Brownell and Liu [66] developed a computer model to select the best option, or the best combination of options, for supply of a given bioenergy plant. The options include a direct chop and delivery option, two bale options (round and large rectangular), and a variation of the large rectangular bale option where these bales are compressed at a stationary location before shipment. Specifically, the options are: (1) loose biomass harvest using a self-loading forage wagon; (2) round bale harvest; (3) large rectangular bale harvest; and (4) large rectangular bale with compression before hauling.

The results of the model provide users a decision matrix which shows the optimized handling scenario for all four handling options analyzed by this program. Cost calculation was based on three plant demands (2,000, 5,000, and 10,000 Mg DM/d) to compare costs as the required production area expanded. Satellite storage locations were established based on the distribution of production fields and the existing road network.

The key constraints used for the simulation are:

1. Only loose biomass can be directly hauled from the field to the biorefinery.

2. Biomass cannot be stored at the plant, 20,000 tons (18,144 Mg) will "stand in field" in surrounding area of the plant and collected on demand as loose material.

3. Baled biomass will be transported to and stored at satellite storage locations.

4. Large rectangular bales may be compressed before transportation.

5. Biomass will be grown on one third of the available acreage, with an average yield of 3 Mg/ha (15% average moisture content).

The land close to the plant is more valuable, from the biorefinery viewpoint, for the production of feedstock. The study found that a plant can afford to pay more for feedstock (landowner gets a higher price) if a farm happens to be closer to the chosen plant location.

The results showed that the size of SSLs is sensitive to the amount of material demanded by the plant. The program tests various sizes, distances between SSLs, and possible overlaps of production fields "served" by each SSL. The programming optimizes the simulation by slowly changing the size and location of SSLs. The program solves for the maximum amount of acreage available in the simulation model. The output is a distance material in each SSL can be efficiently hauled based on if it is harvested and hauled directly from the field or if it is baled and stored before hauling. The program also solves for the lowest cost for a set acreage when the closest fields are field chopped and remaining fields are baled.