Plant Optimization With POSy-System

Plant Optimization With POSy-System

Our optimisation processes also include POSy-SystemTM, as part of our business alliance with Paradox Engineering & Technology Canada, SERVICES | pent (paradoxengtech.com). The POSy-SystemTM optimisation methodology is a true multivariable orthogonal resolver whereby it solves for the most optimal of a set of multivariable Set Points, Outputs, System Gain and Nonlinear Relationships or other Disturbance Variables. POSy-SystemTM can be integrated withing existing Distributed Control Systems (DCS’s), Programmable Logic Controllers (PLC’s) and in existing Advance Process Control (APC’s), and works as an APC / Optimisation numerical engine.

The diagram below illustrates 3 Controlled Variable set-point trials, where A, B & C are considered from other influential strategies to determine the desired true Optimized Set Point X. POSy-SystemTM solves for all of the Optimal Set Points simultaneously over the defined operational scenarios.

• 3 existing systems – experiential, statistical and predictive methods
• All of these systems are not precise
• Large errors are generated (predictive method)

• Time consuming and therefore very costly
• Dynamic changes cannot be captured
• No guarantees – you get what you get

• Work from top down instead of bottom up
• About 10 x faster (cost effective) than traditional methods
• Solutions are much more precise with 95-percentile accuracy

• Increases efficiency/productivity,
• Improves quality and reduces waste all at the same time

• Error generated is minimized
• System more easily adaptable to dynamic fluctuations and trends can be determined

• Provides an excellent Return On Investment (ROI) for consumer (and investor)
• Considered disruptive technology
  • Guaranteed improvements

• Most testing has occurred for manufacturing – non-Newtonian fluids (non-linear systems): achieved 12% to 23.5% efficiency improvements
• Have worked with modelling software in industry: thermodynamic modelling software – works well

• Have worked with modelling software at U of A: simple mixing (works well); SAGD (works well); and Tennessee Eastman Process & HYSYS (works well)