Average EPD

An average EPD is one which declares average results covering more than one product and/or manufacturing site.

In order for different products to be declared in the same average EPD, they need to:

• Perform the same or similar function
• Have the same or similar manufacturing processes (technology and raw materials)
• Have similar emissions profiles*

* Different Program Operators have their own rules about this, as described later in this article.

To know whether products have similar emissions profiles, an LCA needs to be done for each one. These individual LCAs are then combined into one average LCA for all products.

The average LCA should be a weighted average, using annual production volume to ‘weight’ each product based on how much of it was produced in proportion to the other products covered by the average LCA. For example, if Products A, B, and C accounted for 50%, 30%, and 20% of overall production volume in the studied period, Product A would contribute half of the inputs in the average LCA.

To combine multiple LCAs into a single average LCA in One Click LCA, there are three options; 1) Copy individual designs to one average design and use the Allocation % feature to ‘weight’ each LCA’s inputs, 2) Calculate average quantities and apply them to a single average design, or 3) Create an average design feature for Expert users. 

1) Copy individual designs to one average design and use the Allocation % feature to ‘weight’ each LCA’s inputs

PRODUCT NAME	CEMENT (kg)	LIMESTONE (kg)	PRODUCTION VOLUME %
Product A	475	25	50
Product B	450	50	30
Product C	400	100	20

In this basic example, each product has just two input raw materials - cement and limestone. The recipe for each product has different shares of the ingredients, e.g. Product A is 95% cement and 5% limestone, while Product B is 90/10, and Product C is 80/20.

They each have a different share of production volume, with Product A accounting for half of overall volume, Product B 30%, and Product C 20%.

In the software, the inputs for each product are copied to the average LCA and then allocated based on production volume:

This basic example is quite easy to follow. However, in more complex models where there are large numbers of inputs, it may be easier to take option 2.

2) Calculate average quantities and apply them to a single average design

NAME	CEMENT (kg)	LIMESTONE (kg)	PROD. VOLUME %	CEMENT (weighted)	LIMESTONE (weighted)
Product A	475	25	50	237,5	12,5
Product B	450	50	30	135	15
Product C	400	100	20	80	20
Average	-	-	-	452,5	47,5

Average quantities can be calculated in excel for example, by multiplying the percentage of production volume by the input quantities for each product and then adding them together.

Then in the software you have a simpler model which just single inputs for each raw material:

While this requires some calculations outside of the software, it gives a more concise model which makes it easier to understand and replicate mass balance.

3) Create an average design feature for Expert users

Expert users can create a design type called average design:

This allows combining the inputs of two or more designs according to percentages you assign:

A new design will then be created with weighted average results for A1-A3. This makes it faster to create EPDs that require inputs from multiple products modelled in One Click LCA.

Please note that other modules, if included, will be taken from the design the "Create an average EPD" button was clicked from.

Program Operator rules for Average EPDs

A summary of averaging rules for different EPD programs is given below. For more detailed information, please see Rules For Product Averaging and Similarity For Different EPD Programs.

EPD Hub: The allowed variability in A1-A3 GWP-fossil results for covered products and/or sites is ±50%. More information on averaging rules can be found in the EPD Hub General Program Instructions in Annex I.

The International EPD System: There are three options; 1) average results, 2) representative product, 3) worst-case product. For options 1 and 2, aggregated results for modules A-C cannot differ by more than 10%. Variations >10% are possible for option 3.

RTS: The variance in GWP emissions for different products cannot be more than 10% from the highest to the lowest.