Determining the schedule time required for harvest operations is an essential prerequisite to determine herbaceous biomass priority for securing inventory and effectively matching the transport of biomass from SSLs (satellite storage locations) to a biorefinery. Estimating the number of days expected to be available for baling is difficult because agricultural field work is heavily weather dependent and information concerning winter field operations is not well established. Harvest costs depend in part on the investment required in harvest machines (number of machines purchased), and this investment depends on the field capacity of the machines and the number of field workdays available during the harvest window. Therefore, an estimate of the number of harvest days is necessary to determine the total investment in harvest machines required to support a biorefinery.

The optimal perennial grass harvest strategy for a biorefinery remains to be determined. One potential strategy is to delay harvest until the plants standing in the field have transitioned from a non-dormant to a dormant state. Transitional events within the plant include translocation of nutrients from the above-ground plant parts to below-ground plant parts as initiation of degeneration (senescence) of above-ground tissues occur. Delaying harvest until the end of this transitional period, and harvesting over a short time period, can result in near maximum biomass yield, reduced moisture content, and reduced amount of nutrients removed with the biomass. This strategy would suggest an optimum harvest window for switchgrass from October to December. However, a harvest season this short would require substantial investment in harvest machines to complete harvest within a 3- month period. It also requires a large storage for inventory to supply the plant for year — around operation. By delaying harvest until nutrients have translocated, biomass tonnage will be decreased by this loss of mineral content, however since the relative amount of carbon in the material is increased, conversion efficiency and combustion quality may be improved [17]. Moreover, the translocated nutrients stored in the roots can be used for growth and development by the plant year after year, thus, reducing the need for and cost of supplementing the soil with nutrients through fertilization.

A second strategy is to extend harvest over as many months as possible. This would enable more economical use of harvest machines, reduce the quantity required in SSL storage, and potentially better match truck transport with SSL operations. With an extended harvest strategy, harvest would begin in July and extend through the following March. This extended harvest season would allow for spreading the fixed costs of the harvest machines over more Mg harvested per year thus reducing the $/Mg harvest cost. However, the earlier harvests before senescence (July through September) will remove more nutrients, and they will potentially conflict with other farm operations (i. e. grain/hay harvest, small grain seeding). To maintain productivity, additional fertilizer will be required, and this represents an additional cost to the farmer whose fields are harvested prior to October. The farmer whose fields are harvested January through March will have a lower yield because of leaf loss from the standing biomass exposed to winter conditions.

A delayed harvest raises an important issue relative to the payment to a farmgate contract holder. As previously explained, a certain loss occurs when the crop is left standing in the field for delayed harvest (December through March). Also, the delayed harvest increases the risk that a weather event will cause the crop to lodge so the harvest machine leaves material, and the "yield" is reduced. If these same fields were harvested during the optimum window (October through December) and stored longer in an SSL, there is a storage loss that increases with time in storage [18]. Considering the total reliance on stored biomass during the non-harvest months, which strategy is the better choice for the farmgate contract holder (i. e., generates the maximum quantity of biomass to sell to the biorefinery)? Is it better to delay harvest or build a larger SSL and invest in more harvesting machine capacity to complete the entire harvest within the 3-month optimum window? Epplin et al. [19] estimated differences in switchgrass harvest costs for both a 4-month and a 8-month harvest window. They accounted for differences in biomass yield across months, but they did not adjust for fertilizer requirements. They found that the estimated harvest costs varied from $25 per Mg for a 4-month harvest season to $11 per Mg for a 8-month harvest season. These results show that the length of the harvest season is a significant factor in determining the costs of harvesting biomass.

Epplin et al. [19] did not have refined estimates of the number of days that biomass could be harvested. They based their estimates of available harvest days on a study designed to determine the number of days that farmers in Southwestern Oklahoma of United States could conduct tillage operations. To determine more precise estimates of the number of harvest machines required to harvest, and thus obtain a more precise estimate of harvest costs, a more precise estimate of the number of harvest days is required.

Hwang et al. [20] determined the number of suitable field workdays in which switchgrass could be mowed and the number of days that mowed material can be baled. Empirical distributions of the days available for mowing and for baling switchgrass were determined for nine counties in Oklahoma. Distributions were determined for each month and for two potential harvest seasons (short, October-December and extended, July — February). Several conditions are necessary for safe baling of switchgrass. First, the soil must be sufficiently dry to support the weight of harvest machines. Second, prior to baling, the moisture content of the cut switchgrass must be at a level for safe storage. They provided evidence of the difference in harvest days across strategies.

Determining the time available for required harvest operations is a necessary, but not sufficient, prerequisite for determining the optimal harvest strategy. Additional information will also be required. As previously explained, if harvest begins in July prior to maturity and subsequent senescence in October, additional fertilizer will be required (the amounts are undefined) to offset the nutrients removed from the pre-senescence harvest. If harvest is delayed into December or later, the harvestable yield will be less (the amounts are undefined) than if it is harvested in October and November. Thus, the information regarding differences in harvest days across months must be combined with information regarding differences in fertilizer requirements, and biomass yields across harvest months, to determine an optimal harvest strategy.