Summary and conclusions, Miljøstyrelsen

Biogas potential of organic household waste - basic documentaiton

Summary and conclusions

Source separated organic household waste from multi-family and single-family residential areas in 5 Danish towns and cities (Grindsted, Copenhagen, Kolding, Vejle and Aalborg) was sampled twice during an 11-month long period and pretreated by different mechanical technologies: magnetic separator, disc screen and screw separator. In addition, the seasonal variations have been monitored by six samplings in selected residential areas. The pretreated organic waste (called biomass) and the reject have been characterized physically and chemically, and the biochemical methane potential measured in the laboratory during 50 days. Furthermore, 14 of the biomass samples were anaerobically digested in a pilot-scale digester and the methane yield determined in each case after stable digestion was obtained. The digested biomass was afterwards characterized in terms of chemical composition and residual methane potential.

The composition of the source separated organic household waste varied among the residential areas investigated. The compositional variations seemed to correlate with the sorting criteria applied and the type of collection bag used: Cat soil, potted ornamental plants and alike included in the organic sorting criteria correlated with increased non-volatile solids (ash), and the use of plastic bags for collection of the organic fraction increased the plastic content of the organic fraction in excess of the contribution by the actual collection bags. I a few cases the plastic content was very high (>10%). Considering the actual organic waste in the source separated waste (determined as loss on ignition minus the plastic content) no systematical differences were observed between multi-family and single-family sources, between different sorting criteria, or between different collection systems. However, it should be noted that substantial variation in the composition might have masked minor systematical differences.

Among the pretreatment technologies, the magnetic separator in all cases transferred the largest fraction of all parameters to the biomass, since the reject made up only 1% of the mass. Pretreatment by magnetic separator, however, is only feasible if the source separated organic waste is very clean, as was the case for waste collected in Grindsted and Copenhagen. In contrast to the low reject fraction by the magnetic separator, the reject was 34% for the disc screen and 41% for the screw separator based on wet weight. The pretreatment efficiencies varied significantly revealing a relative standard deviation of 10-15%. No systematic differences in pretreatment efficiency were observed with respect to residential area or waste collection system. However, it is important to note that the fraction of dry weight in the biomass varied substantially in time.

The biomass yielded by the screw separator was very clean although small pieces of plastic still could be identified. By weight, however, the plastic content was < 0.5%. The disc screen caused more plastic and larger pieces of paper in the biomass. The reject consisted in both cases primarily of organic matter, typically 90-98 %, although occasionally only 80-85% when the plastic content was high. Other objects in the biomass constituted less than 1%.

The effect of the pretreatment technology varied substantially as to how much of a constituent in the organic waste was transferred to the biomass. This variation was partly related to the waste composition and partly to the pretreatmnet technology. However, no systematical difference between disc screen and screw separator was found with respect to the mass of degradable organic waste in the biomass. In average about 50-55% of all components were recovered in the biomass. However, in the case of waste from Copenhagen and Vejle, the disc screen distributed more of several components to the biomass than was the case for the screw separator.

The composition of the biomass for a defined system (residential area, collection system, pretreatment technology) varied in time depending on the parameter in question. For the most important parameters the relative standard deviation was 3-15%. Only the ash content of the biomass varied among the residential areas, being highest in Kolding and Vejle (15.0-16.7%) and lowest in Copenhagen (6.5-11.2%) and Grindsted (10.0%). The biomass consists of 22-32% dry matter, 83-93% organic matter (VS), 10-14% crude fat, 13-15% crude protein, 10-16% starch, 4-10% sugars and 16-24% crude fiber. The measured components make up in average 80% of the organic matter, and it is assumed that remaining part is "other carbohydrates". The main factor affecting the biomass composition is the pretreatment. In general, the biomass from the hydraulic separator, compared to the biomass from the disc screen, contains more water (relatively 7-20% less dry matter, TS), more crude fat (relatively 10-20% more), less crude fiber (relatively 22-40% less) and more EDOM (EDOM: Enzyme-degradable-organic-matter, constitutes 97-99 % of VS for the screw separator compared to 87-94% of VS for the disc screen) and finally less P (relatively 50% less).

The biochemical methane potential determined over 50 days for the organic household waste in the biomass was in average 465 Nml CH₄/g VS. The determinations revealed some variation, but no systematical differences between residential areas, between multi-family and single-family sources, nor between the different pretreatment methods were found. Methane potentials calculated on the basis of the components or the chemical composition gave, as expected, higher values than the actually measured potentials, but no correlation was found between theoretical values and measured values.

The organic matter in the reject is fundamentally not different from the organic matter in the biomass and did also reveal substantial methane potential, although based on VS, about 25-40% less than the potential in the biomass.

The methane yield of biomass from source separated organic household waste was for 14 samples measured in a pilot-scale digester. The methane yields measured were 300-400 Nml CH₄/g VS, averaging 340 Nml CH₄/g VS and the biogas containing 62% methane. The variation observed was not systematical and no correlation was found with methane potentials measured in the laboratory or with theoretically estimated potentials. The degradation in the digester was 74 - 89 % with respect to VS in the biomass, averaging about 80%. The digested biomass has a potential for additional 40-50 Nml CH₄/g VS originally supplied to the pilot-scale digester corresponding to in average 10-15% additional methane.

Savings in energy, global warming potential and nutrient recovery from source separated organic household waste were modelled for a range of scenarios with different sorting criteria, collection system, pretreatment, digesters and including incineration of the reject. Models were also made considering only incineration of the organic waste. Transport, process energy, energy production as well as substitution of artificial fertilizers are considered in the models.

Savings in energy by digestion of the organic household waste is independent of the pretreatment technology and in general not very different from the savings obtained by incineration of the organic household waste from Grindsted, Copenhagen, Kolding and Vejle, while there is a minor advantage (ca. 9%) in the case of waste from Aalborg.

The digestion of the biomass and the incineration of the reject contribute equally to the production of energy when both systems are operated with power and heat production. The largest saving in energy is obtained when the dry matter is recovered in the reject and the water in the biomass. The savings in energy by substituting artificial fertilizer and the energy used on collection and transport of the waste each corresponds to about 10% of the energy obtained in the system. This suggests that optimisation of the energy savings by digestion of organic waste should focus on optimising the gas production in the digester, the gas utilization and the incineration of the reject.

The overall saving in energy is not very sensitive to changes in the technological system:+ 7 % energy is obtained if the reject constitutes 7 % in stead of usually 30 – 44 %, +5% energy is obtained if no extra fuel is used by having separate collection, -9% energy is obtained if the hauling distance to the digester is increased from 25 km to 150 km, and +9% energy is obtained by a 13 % increase in the biogas production. The crucial issue is in all cases that efficient energy savings require that both electricity and heat are produced: If the gas engine produces only electricity and the heat is not utilized, then the energy savings are reduced by 23 %.

Recovery of N, P and K does not exist by incineration, but by digestion each ton of wet source separated organic household waste contributes with 5-7 kg N, 0.5-1 kg P and 1.5-2 kg K for most of the systems applicable to Copenhagen, Kolding, Vejle and Aalborg. In Grindsted, where the waste is very clean and a magnetic separator is the only pretreatment, about twice as much is recovered in terms of nutrients, since the reject is negligible.

The investigation revealed large geographical and seasonal variations in waste composition, pretreatment efficiencies, methane potentials and in methane yields. However, the large number of samples involved and the extensive characterization performed suggest that the evaluations and conclusions made reasonably well represent typical Danish conditions regarding source separation and digestion of organic household waste.