LCA hotspot analysis: finding the biggest impact by contribution

Hotspot analysis is the contribution analysis that quantifies which processes, which material and energy flows and which life cycle modules the total environmental impact of a product system comes from. Its answer is a ranking: forty per cent of total GWP from A3 electricity consumption, twenty-five per cent from the steel input, fifteen per cent from A4 transport, and so on. ISO 14044 defines it under ‘identification of significant issues’, the first step of the interpretation phase; in practice, hotspot analysis is the point where an LCA stops being a reporting tool and becomes a design tool, because it says where reduction effort should go. This article explains the three axes of contribution analysis, the Pareto logic, how the hotspot shifts with the impact category, and how to move from the analysis to a reduction scenario.

The three axes of contribution: process, flow, module

The same total can be split three ways, and each split answers a different question. Contribution by process distributes the impact across the unit processes in the model tree: steel production, the electricity grid, lorry transport, natural gas combustion. This view answers ‘which supplier or which operation’ and informs purchasing decisions. Contribution by flow distributes the impact across the inventory's inputs and outputs: how many kg of steel, how many kWh of electricity, how many kg of direct CO₂ emissions. This view answers ‘which material to reduce or substitute’ and informs design decisions. Contribution by module distributes the impact across the EN 15804 life cycle stages (A1-A3, A4-A5, B, C, D) and answers ‘in production, in use or at end of life’. Read together, the three views pin the hotspot down: in module A3, in the electricity flow, in the grid electricity process.

The Pareto logic: the top five processes

In most product systems the bulk of the total impact comes from a small number of processes. Even in a model containing hundreds of background activities, the top three to five processes typically carry seventy to ninety per cent of total GWP; the remaining hundreds form a long tail. This distribution is what gives hotspot analysis its practical value: data collection, verification and reduction effort concentrate on the top five. The procedure is simple: processes are sorted by contribution in descending order, the cumulative percentage is computed, and those below the eighty per cent threshold go on the hotspot list. The same list sets the data quality strategy: site-specific primary data is essential for the top five, an industry average is sufficient for the tail. Every hour spent improving the tail with primary data changes the result almost not at all.

The hotspot shifts with the impact category

The hotspot is not a single place; it moves with the impact category. Where the GWP hotspot is an energy-intensive process (a clinker kiln, an electric arc furnace, grid electricity), the same product's water-use hotspot may be cooling or raw material washing; in toxicity categories an additive below one per cent by mass can carry half the total; in land use a bio-based input comes to the fore. That is why a hotspot analysis run for GWP alone can lead to the wrong reduction decision: switching electricity to renewables lowers GWP but may raise abiotic resource depletion (minerals for the solar panels). It is also why the mandatory indicator set of EN 15804+A2 contains thirteen core categories. A good hotspot report shows a separate contribution ranking for each category and flags the processes that jump between categories.

From hotspot to reduction scenario

Once the hotspot is found there are three reduction levers, and all three are run as scenarios in the same model. The quantity lever reduces the flow at the hotspot: less steel, less electricity, a shorter distance; it belongs to product design and process efficiency. The substitution lever keeps the quantity but sources it from a lower-impact origin: secondary instead of primary aluminium, renewable instead of grid electricity, a clinker substitute instead of clinker. The process lever changes the technology at the hotspot: a more efficient kiln, heat recovery, a fuel switch. When a scenario is run for each lever, the reduction percentage and its cost are placed side by side; that is the input to a marginal abatement cost curve. The critical point: the scenario must be run within the same system boundary and the same functional unit as the hotspot analysis, otherwise the reduction figure cannot be compared.

Hotspots in the EN 15804 module breakdown

In EPD studies hotspot analysis is read by module, and the module pattern gives away the product type. For most construction materials (cement, steel, glass, ceramics, insulation) A1-A3 exceeds eighty per cent of the total and the hotspot lies inside production. For energy-using products (lighting, pumps, air conditioning) B6 operational energy exceeds A1-A3 many times over and the hotspot shifts to the use stage, where reduction means efficiency. For heavy products with a low impact per unit (aggregates, ready-mix concrete) A4 transport rivals A1-A3 and the hotspot depends on distance. Module D is read separately: for materials with a high recycling credit (aluminium, steel) D offsets a substantial part of A1-A3, but under EN 15804 it is never netted into the total. The hotspot in the module breakdown also determines which scenario assumptions (A4 distance, B6 consumption, C3 recycling rate) go into the sensitivity analysis.

Pitfalls: background data and allocation

Hotspot analysis has two systematic pitfalls. The first is background data: if the hotspot sits in an ecoinvent activity, the reduction lever may not be in the manufacturer's hands, and the size of the hotspot changes when the activity's geography (RER, GLO, RoW) or version changes. A process modelled with foreground data and one drawn from the background sit in the same list but not with the same reliability, and the report should tell them apart. The second is allocation: how the impact of a multi-output process is shared between its products determines which product the hotspot appears in. Under mass allocation a low-value by-product carries a heavy burden; under economic allocation the same burden shifts to the main product. Hotspot analysis should therefore be reported together with its sensitivity to the allocation choice; if the ranking changes when the choice changes, the hotspot is an accounting decision rather than a process fact.

• Produce contribution on three axes: process, flow, module; define the hotspot at the intersection of the three
• Sort processes by contribution and list those up to the eighty per cent cumulative threshold
• Repeat the ranking separately for every mandatory impact category; flag the processes that jump between categories
• Collect primary data for the top five processes; an industry average is enough for the tail
• Separate foreground from background contribution; state geography and version for a background hotspot
• Vary allocation and scenario assumptions and check whether the ranking stays stable
• Run quantity, substitution and process levers as scenarios for each hotspot and report the reduction percentage with its cost