FARM Environmental Stewardship

Definition

FARM Environmental Stewardship (ES) is the environmental program area of the National Dairy FARM Program. It estimates a farm’s greenhouse-gas (GHG) footprint and energy use from production, herd, feed, manure and energy data, using a life-cycle-assessment model maintained by the program. The result is reported as total emissions and as emissions intensity: kg of CO₂ equivalent (CO₂e) per kg of fat- and protein-corrected milk (FPCM).

It is a modeled estimate, not a measurement. Its uses are to track one farm’s trend over time, to find the largest emission sources (usually enteric methane, then manure storage, then feed production), and to supply milk buyers with supply-chain data for their own sustainability reporting.

ES is generally a data-collection evaluation: the farm is not scored pass/fail against an emissions threshold. Some milk buyers require it of all suppliers; others only of a sample. Cooperatives and processors aggregate the farm results to report emissions to their customers.

The current model is FARM ES Version 3, released October 31, 2024 and run on the Ruminant Farm Systems (RuFaS) model. FARM revised the Version 3 tool on May 19, 2026, including the manure methodology and report changes for GHG Protocol land-sector reporting (as of November 2026, NMPF). Version 3 adds scenario analysis and optional inputs such as reproduction, culling and field practices. Evaluations are run by FARM-certified ES evaluators.

Unverified

Whether FARM ES is required or voluntary at the program level (FARM’s public pages don’t say; buyers set their own rules), and how often a buyer must re-run it.

Data the model needs

Collected for one fixed 12-month window:

• Production and herd: milk shipped, fat and protein tests, and average animal numbers by class (lactating, dry, heifers by age group, calves, bulls).
• Animal management: housing type by group, time on pasture, deaths and culls.
• Feed: dry matter intake by group and ration ingredients (purchased and home-grown).
• Manure: by group, the housing, bedding, collection, solid separation, storage and treatment (e.g. digester, cover) systems.
• Energy: electricity (kWh), diesel, gasoline, propane and natural gas used on the dairy.

Requirements change; confirm the current version with the program or your milk buyer. This page is reference material, not legal advice.

Formula

Fat- and protein-corrected milk (IDF; standard milk is 4% fat and 3.3% true protein, the same equation in Bulletins 479/2015 and 520/2022):

FPCM (kg) = milk (kg) × (0.1226 × fat % + 0.0776 × true protein % + 0.2534)
Emissions intensity (kg CO2e / kg FPCM) = total farm emissions allocated to milk (kg CO2e) / FPCM (kg)
CO2e (kg) = Σ gas (kg) × GWP of that gas

LaTeX:

\mathrm{FPCM} = M \times (0.1226\,F + 0.0776\,P + 0.2534)
\qquad
I = \frac{E_{\text{milk}}}{\mathrm{FPCM}}
\qquad
E = \sum_{g} m_{g}\,\mathrm{GWP}_{g}

Fat and protein are in percent, and protein is true protein (IDF notes that using crude protein instead changes the result). Milk is in kg (1 lb = 0.4536 kg). Part of the farm’s emissions is allocated to meat (cull cows, bull calves), so the milk share is less than the farm total. IDF 520/2022 recommends biophysical allocation between milk and meat and IPCC AR6 GWP100 values (27.0 for biogenic methane).

Unverified

Which GWP set (AR5 or AR6) and which milk/meat allocation method the FARM ES Version 3 (RuFaS) model uses, and whether its FPCM input is true or crude protein. Not stated on FARM’s public pages.

How to pull it in DC305

DC305 supplies the herd section. The rest comes from outside it:

Data	Source
Animal numbers by class	DC305 counts, averaged over the window
Deaths and culls	DC305 DIED / SOLD events
Milk shipped, fat, protein	Milk checks or processor statements, not DC305 test-day data
Feed and DMI by group	Feed management software, ration sheets
Manure systems	Facility description, nutrient management plan
Energy	Utility bills and fuel invoices

COUNT FOR LACT>0 RC<>6          lactating cows (RC 6 = dry; <> = not equal)
COUNT FOR LACT>0 RC=6           dry cows
COUNT FOR LACT=0                heifers and calves
COUNT FOR LACT=0 AGE<12         heifers under 12 months (AGE is in months)
COUNT FOR LACT=0 AGE>=12        heifers 12 months and older
EVENTS\D ID LACT FOR EDAY>-365  DIED / SOLD events in the window (animals that left need \D or \B)
ECON\ID                         past / yearly cow inventory summary
ECON\IHD                        past heifer inventory

ECON\ID and ECON\IHD give past inventories, which suit averages over the window; check that their periods match the model’s.

Average counts over the window: take a count at each month-end (or use daily inventory) and average them, rather than a single count on the day the data is entered. See cow count definitions. Event names for deaths and sales vary by herd; check the herd’s private dc305-vocab mapping.

Typical range

Herd context: US Holsteins, freestall or dry-lot; requirements as of the current FARM ES version.

Measure	Typical US value
Emissions intensity, farm gate	about 0.9–1.3 kg CO₂e / kg FPCM
Largest single source	Enteric methane
Data window	One 12-month period, the same months every year

Thoma et al. (2013) put the US national production-weighted average at 1.23 kg CO₂e per kg FPCM at the farm gate, circa 2008, and found feed-to-milk efficiency and manure system the biggest drivers.

Unverified

The lower end of the intensity range (current US averages are thought to be below the 2008 figure), the enteric methane share of US farm-gate emissions (often quoted as about 35–50%), and current FARM ES average intensities. None was found in a peer-reviewed source that could be read in full.

• Intensity falls with higher milk per cow, better feed efficiency and fewer non-productive animals, because those spread fixed emissions over more milk.
• Manure systems shift the result a lot: liquid storage emits more methane than solid or dry-lot systems, and digesters or covers reduce it.

Common mistakes

• Changing the 12-month window. Use the same months each year, e.g. January–December. A shifted window mixes seasons and makes the trend meaningless.
• Single-day animal counts. Entering the count on the day of data entry instead of the window average, especially in a growing or shrinking herd.
• Mixing herds or sites. Heifers raised off-site or a second milking site entered with the main dairy, or left out with no note.
• Test-day milk instead of shipped milk. Use milk sold over the window from processor statements.
• Mixing units. Entering pounds where kg are expected, or gallons of propane as cubic feet of natural gas.
• Reading the number as a measurement. Year-to-year changes smaller than the model’s uncertainty may only reflect data-entry differences.
• Defaults left in place. Leaving default rations or manure systems when the farm’s real data is available.

Related

• National Dairy FARM Program
• Manure and nutrient management
• Energy-corrected milk
• Dry matter intake
• Cow count definitions