(245b) Process Engineering for Lignocellulosic Ethanol Facilities: a Combined Model- and Experiment-Based Approach and Comparison with Operational Results From a Demonstration-Scale Plant

The Energy Independence and Security Act (EISA) of 2007 mandates that 16 billion gallons per year of cellulosic ethanol shall be produced by 2022. This will require the construction of a large number of facilities using a wide variety of cellulosic ethanol feedstocks. This presents a large technical and economic challenge that can be addressed, in part, by intesified process engineering.

Process design for lignocellulosic ethanol plants is more complicated than for corn-ethanol facilities because the feedstocks have much more compositional variability than common yellow dent #2 corn. Within similar feedstock groups such as coniferous softwoods there is significant variation in the amount of cellulose, hemicellulose, lignin, extractives, acetyl, and ash components, as well as variation in optimal pretreatment severity. Wide variation has been observed even within the same species, depending on the harvest location, age of the plant, and part of the plant that has been harvested. As a result, each new lignocellulosic ethanol facility will typically have a unique feedstock composition and will need significant process design modifications to operate optimally.

Successful, rapid facility design and rapid retrofitting requires a combination of specific process experiments with a robust process model. This paper will describe a methodology to quickly and accurately tailor an entire-facility process design for a unique feedstock or blend of feedstocks. The method has also been applied to convert an existing facility from one type of feedstock to another.

Results from the methodology will be compared with actual operations data from a working, 1.5 million-gallon-per-year lignocellulosic ethanol facility in Upton, WY using a variety of softwood feedstocks.