Abstract
This paper uses an industrial case study of a boron system producing five co-products to examine different allocation methods recommended by ISO 14041 and compare them with the allocation methods most commonly used by LCA practitioners. In particular, allocation by physical causality is discussed. The paper illustrates how the use of whole system modelling can help to identify the correct type of causality for allocation. The case examined here concerns marginal changes of product-related parameters in the system, in this case represented by the output of boron co-products. The analysis shows that in some cases it can be correct to allocate the burdens on the basis of a simple physical quantity, such as mass, as long as the allocation parameter is based on physical causation and is not chosen arbitrarily. In whole system modelling, the correct causality is identified by the model itself, so that the possibility of allocation by an arbitrary parameter is avoided. However, as for system disaggregation and expansion, allocation through mathematical modelling may only be possible if detailed data for the system are available.