Why and when?
For life cycle inventory analysis (LCI) it is common to distinguish between consequential and attributional modelling. The examples on this website are exclusively about consequential LCI modelling. But in this section we explain what the difference to attributional modelling is and why and when consequential modelling is the right choice for your LCA.
A modelling choice
The two Life cycle assessment approaches are defined as follows in the UNEP/SETAC Shonan guidance on LCA (UNEP 2011), italics added:
- Attributional approach: System modelling approach in which inputs and outputs are attributed to the functional unit of a product system by linking and/or partitioning the unit processes of the system according to a normative rule.
- Consequential approach: System modelling approach in which activities in a product system are linked so that activities are included in the product system to the extent that they are expected to change as a consequence of a change in demand for the functional unit.
As it is clear from the definitions the two methods vary greatly. The conceptual difference is illustrated in the figure below.
The two models answer different questions
Another way of describing the difference is in terms of the questions that the two models can answer:
- An attributional product system can be used to answer the question: “Under the specified normative allocation rule, what are (the environmental impacts related to) the allocated shares of the activities that have contributed to the production, consumption, and disposal of the product?” Thus, the purpose of attributional modelling is to trace a specific aspect of the product (as determined by the allocation rule) back to its contributing unit processes. In such a system it is relevant to use data on specific or market average suppliers, and to partition them according to the chosen allocation rule.
- A consequential product system can be used to answer the question: “What are (the environmental impacts related to) the full share of those activities that are expected to change when producing, consuming, and disposing of the product?” Thus, the purpose of consequential modelling is decision support. This implies that in such a system, the consequences are traced forward in time, which means that it is relevant to use data on marginal suppliers and substitution of displaced activities.
Most LCA studies are actually aimed at decision support involving a choice or substitution between two product systems. Even studies of a single product (hot-spot-identification) are typically later used in a comparative context, since the ultimate goal often is to improve the studied systems, thus supporting decisions that involve comparisons, even if indirect. Even when continuing business-as-usual, the consequences of this can be compared to not producing, consuming, and disposing of the product, thus giving the life cycle impact of this product. The same reasoning holds true for Environmental Product Declarations (EPDs) that are used by the customers to make choices between several products. The effect of such choices are that more of the chosen products will be produced at the expense of the competing products – and a consequential LCA is the relevant method to reflect the consequences of chosing one or the other product, which is what the customer needs to make a sound decision (Weidema 2003).
Fully reflecting physical and monetary causalities
One of the most important differences between consequential and attributional models is that only consequential models reflect physical and monetary causalities. Attributional models use normative cut-off rules and allocation to isolate the investigated product system from the rest of the World. And by this, they violate both basic laws of mass and energy conservation and basic principles of economic accounting (the accounting equation).
The below two slide-shows explain five physical and monetary causalities that are included in consequential life cycle models, while not being included in attributional models. The first slide-show covers the most commonly encountered situations, while the second is about situations that are more seldomly encountered (double click for full-screen mode).
Consequential models are neither scenario models nor equilibrium models
Although consequential models seek to identify the consequences of changes, they are steady-state, linear, homogeneous models, with each unit process fixed at a specific point in time. A consequential model does not include any dynamic feedbacks, and it does not necessarily include activities far into the future, if the product system and the consequent changes take place here and now. Consequential modelling is not scenario modelling, but models how activities influence each other and their environment. When modelling a product system that take place here and now, consequential modelling may use data from the recent past: Data on the recent marginal changes and the recent substitutions – thus assuming that the current consequences of producing and consuming a product are similar to those of producing and consuming the same product in the recent past. When the consequences to be modelled happen further into the future, forecasting may of course be used to give a better reflection of the expected future situation. Thus, consequential modelling can be used both as part of a future scenario model and for studying activities that take place here and now.
Although consequential models use market information to identify which activities are affected by a change, a consequential model is not an economic equilibrium model. Equilibrium models focus on the short-term effects, where there are many constraints on changing the supply, and part of the changes are therefore assumed to be changes in demand (consumption), as determined by price elasticities (often assumed or highly speculative). In contrast, consequential modelling focus on the long-term effect of the changes, which in most cases imply that the short-term constraints in supply have been overcome, the production capacity has been adjusted, and the prices have returned to the previous level (as determined by the marginal production costs), which means that the change in demand is full reflected in a similar change in supply, and therefore does not lead to reductions in consumption. Compared to the complex equilibrium modelling, consequential modelling is a simpler, rule-based modelling approach. The modelling may be either marginal (for small changes) or incremental (for larger changes).
How to model consequentially?
The key issue in consequential LCI modelling is the identification of the unit processes that change as a consequence of a decision. This key issue then has implicit consequences for several central elements of the LCA technique:
- How the functional unit and reference flows should be defined.
- How unit processes are linked into product systems via intermediate product flows as identified via the expected reactions of suppliers and users.
- How to deal with unit processes (or product systems) with multiple products.
All of these steps are discussed and exemplified on this website.
Outcome of an LCA
When presenting the outcome of an LCA, it is important to communicate that the result, as e.g. presented in the inventory tables, does not represent the environmental impacts of the functional unit in itself, but the environmental exchanges resulting from adding or subtracting one functional unit compared to doing nothing. When comparing two products, the result is the difference in environmental impact caused by fulfilling the functional unit with one or the other product (obtained by subtracting the result for one of the products from the results for the other product).
CLCA Current best practice for LCA Slideshow
View this video on youtube. The video addresses solutions in consequential LCA with a contrast to attributional LCA options, and generally explains the way CLCA is modelled.