Sensitivity analysis yielded reflux temperature, reboiler pressure and main condenser pressure as the most influential variables for efficiency optimization.
At a Midwest bio-products facility, distillation is used to produce isopropyl alcohol (IPA) that is 80-85% pure from water and other contaminants. The efficiency of the distillation column can be defined as the mass of steam required to produce a given mass of IPA. Using available process data, the baseline average process efficiency was estimated at 1.71 LBprod/LBsteam. Using the same data and Kinergetics understanding of the process, SPO was then used to develop a trained neural network to allow sensitivity testing and identification of operational variables most relevant to improve efficiency while maintaining or improving product purity.
Example dashboard with real time efficiency and purity feedback. Optimal ranges for influential variables are indicated.
The developed model was used to simulate column operation if the identified variables were maintained within determined optimal ranges. The projected steam reduction is 3,204 LB/Hr, corresponding to a new average efficiency of 1.72 LB/LB. The results from the sensitivity analysis are incorporated into a dashboard providing continuous feedback to operators to help ensure the savings are achieved.
The annual projected natural gas reduction is 43,970 MMBTU/Yr (84% boiler efficiency, 8,000 Hr/Yr) with a CO₂ emission reduction of 2,333 metric tons/Yr. At a fuel rate of $4.0/MMBTU, the annual cost savings is $176,000/Yr and the simple return about 6 months.
The Kinergetics SPO process involves management and operations input to help secure buy-in from those responsible for production so that they are vested in the process.