BioCycle April 2005, Vol. 46, No. 4, p. 57
Study at the Edmonton cocomposting facility compares amount of foreign matter in finished product after passing compost through a ball screen, single trommel and double trommel.
Natasha Page, Jerry Leonard and Grant Clark

THE EDMONTON, CANADA Waste Management Centre (EWMC) is composting 200,000 tons of residential municipal solid waste and 22,500 dry tons of biosolids annually – yielding approximately 125,000 tons of finished product. Local residents separate recyclables for the Blue Bag program, and household hazardous waste for local Eco-Stations. The remaining residential MSW is processed in the cocomposter.
Foreign matter (FM) is the undesirable material remaining in the final product that results from composting mixed MSW. FM refers to man-made, inert particles such as glass, plastic and metal. Generally, the amount of FM is determined by the mass of nonbiodegradable material found in the “overs” at a specified screen size, expressed as a percentage of that fraction of compost. Most procedures include drying a mature compost sample, sieving the material, and sorting the “overs” from a specified sieve size.
A study was conducted to compare screening methods to refine the MSW compost. The purpose of this study was to determine whether passing compost through a ball screen, an 18.75 mm trommel screen, or twice through an 18.75 mm trommel screen would lead to the lowest levels of FM. These results were compared with a previous study to determine any changes from the previous levels of FM in the EWMC compost.
DESCRIPTION OF COMPOSTS
The composts in this study were created from MSW and yard waste collected from single family dwellings and processed through the cocomposter at the EWMC.
The types of potential FM found in the compost were divided into the following categories:
(Rock and other naturally occurring materials not included as FM:)
Metal – Any material that is shiny and metallic in appearance and/or resembles metal foil.
Glass – Particles are translucent or clear, brittle, and have obviously sharp edges. They are sharp enough to scratch a plastic petri plate.
Plastic Film – Particulate remnants of plastic bags, sheets, and wrap. Film particles are generally flexible, translucent, and range from 0.05-0.8 mm in width. These particles are also sorted into three different sizes: 0.5 to 2 mm, 2 to 12.5 mm, and 12.5 to 25 mm.
Plastic Chip – Nonshiny, nonearthy colored particles with a rather rigid structure, are generally opaque, and are greater than 0.8 mm in width. Sponge like material is also classified as plastic chips. Many of these particles will be a light green color. These particles are also sorted into three different sizes: 0.5 to 2 mm, 2 to 12.5 mm, and 12.5 to 25 mm.
Manufactured Fibers – All fibers that are not earthy brown, have a constant diameter and shape from end to end, and are over 10 mm in length are considered manufactured. These fibers are long, thin nonbiodegradable particles with a constant diameter no greater than 0.10 mm. Strands of fiberglass are included in this category.
Five composite samples of 10L each were collected from each of the three following treatments: ball screen, one pass; trommel screen, one pass; trommel screen, two passes. From each 10L sample, four 500g samples of each compost type were collected and placed in a 2L polyethylene bag. Each bag of compost was shaken to ensure complete mixing, and one 10g sample was removed to dry for 24 hours at 65°C to determine the initial moisture content. The dried sample was placed in a 95 mm diameter petri plate and hand sorted under a 20x light microscope to visually identify all FM fractions: metal, glass, plastic film, plastic chip, and manufactured fiber.
SORTING THE COMPOST
Three different sorters (two males and one female) participated in this study. They were all instructed in the methodology to be used, but no attempt was made to test or qualify them for sorting before the study commenced. Due to logistical and scheduling problems, each sorter processed a different number of samples from each type of compost. However, for each treatment, at least six samples were processed by each of two sorters.
Each dried compost sample was placed in a petri plate under a microscope and examined at a magnification of 20x. A pair of forceps was used to collect all FM particles from the petri plate, except for manufactured fibers. The different types of FM were collected in preweighed, 20×30 mm plastic bags. Each type of FM particle was sorted to respective plastic bags for collection and weighing. Organic material adhering to the particles was gently removed by hand to avoid an inflated percent of FM. Plastic film and chips were further sorted into the size categories of 0.5 to 2 mm, 2 to 12.5 mm, and 12.5 to 25 mm. Material that was folded upon itself was not manipulated to determine its size if fully stretched. Material smaller than 0.5 mm was not included in the analysis.
To efficiently cover the entire area of the petri plate, compost sorters followed a search pattern to work from one side of the plate, and move up and down in a regular pattern until the entire area was covered. After completely searching through the sample, the plate was turned 90°, and the search pattern was repeated. All the FM particles that were found were collected and sorted into the respective bags. The mass of FM from each compost sample was determined by measuring the total weight of the polyethylene bags and the FM particles inside, minus the weight of the bag.
Particles composed of film plastic or metal foil were often folded upon themselves into an accordion shape, and were measured in that form. Therefore, the stretched length of the particles may have been larger than what was actually measured. This may have caused some error in determining the mass of material for each size category, although the overall mass of each material was accurate.
RESULTS AND DISCUSSION
The ball screen and two passes with the trommel screen were the most effective at decreasing the level of FM in the compost, but were not significantly different from each other. The overall percentage of FM ranged from 1.80 percent for the ball screen, to 2.72 percent for one pass with a trommel screen, and 1.75 percent for two passes with the trommel screen. In all instances, the percent of FM does not meet the minimum voluntary standards for compost quality. Although a statistically significant difference was not always measured, the compost that was refined through the ball screener generally had lower levels of all types of FM, and significantly lower levels of all plastic.
Significant differences were found between the percentage of plastic chips (dry weight) from 2 to 12.5 mm in size. One pass from the trommel screen resulted in 0.754 percent plastic chips, while the ball screen decreased the percentage of plastic chips to a significantly lower percentage of 0.260 percent. Two passes with a trommel screen resulted in 0.530 percent plastic chips, which did not significantly differ from the ball screen or one pass with the trommel.
An average of 31.1 manufactured fibers per gram dry compost was found in the compost treated with the ball screen. This was significantly higher than the 25.3 fibers per gram compost found in the compost treated with one pass through a trommel screen. Two passes with a trommel screen resulted in 28.5 fiber per gram compost, which did not significantly differ from the ball screen or one pass with the trommel.
A bias between sorters was found for the categories of plastic chips and manufactured fibers. This is of obvious concern, as it calls into question the repeatability of the FM sorting methodology. Furthermore, the significant differences in the FM may be a result of the sorter bias, rather than the type of compost screening method used. This brief study cannot fully explain the differences in the sorters, nor can it fully attribute the differences found in FM levels solely to the person who happened to work with a specific sample. However, further investigation is needed to fully quantify the differences in sorting due to the person, and to see if additional training, testing, or qualifying procedures could negate the statistical differences.
The percent FM in the City of Edmonton MSW compost was previously reported to range from 2.43 to 2.17 percent. Therefore, processing with the ball screener or two passes with the trommel screener created compost with a lower percentage of FM.
Natasha Page is a Pesticide/Waste Specialist with Alberta Environment based in Edmonton, Canada. Jerry Leonard, a professor of Bioresource Engineering in the Faculty of Agriculture, Forestry, and Home Economics, is now “on loan” to the Edmonton Waste Management Centre of Excellence. Grant Clark is a Research Associate in bioresource engineering in the Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Canada.