1a. Objectives (from AD-416):
Our overall goal is to develop a productive, efficient, and sustainable biomass feedstock supply system using perennial grasses and legumes as a primary feedstock. This project addresses critical needs for feedstock development using perennial grasses and legumes by developing innovative ways to fractionally harvest and store these feedstocks. Specific objectives are: 1) design and fabricate new harvesting mechanisms to separate the high-protein and high-fiber fractions from these crops at harvest; 2) quantify the machine's field performance using a controlled set of operating variables; 3) use this information to improve the mechanisms through re-design during the off-season; and 4) collaborate to develop on-farm storage and pretreatment systems to preserve and add value to both the high-protein and high-fiber fractions. Additional objectives in the extended project include: 5) determine storage characteristics of switchgrass & reed canarygrass stored under anaerobic conditions in bunker & bag silos at different moisture contents; 6) quantify packing density, porosity & temperature profile of biomass materials during storage; 7) determine aerobic stability of stored feedstocks; 8) assess composition & bioconversion potential of feedstocks before & after storage under various conditions; & 9) estimate storage costs under different conditions, taking into account storage losses & changes in quality, as well as bioconversion potential of the stored material.
1b. Approach (from AD-416):
Design and fabricate equipment for field-fractionation of bioenergy crops during harvest. Test equipment on established fields of alfalfa, switchgrass, and reed canarygrass. Determine yield and quality of fractions obtained, along with power requirements and operating costs. Improve design of equipment to improve performance, reliability, and operating costs. Store harvested materials under different conditions, and determine dry matter losses and quality changes resulting from storage under these methods. We will also compare the yield and quality of switchgrass and reed canarygrass when stored at different moisture levels in either bag silos (a widely adopted storage technology) or in bunker silos (a potentially lower-cost method that does not generate plastic waste). Conventional measures of silage preparation (packing density, porosity, temperature profiles) will be combined with measures of microbial conversion in the silo (fermentation acid production), and of subsequent bioconversion potential of feedstock after storage to fuels and fuel precursors. Because exposure of anaerobically stored feedstocks to air can cause undesirable spoilage by microorganisms, we will conduct aerobic stability tests to determine if storage method affects the rate and extent of feedstock spoilage, and if different microbial agents are responsible for spoilage of different feedstocks stored under different conditions. We will estimate costs of storage of each feedstock that will take into account the dry matter losses and quality change of each feedstock during storage, and the bioconversion potential after storage.
3. Progress Report:
This project was related to Sub-Objective 2A of the parent project: To determine the optimum methods of fractionating, harvesting, and storing biomass materials. Research on the harvesting of perennial forages was conducted by a collaborator at the University of Wisconsin-Madison. Experimental plots (10 acres each) of switchgrass (SWG) and reed canarygrass (RCG), established at the University of Wisconsin Agricultural Research Station, Arlington, Wisconsin, were used for harvesting and storage research over successive years from 2004 through 2011. The most effective harvest strategy for both SWG and RCG was a single-cut system during late summer to early fall. SWG produced higher yields than RCG for each of four successive years studied. Both crops dried more easily than alfalfa or other forage grasses. Wide-swath drying resulted in the shortest drying times and permitted baling at 20% moisture in a single day. An alternative harvest scheme, in which cutting of the crop was deferred until the following spring, resulted in drier material that could be baled immediately. However, dry matter (DM) yields were reduced by 17% (SWG) or 26% (RCG). The density of both SWG and RCG bales was lower than for alfalfa or most forage grasses. This factor must be increased for economical shipping and handling purposes. SWG was more easily baled than RCG, which tended to clog the baler throat. Baling efficiency was improved 46% by net wrapping versus baling with twine. When considering perennial grasses as potential biomass feedstocks, most analysis is conducted assuming that these grasses will be field-dried to below 20% moisture (wet basis [w.b.]) and packaged in bales that will be stored either uncovered or indoors. However, this is not a robust or cost-effective harvest and logistics scenario. These crops are typically harvested in the late fall when drying conditions are difficult. A delay of harvest to spring can insure a dry-standing crop, but the losses over winter are substantial. Future biorefineries will require a very consistent feedstock, but bales stored outdoors have highly variable moisture and quality. Storing bales indoors overcomes this problem, but adds significant cost. Bales require a lot of handling and are quite difficult to size-reduce at the biorefinery, both of which also add to the feedstock cost. Our approach is to direct-cut perennial grasses in the last fall when they are 35-45% moisture (w.b.) and conserve by anaerobic storage. This approach offers many advantages over the traditional bale system. Direct cutting eliminates the mowing, raking, and baling operations. The system is less weather-dependent because no field drying must take place. Use of a forage harvester to chop the material produces a value-added, size-reduced product at the time of harvest. Anaerobically sealed storage produces a consistent product at removal from storage. Finally, with proper packaging, the chopped feedstock can be transported as economically as bales. We found that low-moisture anaerobic storage of round bales reduced DM losses to 1.1%. For comparison, bales stored outdoors for 9 to 11 months had variable dry matter (DM) losses depending on baling method, with plastic film displaying the least loss (3.8%) and sisal twine the most (14.9%). Storage of conventional bales under cover reduced DM losses to 3% and resulted in more uniform biomass feedstock, but these losses still exceeded those of low-moisture anaerobic storage. Additional experiments were conducted to evaluate low-moisture anaerobic storage. RCG and SWG were harvested at two moisture contents (~54 and 43% wet basis) in the fall of 2010 and stored in pilot-scale silo bags until late spring 2011. DM loss during storage ranged from 2.3% to 3.4%, with an average across all treatments of 2.3%. There was no significant difference in storage losses between high- and low-moisture treatments. Both grasses were removed from storage with very uniform moisture through the cross section and length of the bag. Moisture content at removal only varied 1 -2 percentage units from the harvest moisture. Aerobic stability was determined during 2-day and 7-day aerobic exposure after removal from storage. For the two exposure durations, DM losses of 1.6% and 2.3% occurred. Lactobacillus buchneri was applied as a bacterial inoculant at the time of harvest to improve the aerobic stability of the grasses. This inoculant was successful in reducing the heating degree days and temperature rise above ambient in both grasses during aerobic exposure. L. buchneri also produced significantly lower yeast and mold counts than did a control. Inoculation did not have a significant effect on DM loss during storage or aerobic exposure. The harvest and anaerobic storage of moist perennial grasses has been a success thus far. TThe system offers a longer harvest window, produces a size-reduced material at the time of harvest, conserves feedstock DM well, and is aerobically stable (especially when inoculated with L. buchneri). One possible way of adding value to forage crops is to fractionate them to separate the high-protein leaf from the high-fiber stem by stripping the leaf from the stem at harvest. A direct-cut sickle head was modified to mount on a forage harvester. Then three treatments were created: whole-crop, high-cut (mainly leaves), and low-cut (mainly stems). Both RCG and SWG were harvested in this fashion, and all three treatments were preserved by ensiling in a bag silo. Single-pass fractional harvest of RCG, SWG, and mature alfalfa was also conducted using a modified snap bean harvester. The harvester had a multi-tined stripping rotor that stripped leaf tissue off the standing plant. The remaining material, mainly stem, was subsequently cut, windrowed, and baled in separate operations. Between 33% and 60% of the total available DM yield was removed by the stripping rotor. The yield in the stripped fraction was dependent on rotor:ground speed ratio and the depth of penetration of the rotor into the plant canopy. When the crop was fractionally harvested in late August, the crude protein content of the stripped and standing fractions was 8.2 and 5.5 %, respectively, for SWG, and 8.7 and 5.1 %, respectively, for RCG. The moisture level of the stripped fraction was 66% when first-cut alfalfa was harvested in mid-July and the plant was at seedpod maturity. This moisture level would have allowed direct ensiling without additives for preservation, although the crop was extremely mature and rank at harvest. The stripped fraction of the RCG and SWG harvested in late August was 49 and 62%, respectively, so the stripped fraction from these two plants also could be directly ensiled. The stem fraction of alfalfa and RCG was baled in large round bales and stored outdoors to determine storage characteristics. Bales wrapped in twine did not form a good thatch because the rainfall easily penetrated the bales and much spoilage was evident. Bales wrapped in breathable film fared much better in this respect.