Domestic ethanol production has grown substantially over the last 30 years due to a growth in national energy security concerns, more stringent Federal standards for vehicle emissions, and government agricultural incentives. The recently signed Federal energy legislation requires an increase in ethanol production to 7.5 billion gallons a year within the next 10 years. There is, however, a recognized need to reduce the production costs for fuel ethanol and to make its manufacture more energy-efficient. In the U.S., most ethanol derives from corn starch saccharification and fermentation. The production schemes used to produce ethanol from corn utilize mature unit operations (e.g., batch fermentation, distillation, centrifugation, evaporation), such that there are minimal marginal economic benefits for improving the productivities of these process operations. However, via incorporation of improved process operations, it will be possible to substantially decrease the production costs for starch-derived ethanol. The use of selective membranes as substitutes for conventional separation operations in corn refining have demonstrated significant process improvements and projected economic benefits. However, such technologies will require substantial pilot-scale demonstration to ensure that the risk of technical failure is minimized. The subject of this proposal is the development of an integrated membrane-based process that will provide substantial ethanol cost reduction and that builds upon existing membrane products that are unique to the proposing organization. OBJECTIVES: The subject of this proposed SBIR program is the development and demonstration an innovative cost-saving process for ethanol production, which includes the critical features of: (a) substitution of a highly efficient hydrophobic pervaporation membrane step to substantially replace distillation, and (b) use of ceramic crossflow MF or UF as a pre-treatment operation that will optimize the pervaporation (PV) membrane process and provide additional cost benefits. The overall objective of this Phase I program will be to demonstrate short-term membrane performance data that, when scaled to corn-based ethanol plants of 5 MMGPY or larger, is consistent with projected ethanol production cost benefits of greater than or equal to 5 cents/gal ethanol. While the economic projections performed in the Phase I program will determine the range of ethanol plant sizes over which the anticipated production cost benefit target is met (and the degree to which it is exceeded), it is anticipated that the following separation performance targets will be required: (1) for the initial MF/UF clarification step, a steady-state flux of 60 L/m2-h with complete yeast cell retention at a fermentor cell concentration > 109 cells/mL; (2) for the hydrophobic pervaporation step, an ethanol/water selectivity > 60 and time-averaged flux (including time required for membrane cleaning) > 0.75 kg ethanol/m2-h, averaged over the feed ethanol concentration required for the process. APPROACH: An integrated process for cost-effective production of ethanol from starchy crops will be examined. Microfiltration (MF) membranes will be used as a pre-treatment operation prior to the ethanol-selective pervaporation (PV) membranes. MF and UF (ultrafiltration) membrane layers will be applied to porous ceramic multi-channel "monolith" supports using slip-casting. These elements will be qualified using previously developed crossflow retention tests. MF membrane qualification will consist of monitoring retention of micellar casein proteins in a dilute suspension of skim milk in ultrafiltered water. UF membrane qualification will consist of monitoring retention of soluble whey proteins in a dilute suspension of skim milk in ultrafiltered water. A turbidimeter will be used to measure the casein and whey protein retention. Hydrophobic zeolite (silicalite) membranes will be made on monolith supports containing MF membrane layers. Silicalite membranes will be synthesized using parameters defined in recent work at CeraMem. Bubble point and gas permeation testing will be used to initially characterize the membranes. Silicalite membrane elements will be tested using a bench-scale PV test system using a feed of 5 vol% ethanol in water. Permeate samples will be collected to determine ethanol concentration and membrane flux. Laboratory-scale fermentations will be performed. During these fermentations, the beer will be decanted and fed into the crossflow system containing a selected MF or UF module to be tested for beer clarification. Flux will be measured and membrane permeates will be evaluated for clarity using the turbidimeter. Membrane elements will be subsequently cleaned using either a commercial alkaline cleaning solution or a dilute bleach solution. Water flux recovery will be measured for each cleaned element. Several batch fermentations and crossflow clarification trials will be performed to determine the effects of the process parameters. Clarified broths will be subsequently used for ethanol PV using silicalite membranes to determine the effect of the properties of the MF or UF membrane used to pre-clarify the feed stream. Initially, batch fermentation will be performed and the contents of the fermentor will be coupled to the crossflow system while the concentrated yeast slurry will be returned to the fermentor. Fermentation will continue and cell concentration and fermentor ethanol content will be measured (via yeast dry weight measurements and GC). The permeated broth from the crossflow clarification operation will be fed into the PV system. The operation of partial yeast concentration, re-topping of the fermentor, and PV of the clarified broth will be performed. The fermentor productivity and the crossflow and PV membrane performance will be monitored. Cleaning of both the MF/UF and PV membranes will be evaluated. In similar tests, the retentate from the PV system will be periodically removed and recycled to the fermentor to determine if retained soluble compounds in the PV retentate have an effect on fermentation kinetics and/or membrane separations. PROGRESS: 2006/05 TO 2006/12 Note: Full report is confidential until January 1, 2011. The text below is the non-confidential executive summary of the project. The purpose of this USDA SBIR Phase I project was to develop and demonstrate an innovative cost-saving process for ethanol production that includes: (a) substitution of a highly efficient hydrophobic ethanol-extracting pervaporation (PV) membrane step to substantially replace distillation, and (b) use of ceramic crossflow microfiltration (MF) or ultrafiltration (UF) as a pre-treatment operation that can optimize the PV membrane process and provide additional cost benefits. Ceramic membrane elements with characteristic pore sizes ranging from 0.002 to 0.2 micrometers were fabricated and qualified for baseline properties and subsequently evaluated for clarification of various fermentation products, which included fresh, raw (yeast-laden) beer and "barreled" raw beer samples from two pilot-scale ethanol production processes in the Midwest. Crossflow microfiltration readily clarified yeast-laden raw beer to produce a yeast-free permeate from which ethanol could be extracted via membrane PV, and concentrated yeast slurry that could be recycled as a key part of a continuous fermentation process. The steady-state flux of the clarification operation was maintained above 60 liters/m2-h. Clarified beer samples (permeates from the filtration operations) were employed as feeds for hydrophobic membrane PV using zeolitic (silicalite) membranes. Ethanol flux stability and membrane selectivity were found to vary significantly depending primarily on the source of the clarified beer. PV membrane performance did not depend significantly on the specific nature or performance of the crossflow operation used to clarify the initial raw beer. Steep permeate flux decline rates typically were observed as were, in some runs, significant losses in membrane selectivity. This behavior was interpreted as resulting from competitive adsorption of specific low-molecular organic by-products that were present in the beers in low concentrations. Additional experiments were performed to determine the relative effects of specific soluble organic compounds on PV membrane flux and ethanol/water selectivity. Membrane performance data were input into a process model to estimate cost savings for membrane-based processes used to augment or replace conventional separation operations (e.g., distillation and centrifugation) used in ethanol manufacturing. Assuming that the best observed membrane performance data from the experimental program could be modestly improved upon by use of an appropriate pre-treatment process, energy cost savings of ~20 cents/gallon ethanol were projected. Corresponding payback periods were 30 months and less for plants with ca. 30 MMGPY and less capacities, with fermentor ethanol contents of up to 6% (w/v). IMPACT: 2006/05 TO 2006/12 Assuming that the subject technology is further developed in a follow-on Phase II program and ultimately integrated into commercial ethanol plants, it would significantly decrease the energy required to produce ethanol, thereby also reducing production cost. The technology is anticipated to be most economically favorable for small-capacity to medium-capacity plants, and for fermentation streams that contain less than ~8% ethanol. Both of these characteristics are expected to be typical of a growing cellulosic-derived ethanol industry in the next decade.
