Dairy products are among the most GHG-intensive foods in the global food system, driven largely by on-farm emissions. As a result, carbon emissions from dairy have become a central focus for food companies, retailers, producers to reduce Scope 3 emissions. Retailers are asking for PCF submissions by their suppliers to have a granular view over their Scope 3.1.
Yet anyone who has tried to compare the carbon footprint of milk, cheese, or butter across different PCF calculations may encounter a major challenge: the numbers can vary significantly, sometimes by more than a factor of two. Shouldn’t these figures be standardized? This raises an important question: what are the key parameters needed to ensure a consistent and standardized approach to validating and comparing dairy product PCFs?
A common misconception is that a dairy carbon footprint represents a single physical truth; in reality, it reflects a set of modeling choices—including system boundaries and allocation methods—applied to the same underlying production system.
The long and short of it is that dairy production is inherently complex. Milk, both as a raw ingredient or a finished dairy product, comes from multi-output systems that generate several co-products. As a result, the way emissions are allocated among the different co-products plays a decisive role in determining the final carbon footprint.
Understanding why these differences exist and what they mean is essential for the validation and comparability of the carbon footprint of dairy products. The end goal is always to be able to make credible and actionable decisions.
The carbon footprint of dairy products measures the total greenhouse gas (GHG) emissions during its life cycle. According to the IPCC, greenhouse gas emissions are the release of gases such as carbon dioxide (CO₂), methane (CH₄), nitrous oxide (N₂O), and fluorinated gases. The result is expressed as kg CO₂e per kilogram or per liter of the dairy product.
In a cradle-to-gate assessment, the main sources of emissions include direct emissions from livestock, such as enteric fermentation and manure management. But that’s only part of the picture; cradle-to-gate PCFs also capture upstream and on-site emissions needed to produce and process milk into a finished dairy product, including:
.jpg)
Dairy farms operate as multi-output systems, where the same herd and farm activities produce both milk and meat (from culled cows and surplus calves). The method used to allocate emissions between these co-products can substantially affect the calculated carbon footprint.
Let’s look into it using an example:
South German dairy farms (same system, different allocation):
A systematic review also explicitly concludes that allocation choice can strongly influence the final milk footprint, which is why cross-study comparisons often get messy if methods aren’t aligned.
In practice, most dairy-specific guidance recommends using biophysical allocation between milk and meat at the farm gate, as it reflects underlying biological and energy relationships rather than volatile market prices. This approach is notably supported in IDF (International Dairy Federation)-aligned methodologies.
This preference is also consistent with broader environmental footprinting principles. Under the EU Product Environmental Footprint (PEF) rules, the general approach is to avoid allocation wherever possible. When allocation cannot be avoided, a relevant physical relationship should be applied first, and only as a last resort should other methods—such as economic allocation—be used.
If you want comparability, always state:
(1) which allocation method you used (biophysical vs economic vs mass/protein vs system expansion) and
(2) the % allocated to milk—because two “milk footprints” can differ substantially even if farms are identical.
Even after allocation, milk footprints are still sensitive to milk composition (fat and protein vary by breed, feed, season, and region). That’s why many studies report results per fat- and protein-corrected milk (FPCM): it standardizes milk to a common fat and protein content, allowing systems to be compared on a like-for-like basis. The underlying farm emissions do not change; they are simply expressed per unit of standardized “equivalent” milk.
The fact that dairy is a multi-output system means there’s variation in the carbon footprint of dairy. It is unlike single-output processes where emissions can be directly attributed to one product. When it comes to dairy production, there can be several outputs from the same underlying activity.
For example, raw milk is rarely sold as-is. It is typically separated into cream and skim milk, which are then processed into products such as butter, cheese, yogurt, and milk powders. These products may then move further downstream into multiple value chains.
Let’s look at cheese production as another example. In addition to cheese itself, large volumes of co-products—such as whey, lactose, and whey proteins—are generated. The way emissions are allocated among these outputs can significantly influence the calculated carbon footprint of each product.
There is no physical way to indicate that all GHG emissions belong to just one product because all of these products are coming from the same milk and share the same upstream emissions. Assigning all emissions to cheese, butter, or milk alone would misrepresent reality and distort comparisons. For this reason, allocation is unavoidable in dairy carbon footprint calculations.
Allocation allows a systematic way to distribute emissions across multiple products based on a defined criteria such as:
It’s important to know that allocation is the backbone of how the carbon footprint of dairy products are constructed. At the same time, it is also the main reason why reported values can differ so widely across sources.
There are several allocation approaches that are commonly used in the life cycle assessments of dairy products.
There is also system expansion, which avoids allocation altogether by crediting co-products for the emissions they displace elsewhere.
Due to these differing approaches to allocation, there can be a wide variation of published results. For cheese, the carbon footprint of the same cheese produced in the same facility can range from roughly 6 to 13 kg CO₂e per kilogram, purely depending on which allocation method is applied. Milk shows a similar pattern: depending on whether economic, biophysical, or system expansion approaches are used, reported footprints can range from approximately 0.68 to 1.53 kg CO₂e per kilogram of fat- and protein-corrected milk (FPCM). These differences do not indicate better or worse farming practices. They are the result of accounting logic layered onto the same physical system.
Processing further amplifies these inconsistencies. Dairy plants generate multiple co-products, and treating outputs like whey or lactose as valuable products in one study but as low-value by-products or waste in another can significantly shift how emissions are distributed. When allocation rules and co-product treatment vary, comparing dairy carbon footprint values across studies or databases becomes inherently unreliable, even when the underlying production systems are very similar.
In dairy PCFs, the biggest source of “disagreement” is rarely the farm itself—it’s the accounting. Because dairy is a multi-output system from barn to plant, allocation choices and functional units (like reporting in FPCM) largely determine where emissions end up on the balance sheet. That’s why comparability starts with transparency: clearly define system boundaries, use a consistent allocation approach (ideally biophysical at the farm gate), standardize the unit of comparison, and disclose the milk allocation percentage and co-product treatment.
If you want PCFs you can actually trust and compare across suppliers, Carbon Maps helps you get there. Our solution is PACT-conformant, enabling comparable, verified product-level carbon data that retailers and brands can use confidently for Scope 3 reporting and decarbonization decisions. Get in touch to see how Carbon Maps can help you standardize dairy PCFs across your value chain.
See how France's first dairy cooperative, Sodiaal, scaled the product carbon footprint of their dairy portfolio with Carbon Maps.
.svg){kind=icon}
