a primary cause, but it’s not the main cause.” The result is frustration, inefficiency, and solutions that allow a repeat of the same problems.
A better approach to RCA is one built on evidence-based cause-and-effect relationships. The steps are simple: define the problem, explain why it happened, and identify solutions to reduce risk. Its focus is on having more reliable work processes, learning, and improving, which helps engage the frontline and minimize blame.
This session will explain the pitfalls of drifting from scientific problem solving and features a case study demonstrating how first-principles RCA leads to tangible improvements in reliability and human performance.
Participants will learn:
Four common errors that arise when explaining why an incident occurred
The biggest misconception about RCA
Why effective RCA does not require any proprietary techniques, terminology, or software
How RCA can be scaled for low-risk incidents and expanded as needed for higher-impact events.
How to reduce “human error” by involving those who perform the work
Moderator:
Bill Clark, Phillips 66
Speakers:
Bill Clark, Phillips 66
Mark Galley, ThinkReliability