BioCycle September 2003, Vol. 44, No. 9, p. 42
Oregon Department of Environmental Quality study finds that, with proper management, meat discards can be composted just as safely as other components of mixed food residuals.

DISCARDED food – generated daily by millions of households, institutions, restaurants and other businesses – comprise one of the largest identifiable categories of materials still going to waste in the United States. As reported in recent issues of BioCycle, solid waste agencies are increasingly targeting this category for recycling since it makes up approximately 11 percent of the MSW stream by weight. While it is generally agreed that composting is an effective strategy for managing discarded food, a persistent concern for human pathogen transmission remains a key barrier to its widespread implementation.
Some communities have forged ahead in their commitment to recovering and composting mixed food discards, while in others, significant regulatory or operational barriers still exist. Perhaps the biggest barrier to mixed food residuals composting is the popular adage that meat, cheese and dairy should not be included in composting operations. A recent study funded by the Oregon Department of Environmental Quality at a commercial composting facility in Oregon sought to dispel that myth. Overall, the study found that meat discards actually decompose more readily and just as safely as other components of mixed food discards if careful attention is paid to feedstock ratios, composting methodologies, and pathogen reducing practices.
Human pathogens can survive in a number of environments, including soil and water, utensils, the gastrointestinal tract, skin or hides of humans and animals, and even in the air or dust. It is generally agreed that the internal tissues of healthy slaughter animals are free of bacteria at the time of slaughter. However, by the time meat products reach the retail environment, varying numbers and types of microorganisms may be present. These microorganisms have been transmitted to the meat by ‘cross-contamination,’ e.g. from knives and other tools used at the abattoir, from the air, from storage containers or from food handlers. Fruits and vegetables also can be contaminated with pathogenic organisms through field application of manure or during food processing.
Human pathogens include specific bacteria, molds, yeasts, and protozoa. Some strains can cause health problems when ingested, including diarrhea, cramps, and death in extreme cases. While humans are routinely exposed to pathogenic organisms, the potential for infection is based on the number and type of pathogens present, and the overall health of the infected individual.
Concern for pathogen transmission is one of the primary reasons why composting of biosolids has been studied so extensively, resulting in the development and application of EPA’s 40 CFR Part 503 regulations. The 503 regulations require that the final biosolids product meet pathogen reduction requirements via testing for indicator organisms – either fecal coliforms (<1000 MPN per gram of total solids) or Salmonella (<3 MPN per 4 grams of total solids). These standards are widely applied to nonbiosolids compost products as well, regardless of feedstock. Indicator organisms are used because they exist in similar environmental conditions as other human pathogens, and are relatively easy to count and measure. The assumption is that if fecal coliforms and/or Salmonella are eliminated in an environment, it is expected that other human pathogens have been destroyed as well. Studies consistently demonstrate that the time-temperature relationship used in the Process to Further Reduce Pathogens (PFRP) effectively destroys pathogens in compost. BACKGROUND OF COMPOSTING FACILITY STUDY In Oregon, 'green feedstocks' are defined by the state Department of Environmental Quality (DEQ) as those materials low in and unlikely to support human pathogens, including vegetative discards and, ironically, animal manure. Mixed food residuals that include meat or animal by-products are considered a 'non-green feedstock,' defined as those materials high in and likely to support human pathogens. In order to accept feedstocks typically present in an all-encompassing municipal organics stream - a small portion of which would be considered 'nongreen' - current Oregon 'green feedstock' composters would be required to obtain or upgrade their permits to comply with a much more stringent set of regulations. The requirements for siting and operating a fully permitted type of facility in Oregon - as in many other states - are cost prohibitive for many composters. Typically, these requirements may include having fully lined sites, totally covered or enclosed processing facilities, an extensive hearing process prior to opening the facility, and extensive product and/or feedstock testing. There is one 'full permit' facility in Oregon but it is currently nonoperational. [Editor's Note: On August 28, 2003, staff of the Oregon DEQ held a public hearing on a proposed composting permit for ThreeMile Canyon Farms, located approximately 18 miles west of Boardman, Oregon. DEQ proposes to issue a full composting permit to this site for composting green and nongreen feedstocks. The permit allows composting of the following feedstocks: dairy manure and dairy mortalities from the farm; yard trimmings from Spokane, Washington; Portland Metro food residuals; and wood from the Potlatch Tree Farm Plantation. Compost made from dairy mortalities and Portland Metro food residuals would be used solely as crop fertilizer on the farm. A report on the permit as well as composting operations at ThreeMile Canyon Farms will be published in a coming issue of BioCycle.] With Oregon's three largest cities (Portland, Salem and Eugene) exploring how to divert food discards from disposal, and several processors experimenting with animal mortalities and postconsumer mixed food residuals composting, Oregon DEQ has faced increasing scrutiny of its policy that treats 'green' and 'nongreen' feedstocks differently. In fall 2003, DEQ is scheduled to begin the process of updating its operating standards for composting facilities. In conjunction with several project partners, the DEQ-funded study described in this article was subsequently conceived to provide information and address pending issues for the upcoming regulatory review process. Key project participants and the services provided include: the city of Eugene, providing feedstock sourcing and collection services; Rexius Forest By-Products, the compost processing site and services; Tetra Tech, Inc., project design and analytical services; and Oregon State University, monitoring, sampling, and analysis of compost stability issues. According to Marti Roberts-Pillon of DEQ, the results of this study will be among several factors considered by the DEQ when updating its compost facility permit rules. PROJECT GOALS The primary objective of the project was to document how mixed food residuals that contain nongreen feedstocks can be composted thoroughly and safely within reasonable operating parameters. The main components of the project were: Literature Review: Oregon DEQ requested that the consultant begin the field project with a literature search to identify and examine existing research on composting mixed food discards. Process Comparison of Two Low Technology Composting Methodologies: Limited turn windrow (LTW - one turn/week) and passive aerated windrow (PAW) processes were chosen for the project, being the most prevalent methods used in Oregon. The goal was to closely evaluate and document the ability of both methods to compost commercial mixed food residuals (including meat) and achieve adequate pathogen reduction (per Part 503 regulations). The process comparison was conducted twice to assess the potential affects of two distinct weather periods (wet winter vs. dry summer). Establish Stability Test Criteria: Because of the concern for pathogen regrowth, the project included an evaluation of different stability test criteria that could be used by DEQ to determine potential for human pathogen regrowth following PFRP. The intent of a reliable stability test is to show the degree to which human pathogen regrowth potential is eliminated due to depletion of substrate. Regulatory Assessment And Recommendations: The project consultant was asked to provide guidance to DEQ with respect to regulatory criteria for composting of mixed food discards, based on the findings of the process comparison and the literature review. LITERATURE REVIEW As composting regulations often include requirements for meeting the sanitation standards developed for biosolids composting, of special interest were previous studies that evaluated the applicability of federal regulations (EPA 40 CFR Part 503) to feedstocks that include mixed food discards. Somewhat surprisingly, the literature search revealed that prior to this study, the composting of mixed food discards had not been a scientifically scrutinized undertaking. While there have been studies on the collection of mixed food discards and of animal mortality composting (BioCycle, October, 2002), there is a dearth of information as to the fate of pathogens that may be present in feedstocks generated as part of a municipal solid waste stream. Also lacking is information as to whether or not low populations of pathogens can regrow to contaminate material that already has met pathogen reduction requirements. The document that resulted from the literature search (Research Concerning Human Pathogens and Environmental Issues Related to Composting of Non-Green Feedstocks, August 2001, E&A Environmental Consultants, Inc.) concluded that properly designed and operated nongreen feedstock composting facilities should not present a public health or worker health threat. The document also identified specific research needs to be considered in the design of the study: Better data on pathogen destruction in nongreen feedstocks during composting; Data on the destruction of animal pathogens in animal carcasses during composting; Development of sampling methods for pathogens in large animal carcasses; and Evaluation of low technology composting methods (e.g., passive aerated piles, low-tech windrows and stacked piles) as to the effectiveness of pathogen destruction during composting when using nongreen feedstocks. MATERIAL SOURCES AND COLLECTION The literature search was followed by two eight-week composting periods in Eugene, Oregon (pop. 135,000). The first phase began in November, 2001 and the second began in March, 2002. The city of Eugene was responsible for delivering food discards that contained meat to the site for incorporation into two windrows. In order to mimic what might happen during actual processing, the consultant hired to oversee the project required that 20 tons of food be delivered to the processor at a specified date and time. The city worked with three of its licensed haulers to collect source separated food residuals from accounts that each hauler identified as generating significant quantities of mixed food discards. Generators included supermarkets, specialty meat markets, restaurants, a University of Oregon cafeteria, a chicken processing plant, and a slaughterhouse. The idea was to ensure that pathogen levels in the incoming feedstocks would be high. Contamination by noncompostables at each generator site was addressed by placing large signs on the containers utilized by the garbage haulers to collect food residuals. A simple approach of YES and NO was employed. City staff visited with some of the generators to discuss the program with employees and management. An introductory letter was sent, along with periodic updates and a final survey of the generators. While this study didn't focus on collection issues, the mixed food discards were easy to collect because generators were not required to sort food by type, and only the very obvious materials such as glass, plastics, ceramics, and metals needed to be excluded. Collection began two weeks before scheduled delivery to the composting site. Two drop boxes were placed at Eugene's largest transfer station for temporary storage of the collected materials. To contain any liquids, approximately five yards of sawdust were layered in the bottom of each drop box. Subsequent layers of sawdust on top of the food discards were added to minimize odors during the two weeks of storage. This approach worked well, with no adverse odors or leakage reported. On the delivery day for each study period, the drop boxes were tipped at the composting site. Given the composition and layering of material in the drop boxes at the transfer station during the two-week collection period, the composting process already was underway. Additional materials were delivered to the site directly on pile building day, including nine tons of chicken offal, various volumes of mixed food discards that the haulers collected that day, and small quantities of feedstocks collected directly by city staff. During the first phase of processing, 21 tons of material were delivered. During the second phase, 24 tons were delivered. Due to the nature of this study - to monitor pathogen reduction in a low-tech composting environment - incoming feedstocks were selected for a higher meat component than would be expected in a typical municipal collection program. Of the 24 tons of material collected during the second period, for instance, approximately 16 tons were either offal from a chicken processor, sheep butcher, fish market, or were out of date meats, poultry, and seafood from a supermarket. The feedstocks in period one were slightly less concentrated with meats, although there were more cooked and processed meats. Characteristics of the feedstocks processed are shown in Tables 1 and 2. PROCESSING Once the feedstocks were delivered to the site, they were mixed with bulking materials and carbon sources to create the desired pile characteristics. The goal was a C:N ratio of approximately 20-25:1 and a moisture content of about 65 percent. Sawdust and ground yard debris were used as a carbon source, with additional volumes of steer manure added to ensure fecal coliform contamination. Feedstocks were mixed using a standard front loader, with no pregrinding of the mixed food discards. Grinding was not considered an option due to the concern for contaminating materials that would be ground subsequent to the mixed food discards. One of the protocols established for the study was the sanitizing of equipment that might come into contact with feedstocks that were not part of this study as well as minimizing the chances for contamination even between the two windrows studied. After mixing the bulking agents with the delivered food discards, the blended materials were then divided into two piles representing the composting methodologies to be examined. The first pile was characterized as a 'limited turn windrow' (LTW). This was a conventional static pile containing approximately 110 cubic yards of material. It was turned at two weeks and weekly thereafter. The second pile was built as a 'passively aerated windrow' (PAW), also with a volume of 110 cubic yards but constructed using perforated pipes laid every two feet at a height of approximately two feet above the liner. It was turned at four weeks, and then not turned again. Both piles were capped with approximately eight inches of sawdust in order to decrease the probability that birds, rats, dogs and other vectors would be attracted to them. FIELD MONITORING AND SAMPLING A comprehensive monitoring strategy was designed and implemented throughout the study that involved a combination of field measurements and sample collection. Field monitoring consisted of visual observations and measurements of both oxygen and temperature. Qualitative observations were made daily of odors, vectors or other remarkable conditions. Temperature and oxygen levels were measured daily ('routine' monitoring) at four locations within each pile (six locations during Period 2). A more extensive profile of oxygen and temperature was obtained weekly ('intensive' monitoring), using four cross sections through each pile. At each cross section, readings were taken at three different heights along the cross section and at three different depths within the pile. Weekly samples were taken for pathogen and stability testing. Composite sampling (mixing several subsamples from different areas of the pile) was used for pathogen samples; grab sampling (each sample from a single location) was used for stability samples. Leachate samples were taken for the first six weeks, but a subsequent decrease in moisture content by week eight of each period rendered insufficient volumes of leachate to sample. Final product analysis was performed at the end of each phase (week 14 of both periods) for public health parameters (Salmonella and fecal coliforms), stability parameters (CO2 respiration, Solvita stability, organic carbon and TKN), and an array of additional product quality parameters (total solids, conductivity, sieve analysis, pH, ammonium, nitrate, plant nutrients, and trace metals). RESULTS, PERFORMANCE Performance of the two composting methods provides a basis for determining suitability as approved processes for composting nongreen feedstocks in Oregon. While both processes produced successful results in the context of this study, additional demonstration of their performance over a wider range of operating conditions is required for further validation. Key operating parameters that influence the composting process are weather, initial mix, moisture content, time and temperature. Success was measured in this project by pathogen reduction, leachate quality, stability and product quality. The following is an abbreviated overview of the results and analysis provided in the project report. Readers may view the data and full analysis in the final report referenced at the end of this article. Moisture: Period 1 weather was very wet when piles were constructed and for two days following pile construction. At the beginning of Period 2, significant rainfall occurred over the week prior to and on the day of feedstock mixing and pile formation. Otherwise, rainfall was less than normal for the duration of both Period 1 and Period 2. The wet weather at the beginning of both periods may have affected the initial moisture content of the piles. The increasing moisture content of both LTW and PAW processes during Period 1 indicates a need for caution when rainfall exceeds normal or when feedstocks do not include a significant fraction of high energy materials. These conditions could limit the ability of the PAW and LTW processes to achieve and maintain temperatures. In addition, the reduction in moisture content during Period 2 indicates a need for caution during summer months and when a lot of high energy materials are included in the mix. Overall, both processes provided acceptable performance for the weather conditions experienced during the test periods. While not a 'worst-case' situation, the conditions during the project show that these processes can perform in non-optimum conditions. Temperatures: The PAW process very quickly achieved pathogen reduction temperatures (>55°C) and consistently maintained high temperatures throughout the composting period. The LTW process also achieved adequate temperatures, though not as quickly as the PAW, and saw expected temperature reductions and rebounds after turning. Excessively high temperatures were observed in both piles (as high as 165°F/74°C). The protein-laden feedstocks appear to have contributed to the high temperatures, as bacteria quickly reproduce in the presence of food that meets their immediate energy needs. Within two weeks of pile construction, most evidence of any meat products was gone. Materials such as potatoes and citrus skins took longer to decompose, as they are chemically more complex.
The daily temperature (averaged from the four routine monitoring points) for each of the four test piles was calculated. A comparison of average temperatures for the LTW piles indicates a rise over 21 days to 65°C with a continual increase to almost 70°C through day 45. The PAW piles heated to 65°C in 10 days and 70°C in 14 days, then ranged between 70 and 75°C through most of the composting process.
Figures 1 and 2 provide similar plots for maximum and minimum temperatures. The maximum temperature plots indicate that the Period 2 piles generally had hotter zones than the Period 1 piles. The PAW Period 2 pile in particular consistently had zones where the temperatures ranged between 75°C and 80°C. The minimum temperature plots show more rapid early heating at all points in the PAW piles and maintenance of warmer minimum pile temperatures throughout the composting period until day 42.
Pathogen Reduction: The intensive temperature data for both piles during Periods 1 and 2 indicate that all met both the pathogen reduction and vector attraction reduction temperatures for much longer than required by 40 CFR Part 503. Pathogen indicator data (Tables 3 and 4) show that all four piles (two systems, two seasons) had fecal coliform levels below 1,000 MPN/gram dry solids within the first five weeks of composting. During Period 1, in both systems, Salmonella levels remained above 3 MPN per 4 grams dry solids until the seventh week. During Period 2, Salmonella were below the limit within the first five weeks of composting.
In all four test piles, either Salmonella or fecal coliform levels increased later in the process. The indicator organism data show evidence of regrowth or contamination with fecal coliform in both process piles during Period 1, and evidence of regrowth or contamination with Salmonella in both process piles during Period 2. The other performance data points offer no valid rationale for either increase, and all temperature data met both the pathogen reduction and vector attraction reduction requirements for much longer than required. The Part 503 regulations require that either fecal coliform or Salmonella standards be achieved in the finished product; therefore, adequate pathogen reduction was met in all test piles. However, a concern exists that these organism populations increased while operating temperatures remained high, possibly explained by undetected cooler zones. Since regrowth should not occur at the documented temperatures, contamination during material movement or during sampling could be the culprit. If truly representative of pathogen regrowth, both the Salmonella and the fecal coliform levels would be expected to rise.
Leachate Quality: There are no human pathogen standards for leachate in the Part 503 regulations. Water quality standards in the region apply to receiving waters, not leachate. Therefore, existing water quality standards only would be applicable if the leachate were to be discharged directly into freshwater or estuarine water. Leachate quality data shows the Salmonella levels throughout the composting processes during Periods 1 and 2 to be less than 0.3 MPN/100 ml. Period 1 fecal coliform levels for the LTW process vary from 24 MPN/100 ml to undetectable, and for the PAW process vary from 1100 MPN/100 ml to undetectable.
Stability: Stability data were collected for both processes during both seasonal periods. Stability parameters were carbon dioxide (CO2) evolution rate, Solvita maturity index, and Vector Attraction Reduction (VAR) time and temperature requirements.
Both the LTW and PAW processes met the Part 503 VAR requirements during Periods 1 and 2 within the first 17 to 25 process days. The regulations do not require further stabilization after VAR requirements are met; however, material meeting VAR requirements alone is not necessarily defined as stable according to US Composting Council (USCC) TMECC 05.08-B standards that are based on CO2 evolution. Additional composting and curing further ensures pathogen reduction and is required to produce USCC-defined stable and mature compost. A low CO2 evolution rate indicates low microbial activity and high stability. According to TMECC 05.08, a CO2 evolution rate ranging from 2 to 8 mg CO2-C per gram compost organic matter per day indicates a stable compost, and a CO2 evolution rate less than 2 mg CO2-C per gram compost organic matter per day (4 mg CO2-C per gram compost organic carbon per day) indicates a very stable product. Stability results for this study according to TMECC 05.08-B are shown in Table 5.
The PAW process consistently exhibited higher CO2 evolution rates, indicating less stability at all stages of the composting process during both periods. This may be the result of continuously high temperatures that limit the active population of microbes. It could also be a result of different feedstock characteristics.
An independent objective within the study was to evaluate a variety of methods to measure the CO2 evolution rate in an attempt to develop improved stability testing protocols that are simple enough for on-farm use, but sophisticated enough to be useful in compost marketing and in compost quality control programs. Researchers at Oregon State University (OSU) conducted this part of the study. The alternative alkaline trap test procedures used by OSU were a variation of the USCC standard test method TMECC 05.08-B. The OSU procedures used an incubation temperature equal to 25°C with a 2-day incubation time versus the TMECC 05.08-B method that incubates at 37°C over a 4-day period. OSU also looked at a 4-hour alkaline trap test, the Draeger tube 4-hour test (with and without refrigeration prior to respiration measurement) and Solvita CO2 4-hour test in comparison with the alternative alkaline trap procedures. Direct comparisons with TMECC 05.08-B were not performed as a part of the study. The average stability results for the OSU alternative alkaline trap 24 and 48-hour test procedures indicate that both processes during Periods 1 and 2 produced a very stable product after ten weeks of composting.
Product Quality: Product quality was measured at the completion of the composting process (week 14) by a comprehensive series of characterization analyses developed for compost by the USCC. The product quality reports (analyses performed by Soil Control Lab) conclude that both the LTW and PAW processes during Periods 1 and 2 produced a high quality product. The only parameters of concern are a low Period 2 moisture content and a high C:N ratio of each of the products, as shown in Table 6. The moisture content of the material in both the LTW and PAW processes was greater in Period 1 than Period 2. Material in both processes dried significantly during Period 2. Consistent with the analytical C:N data, the product has a visibly high woody fraction. Germination and growth analyses yielded 100 percent emergence and seedling vigor for all material at week 14.
REGULATORY RECOMMENDATIONS
The project consultant provided detailed analysis to Oregon DEQ for consideration in its upcoming review of composting facility regulations, specifically to address pathogen destruction in mixed food residuals composting. The report’s conclusions and recommendations to DEQ are as follows:
1. Both of the low technology composting methods field tested (LTW and PAW) demonstrated the ability to effectively reduce pathogen-related risk to acceptable levels during the wet weather conditions experienced in western Oregon. The test periods had lower than normal rainfall, so continued demonstration of the technology during wetter than normal conditions is advised. This can be done as part of a provisional longer term demonstration of the processes at composting facilities.
2. Proceed with development of regulations that allow and encourage the composting of nongreen classified food discards using the PFRP composting processes identified in 40 CFR 503 or the two processes (LTW and PAW) that were demonstrated as part of study. The recommendations include: a) The PFRP processes are suitable for outright approval based on history and experience; b)The LTW and PAW processes should be approved on a provisional basis with additional monitoring required at each facility to provide sufficient operating experience to justify full approval.
3. Base nongreen feedstock composting regulations on the regulatory concepts and procedures used in the federal sewage sludge regulations (USEPA 40 CFR Part 503). The recommendations include: a) Expose all material to 55°C for a minimum of three consecutive days to provide pathogen reduction; for a turned windrow process this requires five turns after each of which the three day criteria must be met to assure that all material is exposed to those conditions; b) Expose all material to an average of 45°C with a minimum of 40°C for a minimum of 14 consecutive days to provide Vector Attraction Reduction; and c) Test the product for indicator organisms, fecal coliform <1000 MPN per gram of total solids (dry weight basis) or Salmonella <3 MPN per 4 grams of total solids (dry weight basis). 4. Require that the product be stable to the standard set by the US Compost Council Test Methods for the Examination of Composting and Compost (TMECC) 05.08-B Carbon Dioxide Evolution Rate (4 mg CO2-C per gram compost organic matter per day). 5. Require that each nongreen feedstock composting facility develop plans for implementing the regulation with particular focus on achieving pathogen reduction and a stable product, runoff management and management of materials that fail to comply with pathogen reduction requirements. 6. Provide a procedure for demonstrating new low technology composting options and criteria for approving those that perform as required. CONCLUSION Composting meat-laden food discards can be accomplished as safely as composting animal manure. Meaty tissue within a healthy animal's body is absolutely free of pathogens. They are found there only after contact with cutting tools, surfaces, or human hands that are contaminated with fecal material. Once contaminated, however, pathogens can populate food residuals in significant numbers and mixed food residuals that contain offal are certainly a source of pathogens. As in biosolids composting, time and temperature regimes are effective at creating an environment inhospitable to the microorganisms that are dangerous to human beings. Whether or not mixed food discard composting can be done profitably depends on local characteristics such as the market for finished product, the competition that exists from other disposal options, the collection infrastructure, and the regulatory climate of the region. In Oregon at least, composting regulations that are based on assumptions about the ability of incoming feedstocks to support pathogens has been one barrier to the implementation of food discard collection programs. As additional and more difficult feedstocks are viewed as compostable, the composting industry will face the challenge of its dual role as municipal waste manager and producer of a quality end product. It is hoped that the results of this study will diminish the difference between those pursuits. For a full review of this study, visit the Oregon DEQ website at http://www.deq.state.or.us/wmc/solwaste/rsw.htm. Cindy Salter is a consultant and director of the Compost Tea Industry Association. Alex Cuyler is chair of the Association of Oregon Recyclers.