2010 AIChE Annual Meeting Salt Lake City, UT November 7-12, 2010 Advanced Process Control in ExxonMobil Chemical Company: Successes and Challenges Tyler A. Soderstrom PhD. Yang Zhang PhD. John Hedengren PhD. Session 10B01: In Honor of Tom Edgar’s 65 Birthday II
12
Embed
Advanced Process Control in ExxonMobil Chemical Company ...€¦ · Advanced Process Control in ExxonMobil Chemical Company: Successes and Challenges Tyler A. Soderstrom PhD. Yang
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
2010 AIChE Annual MeetingSalt Lake City, UTNovember 7-12, 2010
Advanced Process Control in ExxonMobil Chemical Company: Successes and Challenges
Tyler A. Soderstrom PhD.
Yang Zhang PhD.
John Hedengren PhD.
Session 10B01: In Honor of Tom Edgar’s 65 Birthday II
222010 AIChE Annual Meeting
November 8, 2010, Salt Lake City, UT
Outline
• Process Industries Advanced Control Toolbox
• ExxonMobil Chemical’s Advanced Control Experience
• Engineering Specialists: Process Control
• Advanced Control Improvement Needs
• Tom Edgar’s Impact
• Summary & Conclusions
332010 AIChE Annual Meeting
November 8, 2010, Salt Lake City, UT
Process Industries Advanced Control Toolbox
Closed Loop Multi-period economic
Closed Loop economic
Multi-variable constraint
Base Regulatory
Primary Function
Process Characteristic:
T P F A
Continuous Cyclic Semi- Continuous
Technology MaturityModel Rigor
Technology Maturity
Model Rigor
Sequentiallogic-based, discrete
PIDsingle input-single output control, multilevel
cascade, surge margin
MPRTOMulti-Period Real Time Optimization
Approximate nonlinear economicoptimization over a time horizon
RTOReal Time Optimization
Provides economically optimal targets to LMPC
LMPCLinear Model Predictive Control
stabilizes plant, push linearity to constraints
DynOptDynamic Optimization
provides nonlinear economic optimization over a time horizon
NMPCNonlinear Model Predictive Control
provides combined nonlinear economic optimization and control
442010 AIChE Annual Meeting
November 8, 2010, Salt Lake City, UT
Linear Model Predictive Control (LMPC)
• LMPC is the most widely used advanced control technology• Medium Size application routinely delivers significant energy savings as
well as additional production
• Example: Butadiene Recovery Unit, Baton Rouge Chemical Plant• 40 Manipulated Inputs, 50 Controlled Variables