Case Study: Reduce Energy Costs, Increase Production with Model Predictive Control

By Michael Tay | August 27, 2007
Mid-Missouri Energy Inc. (MME) is a 50 MMgy farmer-owned ethanol plant in Malta Bend, Mo., which started distilling corn in early 2005. MME employs 40 people while taking advantage of the area's abundant corn supply, using nearly 18 million bushels of locally produced corn.

The plant has helped create nearly 1,800 additional Missouri jobs and increased Missouri economic activity by more than $169 million. The facility is adding value to the local farmers' corn crop while helping to reduce the nation's dependence on foreign oil.

As MME entered the ethanol business, U.S. fuel prices skyrocketed and new legislation, including the renewable fuels standard (RFS), strongly encouraged ethanol use. In fact, national ethanol use goals set by the RFS will require 7.5 billion gallons of ethanol use annually by 2012, up from 2006 usage of approximately 4 billion gallons.

MME immediately began to help satisfy the need for more ethanol. In less than a year, a group of 15 farmers created a cooperative with more than 700 members and raised more than $20 million in equity, breaking a national record for producer investment in its initial equity offering. The plant, designed to produce 40 MMgy, has maintained operating production in excess of 50 MMgy. In fact, in its first fiscal year ending just seven months into operation, investors enjoyed a more than 30 percent dividend. While financially strong, MME has also invested in state-of-the-art technology, enabling it to more efficiently utilize energy and thereby reduce the environmental impact of its production process.

Plant operators decided to study the plant's stillage processing and distillers dried grains with solubles (DDGS) drying, which was the location of the plant's primary throughput bottleneck. MME chose to focus on their stillage processing section including DDGS dryers, thin stillage evaporators, centrifuges, stillage tanks and thermal oxidizer/heat resource steam generator to manage this complex plant section along with a principal energy consumer in the plant.

To further succeed in the dynamic ethanol market, MME sought to optimize its dryer performance, which would also allow it to increase production capacity without building additional process units.

Applying a Model Predictive Control Solution
To find an optimization solution to meet its demands, MME turned to Pavilion Technologies. Pavilion's Ethanol Solution, which leverages the company's trademarked Pavilion8 modular software platform and is based on model predictive control (MPC) technology, has enabled ethanol manufacturers to improve the performance of the dryer, evaporator, thermal oxidizer, molecular sieve distillation and fermentation stages of the production cycle.

In the late 1960s, the term advanced process control (APC) was popularized to refer to any algorithm or strategy that went beyond classical, ratio or proportional plus integral plus derivative control. Today, APC encompasses a variety of control technologies and methods such as supervisory, inferential, feedforward, adaptive, multivariable, nonlinear and MPC. Multivariable MPC has emerged as the prevailing method of APC, and there are more than 10,000 MPC applications currently in use, according to ARC Advisory Group. The MPC approach uses a reference model of the process to predict future process behavior and calculate an optimum set of control moves to minimize deviations from the desired control objective. MPC accounts for the interactions among process variables, allowing plant operators to reduce process variability and ensure the production process stays on target to meet desired business results.

MPC is a proven technology that enables an ethanol manufacturer to optimize process units, like fermentation or distillation, or their entire production process to increase production, reduce energy consumption and costs, improve product quality and increase yield. Ethanol manufacturers that strive to capitalize on the growing demand for ethanol are increasingly turning to MPC solutions to improve cost efficiency and environmental responsibility in order to maximize profit. MME selected Pavilion's MPC-based dryer control application to improve process efficiencies and increase its ethanol production capacity by de-bottlenecking the drying and stillage processing stage.

A Deliberate Process Yields Results
Following its trademarked ValueFirst customer engagement methodology, Pavilion worked closely with plant personnel to tailor the application to MME's site-specific needs. Iterative testing ensured the consistency of product quality and accuracy of the dryer control models. In MME's initial audit test period, beer rates increased significantly beyond any previous sustained period, and MME recorded a 6 percent increase in its 200 proof ethanol production rate. With production maximization on the dryers, dryer capacity demonstrated more than 6 percent higher capacity availability than previously sustained records. Energy consumption was also reduced by 2.1 percent on a British thermal unit per gallon basis.

The stillage processing and drying application includes MPC covering stillage tanks, centrifuges, DDGS dryers, syrup evaporator, beer column/feed and thermal oxidizer/heat recovery steam generator. A key objective is the stabilization of both the syrup and DDGS qualities for easier handling and higher yields along with continuous production maximization through the management of stable base losses in the beer column. This is done while using higher levels of steam and balancing stillage levels through the evaporator. Along with the MPC system directly setting targets for up to 13 existing regulatory system controllers (e.g., beer column and centrifuge feed flow) simultaneously, three continuous inferred property models were developed and run in real-time. The MPC system manages and provides operator feedback in real-time on inferred quality of syrup solids, and Dryer A and DDGS moisture. Operators and the control system are aided by minute-by-minute qualities supported by the normal frequency of operator sampling.

A supplemental benefit of a stillage processing application is due to the fact that the operator attention required to manage dryers and evaporators is higher than justified by the revenue produced in this plant section. So while DDGS, distillers wet grains with solubles and syrup coproducts provide a stable additional revenue stream, operator attention is generally higher on this section of the plant, particularly when weighted in relation to the fraction of plant revenue produced. Most dryer projects, including this one, achieve supplemental benefits allowing a refocus of operator attention to other areas of the plant including fermentation and higher revenue productive operations.

"Pavilion's Ethanol Solution has made a significant impact on our bottom line in a short amount of time," MME Plant Manager Chris Wilson says. "Our ability to de-bottleneck the dryer process resulted in immediate value and paved the way for our plant-wide deployment."

MME completed the project in just five months, and the Pavilion Dryer Control Application has delivered benefits to MME in excess of $3 million per year based on 2006 market prices, enabling the company to achieve its return on investment for the project in less than six months. The project was so successful that MME has elected to expand the deployment to include additional Pavilion optimization applications across the entire production process.

Michael Tay is a technical account manager with Pavilion Technologies. Reach him at [email protected].

The claims and statements made in this article belong exclusively to the author(s) and do not necessarily reflect the views of Ethanol Producer Magazine or its advertisers. All questions pertaining to this article should be directed to the author(s).