Material Flow Analysis (MFA)

The material flow analysis (MFA) method is described in the “Practical Handbook of Material Flow Analysis” by BRUNNER & RECHBERGER (2004). Material flow analysis contains the following main steps:

• Identification of the key (material flow related) issues.
• System analysis (selection of the relevant matter, processes, indicator substances (elements), and system boundaries).
• Quantification of mass flows of matter and indicator substances.
• Identification of weak points in the system.
• Development and evaluation of scenarios and schematic representation, interpretation of the results.

“Indicator substances”considered for MFA are generally chemical elements such as nitrogen (N) and phosphorus (P), carbon (C) or CO₂ and NH₃. These substances can be either pollutants (e.g. eutrophication) or resources (e.g. fertiliser in agriculture).

“Goods”represent a substance or a mixture of substances with a function valued by man such as food, solid waste and wastewater.

“Processes” describe the transformation, transport or storage of “goods” and “substances” such as households, wastewater treatment plants or agricultural soils.

The first and second step of MFA consists in the identification of the key material flows and the setting of system boundaries (selection of the relevant matter, processes, indicator substances). For the indicator substance nitrogen, for example, the relevant “goods” are the inputs and outputs to and from the household, which contain nitrogen: that implies food, excreta, blackwater, greywater and organic solid waste. In order to have an overview of nitrogen pathways at household level, the processes are the household (food and water consumption, digestion, and kitchen wastes) and the on-site sanitation system. Nitrogen containing “goods” are being transformed in these two processes. For example, nitrogen enters the human organism as food (household) and leaves it, after digestion, with the excreta (on-site sanitation system). The system boundaries includes all household activities and the on-site sanitation system.

The third step consists in quantifying “goods” and substance flows in the system. Only the substance-flows (nitrogen-flows in this example) are of interest in this case. “Substance” flows are often calculated based on the mass flows of “goods” and substance concentrations in these “goods”. For example, the nitrogen flow in excreta can be assessed by multiplying the quantity of excreta produced per capita per day with the nitrogen concentration in excreta. Source data can be obtained from literature, estimations, measurements or calculations.

In the fourth and fifth step, the assessed “good” and/or substance flows are represented graphically and interpreted (see picture below). The software STAN (Software for Substance Flow Analysis), which is open source can be used for graphic representation. STAN was developed by the Vienna University of Technology and can be downloaded on the internet for free. It can be used as a base for modelling material flows for the assessment of the economic resources and environmental values of materials (CENCIC & RECHBERGER 2008). The Vienna University of Technology also offers workshops on the use of MFA and STAN; you can subscribe to their newsletter to stay informed. 

Because of the amount of data needed, the implementation of MFA is difficult in the contexts of developing countries. MONTANGERO (2006) suggested therefore using the method of “eliciting expert judgement” to deal with this problem. Eliciting expert judgement consists in the translation of expert knowledge (subjective judgements) into probabilities for real values (data). Eliciting expert judgement comprises seven steps, the first step consists in the identification of the parameters for which expert assessment is needed; the second step consists in the recruitment of experts; followed by motivating experts; the development of a general model that reflects the experts’ thinking; the training of the experts in probability assessment; the required probability assessments and a documentation of the reasoning behind the assessments; and finally the aggregation of experts’ probability distributions (i.e. if experts’ views on the same parameter differ, it is important to try to understand why they reach different conclusions). Additional knowledge subsequently gained through measurements for example can be combined with expert knowledge.