Managing an ethanol biofuels plant is challenging enough. Profit margins are razor thin to start with, so what do you do when a drought kills off most of your locally sourced corn supply? Such was the grim scenario facing one of the Midwest's largest ethanol plants, Hankinson Renewable Energy LLC in Hankinson, North Dakota, when a crippling drought dragged on through much of 2012's spring and summer.
"Last year was very bad," said Gary Jubien, solutions consulting manager at Honeywell Process Solutions. "Much of the corn crop was lost; many supplies were unavailable; prices went up; and the plant's operations were really threatened. So, assistance from something like advanced process control (APC) could really help them pull through." Jubien presented "APC Helps Ethanol Plant through Tough Times" this afternoon at the Honeywell Users Group Americas 2013 conference in Phoenix, Ariz.
The plant was designed by ICM Inc. of Colwich, Kansas, which also supplies the licenses for more than 100 plants nationwide. Its first-generation plants use corn or grain sorghum as feedstock, while its second-generation plants for cellulosic ethanol are beginning to roll out this year. Hankinson is at the generation-1.5 stage, which means it can use add-ons like kernel husks as complementary feedstocks.
The objectives of Hankinson's APC project were to improve control and reduce variability, reduce energy use, and increase the production of dried distiller's grain with solubles (DDGS) -- a valuable by-product that can significantly increase operational profit. "The plant's dryers have four disturbance variables each, and PID controls poorly," explained Jubien. "Each set of three centrifuges has a dead-end feed header, so they load unevenly, and APC could adjust each feed while balancing amps. Also, molecular sieves at the end of the production process cycle every four minutes, so steam demand changes frequently. And, the plant's thermal oxidizers (TOs) had no air-flow measurement, so we needed to added oxygen sensors for trim. The TOs and dryers also use all the natural gas in the plant, which is its second highest direct cost after grain."
Prior to the drought, Honeywell began working with ICM and Hankinson to implement some APC solutions at the facility. "Our procedure includes kick-off and pre-testing, preliminary design, step testing and model identification, commissioning, and as-built documentation," said Jubien.
Consequently, the plant adopted Honeywell's Profit Controller multivariable control software; Profit Stepper automatic step testing and identification software tools; and Profit Suite operator interfaces. As a result, several APC models were implemented, generating some extremely useful results.
For example, the key control variables (CVs) for the TOs were furnace temperature, steam pressure and oxygen, while the manipulated variables (MVs) included valve and damper positions, natural gas flows and other factors. Understanding these factors in detail and their performance allowed Hankinson's operators to run the TOs closer to their limits, which saved gas and other energy costs.
Likewise, APC enabled the plant's dryers to more consistently control the moisture of its DDGS, increase its moisture by 0.7%, and increase profits without going off-spec. Also, APC applied to the plant's centrifuges enabled them to balance their loads and performance much more efficiently.
Thanks to implementing APC, the gas/steam ratio of Hankinson's two TOs were reduced by 1% and 5.6%, respectively. Meanwhile, besides increasing product moisture while staying in spec, the dryers also reduced variability by 20%, reduced gas use by 4%, and increased production by 6%.
"Overall, the plant reduced its natural gas costs by more than $116,000, and increased its DDGS revenue by more than $675,000," added Jubien. "The payback period of the APC project was about eight months: the plant's also reduced gas use, but still increased its production by 6%."