Grant: Discovery Grants Program - Individual
University of Alberta
$26,950.00
Performance analysis for online chemical process optimization systems
FY 2012-13
Summary
Online optimization is widely used in the chemical process industries to ensure competitiveness though a high level of process performance. Common applications of online optimization in these industries include real-time optimization (RTO), model-predictive control (MPC), reactive scheduling, supply-chain optimization, and so forth. These online optimization technologies are, typically, complex and require regular maintenance to ensure their continued high performance (i.e., the performance of these automation systems degrade over time due to a number of factors). Thus, a key, but currently missing, element of online optimization systems is an effective performance analysis subsystem, which would include: performance monitoring, failure detection, and failure diagnosis.
The work proposed in this application will focus on RTO and MPC, which are two of the most common online optimization technologies. One of the critical challenges in analyzing the performance of RTO and MPC using operating data is that their operation is very dependent on the fidelity of the process model on which they are based. Any modeling errors (e.g., parameter mismatch or structural mismatch) can result in the underlying optimizer incorrectly identifying process constraints that should be limiting the achievable level of performance and the reduced gradient of the performance objective. This can result in the optimizer driving the process into areas of operation that actually decrease performance. Furthermore, unlike the currently available controller performance assessment techniques, there are no performance benchmarks (e.g., the minimum variance benchmark) available for comparison purposes.
The proposed research program is targeted at the development of the set of fundamental tools and ideas from which a comprehensive performance analysis and failure diagnosis framework can be constructed. The results from the research program will be tested using an industrially-relevant, pilot-scale distillation system and industrial case studies.
Program
Discovery Grants Program - Individual
Recipient
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
194372-2008
Name nom
Forbes, JFraser
Department département
Chemical and Materials Engineering
Organization id
9
Institution établissement
University of Alberta
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2008
Award amount
26950
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
4
Committee name en
Chem. & Metallurgical Eng.
Committee name fr
Génies chimique et métal.
Area of application code
500
Area of application group en
Manufacturing processes and products
Area of application group fr
Procédés et produits industriels
Area of application en
Manufacturing processes and products
Area of application fr
Procédés et produits industriels
Research subject code
1800
Research subject group en
Chemical engineering
Research subject group fr
Génie chimique
Research subject en
Chemical engineering
Research subject fr
Génie chimique
Partie
5
Nb partie
5
Application title
Performance analysis for online chemical process optimization systems
Keyword
NSERC
Application summary
Online optimization is widely used in the chemical process industries to ensure competitiveness though a high level of process performance. Common applications of online optimization in these industries include real-time optimization (RTO), model-predictive control (MPC), reactive scheduling, supply-chain optimization, and so forth. These online optimization technologies are, typically, complex and require regular maintenance to ensure their continued high performance (i.e., the performance of these automation systems degrade over time due to a number of factors). Thus, a key, but currently missing, element of online optimization systems is an effective performance analysis subsystem, which would include: performance monitoring, failure detection, and failure diagnosis.
The work proposed in this application will focus on RTO and MPC, which are two of the most common online optimization technologies. One of the critical challenges in analyzing the performance of RTO and MPC using operating data is that their operation is very dependent on the fidelity of the process model on which they are based. Any modeling errors (e.g., parameter mismatch or structural mismatch) can result in the underlying optimizer incorrectly identifying process constraints that should be limiting the achievable level of performance and the reduced gradient of the performance objective. This can result in the optimizer driving the process into areas of operation that actually decrease performance. Furthermore, unlike the currently available controller performance assessment techniques, there are no performance benchmarks (e.g., the minimum variance benchmark) available for comparison purposes.
The proposed research program is targeted at the development of the set of fundamental tools and ideas from which a comprehensive performance analysis and failure diagnosis framework can be constructed. The results from the research program will be tested using an industrially-relevant, pilot-scale distillation system and industrial case studies.