Many of the costs associated with food waste disposal are highly dependent on the amount of food waste generated. It is therefore necessary to obtain information regarding the amount of food waste produced to allow for an accurate assessment of cost. In May 1999, Dennis Easley (Superintendent of Grounds) calculated the cost of five methods of waste disposal using data extrapolated from a St. Olaf food study in a summary of kitchen waste options. The report assumes that each student creates 1.5 pounds of waste in one meal, but these are rough estimates and are highly uncertain. Consequently, calculations were also made in the formal report assuming .75 pounds per person per meal, reducing the waste stream by one half. Data from Endres Processing, a company that reprocesses food waste into animal feed, estimates restaurant style food waste at .4 pounds per person per meal. In addition, a food audit performed at each of the 16 dining halls at Harvard University found students created an average of .25 pounds per person per meal of post-consumer food waste (total food waste is then assumed to be .5 pounds per person per meal).

Due to the great deal of inconsistency in these estimates, we deemed it important to measure how much food waste Carleton students produce through use of a food audit. Food waste is usually broken down into two categories: pre-consumer, which includes all preparation scraps as well as any leftovers which must be discarded, and post-consumer waste, which is the food left on students' trays when they are sent to the dishroom. The Carleton Food Audit examined pre-consumer and post-consumer wastes. Both are usually assumed to generate fairly equal amounts. Thus, according to Endres Processing estimates there are roughly .2 pounds of pre-consumer waste and .2 post-consumer per person per meal. Similarly, by St. Olaf estimates, each meal served results in .75 pounds of pre-consumer and .75 pounds of post-consumer waste.

All data collected is from Burton Dining Hall. This facility serves approximately 1500 meals each day, similar to the expected capacity of the new dining hall. To gather post-consumer data, students collected all of the food waste (edible and non-edible) returned to the dish conveyor belt. At the conveyor belt, food was scraped into a bucket before the trays were sent to the dishroom. This waste was separated into liquid and non-liquid waste, and each was weighed. To track the number of meals served, the number of trays was also recorded. Additionally, every tenth tray was weighed to measure variance in food waste. On the second night of data collection, a slightly different method was employed. In response to student comments that people were leaving less waste than usual when they saw that it was being measured, the waste collection took place "behind the scenes" in the dishroom. Surprisingly, there was little variance in the data. On the first night, students left an average of .23 pounds of waste on their plates (138 pounds on 600 trays). On the second night, this number was .22 pounds (145 pounds on 660 trays). The sampling of every tenth tray was higher than both of these values, at .61 pounds per tray sampled per meal. Because at times there were several trays being scraped at once, it was difficult to determine which trays qualified as the "tenth;" as a result, we may have been biased toward plates with more food waste.

The food audit also measured pre-consumer waste. Over a period of two days, Marriott workers collected food preparation scraps and leftovers in several garbage containers. The food collected included all waste from Friday (breakfast, lunch, and dinner) and all from Saturday up to dinner (breakfast, brunch, and dinner preparation but not leftovers). The food was weighed on a platform scale at the Cannon Valley Co-op. A total of 480 pounds were collected. Because the collection days were Friday and Saturday, the data may be slightly different from that produced in a normal week. Fewer people eat in the dining halls on Friday nights and on Saturday, and fresh fruit is served on Saturdays. Because there would be less preparation waste if less food was prepared and because the fruit waste is significantly heavier, Tony Sobiech, Sodexho-Marriott Executive Chef, postulated that these two factors would balance out to an average day of pre-consumer food waste. If this is the case, each meal served generates about .16 pounds of pre-consumer waste per person.

Based on these calculations, we are able to assume that the amount of waste produced for lunch and dinner is roughly .4 pounds per person per meal. The waste would theoretically be less for breakfast. These numbers give us a reasonable range to use in assessing the costs associated with various disposal methods.