Socio-environmental Opportunities for Organic Material Management in California’s Sustainability Transition

Contemporary resource management is doubly burdened by high rates of organic material disposal in landfills, generating potent greenhouse gases (GHG), and globally degraded soils, which threaten future food security. Expansion of composting can provide a resilient alternative, by avoiding landfill GHG emissions, returning valuable nutrients to the soil to ensure continued agricultural production, and sequestering carbon while supporting local communities. Recognizing this opportunity, California has set ambitious organics diversion targets in the Short-Lived Climate Pollutant Law (SB1383) which will require significant increases (5 to 8 million tonnes per year) in organic material processing capacity. This paper develops a spatial optimization model to consider how to handle this flow of additional material while achieving myriad social and ecological benefits through compost production. We consider community-based and on-farm facilities alongside centralized, large-scale infrastructure to explore decentralized and diversified alternative futures of composting infrastructure in the state of California. We find using a diversity of facilities would provide opportunity for cost savings while achieving significant emissions reductions of approximately 3.4 ± 1 MMT CO2e and demonstrate that it is possible to incorporate community protection into compost infrastructure planning while meeting economic and environmental objectives.


S1.1 Data Sources
Organic materials, or feedstocks, are defined as the organic materials that will be processed in composting facilities (rather than landfills).MSW feedstocks were provided at the census-tract level, while agricultural materials were assigned to the geographic county centroid.Municipal solid waste (MSW) consists of everyday items used and disposed of, ultimately ending up in the landfill.Our study was concerned with the organic fraction of MSW targeted by SB1383, which is made up of primarily food scraps and green waste (landscaping and yard trimmings) and estimated to be approximately 12 million wet tonnes of material annually 1 .We relied on 2025 food and green waste disposal rates as estimated by the 2014 CalRecycle waste characterization assessment composition values.Per SB1383, we assumed that at least 75% of this material needs to be diverted from landfill.
Although not directly targeted by SB1383, California generates organic waste from the state's vast agricultural sector in the form of manure and woody crop residue from orchards and vineyards that are currently underserved as biomass conversion facilities shut down across the state.As jurisdictions pursue additional composting infrastructure investments, it will likely prove beneficial to consider these important feedstocks as well.We relied on data from the Joint BioEnergy Institute California BioSiting website 2 .Manure data were from 2012.Orchard and vineyard residue data were projections for the year 2020.This feedstock is composed of naturally shed leaves, trimmings and prunings removed to improve plant yield, and tree removals.In line with previous studies, we assumed that 20% of total combined manure and orchard residue material will be processed in new composting facilities in order to replace the current volume of manure being applied to land as slurry 3 .
Finished compost, or 'compost,' is the final product resulting from the composting process.This product is distinct from 'feedstock', which is the raw material, such as food scraps, livestock manure or yard trimmings used to produce the compost.We assumed that the new facilities in this study will generate a relatively homogenous product with a high C:N ratio (low Nitrogen) in accordance with the material characteristics of our organic feedstock inputs, which consist of a balance of nitrogen-rich substrates and carbon-rich bulking agents 4,5 .We used a consistent ratio of food waste and green waste from the MSW stream, and of manure and orchard residue from the agricultural residue stream to generate the required balance of nitrogen-rich (food scraps, manure) and carbon-rich (green waste, orchard residue) input materials to generate a balanced final compost.
Data on existing composters were obtained from the CalRecycle Solid Waste Inventory System (SWIS) 6 , which records information facility size, location, status, activity category and accepted feedstock, and from CalRecycle's report "Analysis of the Progress Toward the SB 1383 Organic Waste Reduction Goals" 7 , which provides summary metrics for compost facilities across the state.There are 293 active composting facilities (or "composters") in California (Figure 2 in main text) that currently process approximately 6 million tonnes of organic material annually, though they are permitted to accept up to 10 million tonnes.Composting operations can vary considerably in size and scale, even within the typology we describe in the main text.We use average values of throughput capacity in this model rather than land area, which often depends on local conditions 8 .
We included two types of end-markets where finished compost will likely be land applied to promote soil health and carbon sequestration benefits.The first is rangelands, or grazed grasslands, which were acquired from the California Department of Conservation Farm-land Monitoring and Mapping program 9 .This dataset includes only areas where vegetation and land characteristics are suited to grazing of livestock as determined by a collaboration between the California Cattlemen's Association and UC Cooperative Extension.We also included perennial cropland, derived from California Department of Water Resources 10 .We focused on high-value tree crops (including citrus, subtropical fruit, and nut trees and vineyards) as these are currently the primary agricultural markets for compost 1 .
Different land types across the state are associated with different soil emissions and carbon sequestration rates 11 .An application rate of 4.7 dry tonnes compost/acre was used based on CDFA recommendations 12,13 .In this recommendation, application is assumed to occur once annually before planting or early in the growing season, and results in a projected annual greenhouse gas reduction/sequestration.Net sequestration rates were based on county and land type as given by the California Air Resources Board's (CARB) use of the DeNitrification-DeComposition (DNDC) Model 12 .Land use characteristics from the National Land Cover Database were also included to allocate new composter facility types based on factors such as urban or agricultural density.
CARB Air Quality Districts were used to determine sensitive air districts that have highly restrictive air pollution threshold limits.The CalEnviroScreen dataset was used to determine the median environmental justice score for each zone, as well as whether that zone contained any census block designated as a Disadvantaged Community (DAC), as indicated by a score of 75 or above 14 .CalEnviroScreen considers a comprehensive set of environmental, health and social indicators, including pollution burden, socio-economic vulnerability, and public health concerns, to assess conditions and score census tracts across the state.These values were also used to set constraints in the model.All data-cleaning and processing was done in Python and QGIS.

SI1.2. Siting Model Formulation
Indices, objective functions, decision variables, and constraints of the spatial optimization model are detailed here.

Sets q
Zones (1 x p) T Composter Types: Industrial (IF), On-Farm (OF), Community (CC) i Organic Feedstock Type: MSW, MSW + Agricultural j End-Use Type: Grazed Grassland, Perennial Cropland Dqq' Distance matrix between zone of origin and zone of new composter Bqq' Distance matrix between zone of new composter and zone of land application The decision variables solved in this model are the quantity of annual throughput capacity (in tonnes) of each composter type T to build in each zone q, and how best to allocate organic materials between generating zones and land application areas.Movement of (uncomposted) organic feedstocks of different material classes are modeled separately, as are final land use types receiving finished compost.The decision variables are displayed below.

Decision Variables
Xq,T Annual tonnage of new facility of type T in zone q Fi,qq' Tonnage of (uncomposted) feedstock moving between zones Lqq',i Tonnage of finished compost moving between zones As noted in the main text, we constructed two objective functions to reflect policy orientations that prioritize cost or emissions.The first, minimizing cost, is a function of collection (Ccollection) and hauling cost (Chauling), which itself is dependent on distance traveled, the upfront capital required for new composting facilities (excluding land price and labor) (Cconstruction), and the cost to physically apply finished compost to land (Cspreading), where Aq equals the amount of finished compost applied, as defined below.Fi,qq' is the amount of materials moving between zones into a site of new infrastructure, and Lqq',j is the amount of material moving between zones for land application.Dqq' and Bq'q are the distance matrices between zones.In this function, we did not include costs from labor and land due to limited availability.
Objective Function 1: Minimize Cost The second objective function, minimizing emissions, evaluates total kgCO2eq from a solved arrangement of new composting infrastructure, including roadway collection (Ecollection) and hauling emissions (Ehauling), annual compost processing emissions (Eprocessing), and emissions released during the spreading of finished compost to land (Espreading).We subtract from these emissions the avoided greenhouse gas emissions resulting from diverting organic feedstocks away from landfill (Elandfill) as well as the expected carbon sequestration benefit of applying compost to range or cropland (Esequestration).In these functions, we did not include fossil fuel emissions from machinery during compost processing.
Objective Function 2: Minimize Emissions The above objective functions are each subject to two kinds of constraints.First, a series of physical plausibility constraints were imposed.For instance, total build and build for each type of facility must be positive (equations 14 and 15).Further, material flow from a given zone must be less than or equal to total available organic feedstock in that zone (equation 17).Similarly, the flow into a zone is constrained by the amount of appropriate range and cropland in that zone (equation 18).The build in any given zone is limited to the flow into that zone by material type, and the total flow into new composting facilities (Fi,qq') in each zone must be balanced by the flow out of those facilities onto land (Lq'q,j).(14)  0 ≤ ∑ ' 0,56 +  0,76 +

Figure S1 .
Figure S1.Industrial Facility Capacity for scenarios where only industrial composters were allowed.Panel (a) shows the distribution of new industrial capacity under emission minimization runs.Panel (b) shows the distribution of new capacity under cost minimization runs.The second row displays results where there was no limit on industrial facilities within each zone.As a result, there tends to be higher concentrations of industrial build in the South Coast and Bay regions, surrounding population centers.When a cap is imposed, industrial facilities are distributed across more zones in those regions.

Figure S2 .Figure S3 .
Figure S2.Distribution of new capacity by facility type across scenarios using only MSW feedstock, under cost minimization objective

Figure S5 .
Figure S5.Industrial facility capacity for the Land-use and EJ & AQ scenarios, highlighting the areas where industrial capacity expands or contracts under environmental justice and air quality constraints.Panel (a) shows the distribution of new industrial capacity under emission minimization runs.Panel (b) shows the distribution of new capacity under cost minimization runs.

Table S1 .
Scalar parameter values, descriptions, and sources

Table S2 .
Vector parameters and descriptions

Table S3 .
Sensitivity Analysis Results