Here are the definitions for each scenario:

• Near-term operations

  • Refers to less optimized Mill operations projected for the beginning of 2025 through the end of 2026 where resource efficiency, energy efficiency, transport distances, and manufacturing have room for improvement with time and product/process refinements.
    
    

• Conservative assumptions

  • Selected to emulate a more pessimistic scenario, where processes are, for example, assumed to be inefficient, energy-intensive, geographically further apart, and products are shorter-lived and heavier. Often, current operational efficiencies were used for conservative assumptions. Because Mill has a continuous improvement culture, most of the current operational efficiencies will only improve over time, and can be considered worst-case scenario values.
    
    

• Optimistic assumptions

  • Selected to emulate a more optimistic scenario, values were selected that were considered by the engineering and operations teams to be realistically achievable within the near-term timeline (over the next two years).

[1] For activities that do not emit methane, the same EF was used for both the 20- and 100-year GWP scenarios, as most other non-CO2 greenhouse gases are long-lived in the atmosphere, and the relative amounts of heat trapped do not significantly vary between 20- and 100-year time horizons.

*The base case scenario assumes scraps were managed in accordance with the U.S. average residential food scraps management mix (according to 2019 EPA data), and the Mill customer average Food Grounds management mix (based on January 2025 customer survey data). See the Impact Matrix appendix below for impact estimates of additional scenarios. The values shown in the table account for both improved scraps management (476.7 for the mean) and source reduction from engagement-driven behavior change (258.2 for the mean).

Below is a waterfall chart showing the breakdown of the impact of the base case scenario:

This also assumes that the 20% reduction in mass additions to the bin was not attributable to an increase in placing those food scraps in the garbage. This assumption is supported by an increase in Mill customer Net Promoter Score, a metric used to quantify customer satisfaction with a product or service, over the same period. Further, engagement, defined by the number of lid-opens per week, remains consistent over the same period, suggesting that this decrease in mass additions is not attributable to less use of the bin.

A 20% reduction in food scraps generated was estimated to equate to a 4.4% reduction in food purchases - which may not be detectable to the average household given week-to-week fluctuations in purchases. Even with this supporting evidence, the extent to which source reduction typically occurs has not been validated.

Below is a waterfall chart showing the breakdown of impact from near-term scenarios with source reduction included:

We organized the results from Mill Food Recycler operations into the following groups:

• Device: manufacture, package, ship, refurbishment, and End-of-Life (EOL) management
  
  

  • Scope: Includes Mill Food Recycler manufacturing (extraction and processing of raw materials, component manufacturing, and assembly), packaging, energy to transport the bin to the customer’s home, reverse (and subsequent forward) logistics for bin refurbishment, manufacture and assembly of components replaced during refurbishment, and management of end-of-life components. 
    
    

  • Result: The majority of the impact of this process group can be attributed to the annualized impact of manufacturing the bin. Refurbishment activities are the secondary source of impact.
    
    

• Device use: electricity usage & charcoal filter replacement
  
  

  • Scope: Includes electricity consumption during normal bin use as well as the production and shipping of replacement charcoal odor filter cartridges.
    
    

  • Results: Electricity consumption is the primary source of emissions for this process group. It is also the primary source of emissions for the Mill bin overall, although still small relative to the counterfactual.
    
    

• Food Grounds to feed
  
  

  • Scope: Includes the emissions invested to make Food Grounds mailer boxes and liners, to ship multipacks of mailers to member homes, and to ship Food Grounds-filled mailers back to Mill’s feed ingredient manufacturing facility. It also includes the energy used by the manufacturing facility, and to ship the feed ingredient to a
    formulator.
    
    

  • Result: Although the impact of shipping is non-zero, it is small relative to the bin’s electricity use. By leveraging trucks that are already on the road, shipping transport emissions are measured as nominal ton-miles – the energy to carry the additional mass of Food Grounds from the home to the feed ingredient manufacturing facility. Shipping Food Grounds is more impactful (has a greater GHG impact) than the feed ingredient manufacturing process.
    
    

• Food Grounds to compost
  
  

  • Scope: Includes the emissions resulting from backyard composting as well as commercial composting of Food Grounds (and the transport to the facility). Compost process fugitive emissions (non-CO2 biogenic emissions) are also accounted for.
    
    

  • Result: Composting (both commercial and backyard) is typically an approximately net-neutral process (the small quantity of fugitive emissions are typically offset by the small portion of compost carbon which becomes sequestered in the soil which the compost amends), so it has a negligible impact on the LCA overall.
    
    

We organized the results from the counterfactual in the following groups:

• Food scraps kept out of landfill:
  
  

  • Scope: This process estimates the impact of landfilling food scraps before Mill. This includes the collection and transportation of scraps to the landfill, landfilling biogenic methane release, and emissions from landfill equipment. The portion of food scraps that were landfilled before Mill was estimated based on EPA residential food waste management data FY2019, reflecting the behavior of the average U.S. household.
    
    

  • Result: This is typically the greatest source of impact in the whole system. Because of its potency, the modest mass of methane emitted from landfilling one household’s food scraps (roughly 10 - 20 kg of methane) results in hundreds of kilograms of CO2e. The portion of food scraps landfilled were sourced from an April 2023 EPA report (2019 Wasted Food Report). The landfill emissions calculations were based on an October 2023 EPA report (Quantifying Methane Emissions from Landfilled Food Waste). The model accounts for landfill gas capture, methane oxidation (to CO2) in the near-surface of the landfill, landfill gas electricity generation, and the non-captured methane released to the atmosphere. See the Landfill Methane section below for more information.
    
    

• Avoidance of other food scraps management methods:
  
  

  • Scope: This process estimates the impact of all raw food scraps that weren’t landfilled before getting Mill. It calculates this impact by referencing EPA’s 2019 Wasted Food Report for the relative portions of U.S. residential food scraps managed by different methods. It includes food scraps sent to waste-to-energy waste incineration facilities, food scraps backyard and commercially composted, feeding raw scraps to backyard chickens, putting food scraps down the sink or sink garbage disposal, and scraps anaerobically digested. The impact of transportation, fugitive emissions, fossil process emissions, electricity generation credits, and any carbon sequestration that results are all accounted for in these pathways.
    
    

  • Result: Of all of the food scrap management methods described in this process, sending food down the sink or sink garbage disposal was estimated to be the most impactful per pound of food (although less impactful than landfilling food). Because much more food is typically landfilled than is sent down the drain (and because the non-sink management methods for managing food scraps have relatively lesser impact), this process was not nearly as impactful as landfilling food scraps.
    
    

• Conventional feed production avoided:
  
  

  • Scope: For Food Grounds turned into a commercial feed ingredient, or those fed directly to chickens, some conventional feed is displaced. This process estimates emissions avoidance from that displacement. It includes conventional grain agriculture, grain transport, grain processing into a feed ingredient, and transport to a customer.
    
    

  • Result: Most of the impact results from grain agriculture. Transport and processing are relatively efficient on a mass basis (per pound of grain, for example).
    
    

Below is a chart of the results for each scenario using GWP20 and GWP100:

Local grid energy mix

• The proportion of a Mill member’s local electric grid mix that is fossil-derived energy affects the emissions of using the Mill Food Recycler. Although the grid is becoming progressively cleaner over time, this is a slow process. Households do have the power to significantly reduce the footprint of bin electricity use, and of their home overall by purchasing renewable electricity. For conservatism, enrollment in renewable energy was not considered in the 2025 model (i.e. the average grid mix was used).
  
  

• For both the conservative and optimistic scenarios, we used the 2022 state-based EPA eGRID emission factors. A single nationwide grid mix emission factor was calculated by taking the weighted average of state-based emission factors and the relative proportion of Mill customers in each of the continental U.S. states. Although the grid energy mix will likely become cleaner than it was in October, 2024 by the end of 2026, the same value is used for both scenarios for conservatism.

Landfill methane

• Because landfilling food scraps is so impactful, care was taken to fairly develop the underlying calculations. The food waste landfilling calculations are highly sensitive to numerous assumptions. The table below includes some of the key assumptions. Above the assumptions, the calculations carried out in the model are shown. They include the impact not only of fugitive emissions but also of methane oxidation within the near-surface of the landfill, landfill gas capture and flaring, and landfill gas-fired electricity generation.
  
  

• Two of the key studies referenced in the calculations were:
  
  

  • EPA WARM 2023 v16
    
    

  • Quantifying Methane Emissions from Landfilled Food Waste - EPA, 2023-10
    
    

• The following calculations are applied to estimate carbon avoidance from keeping food out of landfill:
  
  

  •  (kg of food scraps generated by household per year) * (% food scraps were landfilled in the counterfactual scenario) = (kg food scraps landfilled per year)
    
    

  • (kg food scraps landfilled) * (1 - % Moisture Content of food scraps) = (kg of food scraps dry matter)
    
    

  • (kg of food scraps dry matter) (Weight % of food scrap dry matter is carbon) (Weight % of carbon is converted to methane) = (kg of methane generated)
    
    

  • (kg of methane generated) (% of landfills with gas capture) (% of methane captured by landfills with gas capture) = (kg of methane captured)
    
    

  • (kg of methane generated) * (% of methane generated that is oxidized before reaching landfill surface) = (kg of methane oxidized)
    
    

  • (kg of methane generated) - (kg of methane captured) - (kg of methane oxidized) = (kg of methane released)
    
    

  • (kg of methane released) * (kg CO2e / lb methane released) = (kg CO2e from landfilled food scraps methane release)
    
    

  • (kg of food scraps landfilled) (1 short ton / 907.2 kg) (% of landfills have gas capture systems with electricity generation) (kWh / ton food scraps landfilled) (grid electricity mix carbon avoidance / kWh generated) = (kg CO2e credit from electricity generation)
    
    

  • (kg of food scraps dry matter) (Weight % of food scrap dry matter is carbon) (Weight % of carbon remains sequestered in the landfill indefinitely) * (kg CO2e / kg C) = (kg CO2e credit from food scraps carbon that remains sequestered in the landfill)
    
    

  • (kg CO2e from landfilled food scraps methane release) - (kg CO2e credit from electricity generation)
    
    

  • (kg CO2e credit from food scraps carbon that remains sequestered in the landfill) = (Net kg CO2e from landfilling food scraps)

*The portions of methane captured, oxidized, and released are calculated in alignment with EPA’s October 2023 report, Quantifying Methane Emissions from Landfilled Food Waste, by manually integrating the quantities of methane generated, oxidized, captured, and released year-over-year over a 30-year period. The respective summed values of oxidation, capture, and release were divided by the summed value of landfill gas generated.

Food scraps generation rate

• The amount of food scraps generated per person and the number of people in a household affect this rate. If a Mill customer lives alone or travels a lot, they could have a lower food scrap generation rate per device per day. The 2025 model assumes 1.1 lbs of food scraps / household / day, before behavior change, based on a significant body of field data. After three months of bin use, this reduces to 0.87 lbs / household / day (reflecting the post-behavior change generation rate).

Counterfactual fate of food scraps

• For the base case scenario, the 2025 model assumes food scraps were managed in alignment with the U.S. average food scraps management mix, according to 2019 EPA data. The specific scraps management mix is: 
  

  • 66.2% of scraps are landfilled

  • 15.1% of scraps are incinerated (sent to waste-to-energy facility)

  • 3.7% of scraps are commercially composted

  • 15% of scraps are sent down the drain (via the sink or sink garbage disposal)
    
    

• The 2025 model is structured to enable the assessment of alternative counterfactual scenarios, to tailor the math to specific use cases (such as 100% of food scraps being landfilled). See section below for impact estimates of alternative scenarios.
  
  

Source reduction

• To estimate the impact of source reduction, a calculation is needed to translate the observed reduction in food scraps added to the bin (assumed to be a reduction in food scraps generated) to a reduction in the mass of food purchased. The math suggests a 21% reduction in scraps generation corresponds to a 4.4% reduction in mass of food purchased, which corresponds to a carbon avoidance of 258 kg-CO2e / household / year. 
  
  

• This works by assuming that the amount of inedible food scraps generated remains the same during behavior change. So, for example, if a person generated 5 banana peels a week before getting Mill, they will continue to generate 5 banana peels a week after having Mill for a few months. Thus, the lesser mass of scraps going into the bin is edible food parts that are now eaten rather than being tossed, and this is edible food that no longer needs to be purchased at the store.

Sensitivity Analysis

A sensitivity analysis was conducted to understand the relative sensitivity of estimated annual emissions avoidance to key model inputs. 

The analysis was carried out by initializing the model to represent the base case scenario. The base case scenario attempts to reflect the typical American household in terms of their diet, food scrap generation behavior, food scrap management behavior, and the typical behavior change with Mill. With the base case setup, a series of input values were varied, one at a time. The output (estimated annual carbon avoidance) value was recorded after each variation.

For comparability of input value sensitivity, inputs were varied by a standardized value of + / - 25% (percent, not percentage points). To establish the “low” values, the input value was multiplied by 0.75, to establish the “high” values, the input value was multiplied by 1.25. The only case where this deviated from was the raw food scrap wet weight moisture content, where the base case value of 72.2% was multiplied by 0.875 to get the low value and 1.125 to get the high value. This was done because dry matter content changes significantly with a very small change in wet-weight moisture content, particularly when that moisture content is high. For example, a change from 75% to 87.5% moisture content would halve the dry matter content.

As is clear from the analysis, the LCA model is most sensitive to the baseline daily weight of food scraps generated. This is logical because improved scraps management is directly proportional to the amount of scraps that were wasted before, which are being beneficially used with Mill. In other words, Mill reduces the impact of food scraps management, and thus Mill can only help reduce this impact by (it is constrained to reducing it by) as much as the impact of managing the quantity of scraps generated before Mill. If a large household generated a lot of food scraps before, their behavior before Mill resulted in more emissions than a smaller household which generated less food scraps.

Beyond the baseline (counterfactual) food scraps generation rate, the 2025 model is most sensitive to:

• The extent of behavior change that takes place (specifically the amount by which scraps generation decreases following the adoption of Mill)
  
  

• The moisture content of food scraps (which is typically outside of the control of the household)
  
  

• The portion of food scraps that were landfilled before Mill

Mill has attempted to account for these sensitivities by conservatively structuring the base case scenario. Daily mass additions of food scraps to the bin (and the percent reduction of mass added over time) were derived from more than a year’s worth of daily mass additions from thousands of bins. The raw food scraps moisture content value references EPA WARM v16 (72.2%) - which is lower than other studies suggest (80%), a conservatism for carbon avoidance. The portion of food scraps landfilled before Mill figure references the latest EPA data.

Impact Matrix

Below is a matrix showing a series of counterfactual scraps management x Mill Food Grounds management scenarios. There are two sets of values included:

1. The estimated carbon impact resulting from improved scraps management with use of the Mill bin
   
   

2. The estimated carbon impact from both improved scraps management and source reduction*.

*The source reduction calculation estimates that a 21% reduction in mass additions to the bin over the first three months of bin use corresponds to a 4.4% reduction in purchases of food for home use.

Negative values (those in parentheses) are avoided emissions, positive values are additional emissions. 

For the scenarios where 100% of food scraps were commercially- or backyard-composted, there are additional emissions when adopting the Mill bin, when source reduction is not included. This is because no landfill avoidance can be accounted for.