Nonwoven Fabric Producer Recovered $4M/Yr.

Improvements to Product Yield based on Optimizing Feed Network

• A network of feed Sources being fed to two Tall Oil Refineries.
• Current optimization was based on minimizing logistics costs.
• Feeds were being blended at each refinery into one large mixing tank.
• Distillation parameters were fixed with little feed-forward automation into Tower Control.
• Feed pricing was based on a negotiated price versus taking into consideration Feed Quality and Tower Yield Effects.

Feed Composition and Tower Yield Analysis and Modeling Findings

• Significant differences in quality and composition from source to source – differences in paper mill processing and regional differences due to the trees and types used.
• Some qualities had plugging and adverse effects on processing that were not clear to customers as well as a lower number of high-value materials.
• Purchasing contracts did not consider the quality and adverse effects on the process. Different feeds also had different effects on the two processing plants.

Developed Refinery LP Model that simulated operations, assisted in monthly planning and ranked feeds based on quality, and impact on the plants that:

• Established profitability to the two plants by feed. Optimized feed distribution to two plants (ex: some feeds close to one plant had better yields at the other offsetting extra transportation costs)
• Ranked Feeds and established 3-grade levels for feed.
• Established better criteria for negotiating price.
• Optimization of feed network also reduced production and downgrading of low-value products.
• Ranked new feeds and determined optimum pricing relative to the feed yield database as new feeds were acquired.