System expansion solves the multi-product problem differently to allocation. Instead of splitting impacts between products, you expand the system boundary to account for what those products replace in the wider economy.
ISO 14044 actually prioritises system expansion over allocation. The standard says: try to avoid allocation first. System expansion is how you do that.
Why System Expansion Matters
Return to the WHY. When a process makes multiple products, those products go into the economy and displace other products. A waste treatment plant that generates electricity displaces grid electricity. A biorefinery producing biodiesel and animal feed displaces both fossil diesel and soy meal. These displacement effects are real market consequences.
System expansion captures this reality. Allocation just splits up the impacts. System expansion asks: what would have happened otherwise?
Waste Incineration Example
A municipal waste incinerator processes 100,000 tonnes of waste annually. The process generates heat, which drives a steam turbine producing 50,000 MWh of electricity sold to the grid.
Allocation approach: Split the waste treatment impacts between waste management (the primary service) and electricity generation (the co-product). Perhaps 80% to waste treatment, 20% to electricity based on economic value.
System expansion approach: The system boundary includes both the incinerator and the grid electricity being displaced. The 50,000 MWh fed into the grid means 50,000 MWh of conventional grid generation doesn’t happen. Credit the waste treatment system with avoided impacts from that displaced electricity.
The expansion reflects reality. That electricity genuinely prevented a gas power station or coal plant from generating the same amount. The waste treatment now carries its full process impacts minus the credit for displaced generation.
When System Expansion Works
System expansion requires knowing what’s being displaced. For commodity products with established markets, this works well. Grid electricity has a known generation mix. Biodiesel displaces fossil diesel with known production impacts. Animal feed displaces soy meal or other feed ingredients.
The method handles multiple co-products naturally. A palm oil mill produces crude palm oil, palm kernel oil, palm kernel meal, and process steam. Each product displaces something in the market. Expand the system to include all displacement effects, credit each one, and the primary product (crude palm oil) carries the net impact after all credits.
When System Expansion Fails
Displacement must be identifiable and credible. If a co-product has no established market or gets landfilled rather than sold, there’s no displacement to credit. The system expansion becomes speculative.
Novel products create problems. What does lab-grown meat displace? Conventional beef, certainly, but also potentially chicken, pork, or plant proteins depending on consumer choices. The displacement isn’t clear-cut, making system expansion assumptions arbitrary.
Market dynamics complicate things further. Producing more biodiesel doesn’t simply reduce fossil diesel production one-for-one. It affects prices, which affects demand, which affects production decisions across multiple sectors. True displacement effects ripple through economic systems in ways LCA struggles to capture.
Allocation as the Fallback
When displacement is unclear, ambiguous, or speculative, allocation becomes necessary. ISO 14044 acknowledges this – it presents allocation as the method to use when subdivision and system expansion aren’t feasible.
The gold mine from the allocation article demonstrates this perfectly. The waste rock displaces nothing. It has no market, no buyers, no economic function. System expansion offers no credit because there’s no displacement. Allocation then becomes the only practical approach, and economic allocation correctly assigns all impact to gold because that’s WHY the mine exists.
The Methodological Hierarchy
ISO 14044 establishes a clear preference order:
- Avoid allocation by subdividing processes or expanding system boundaries
- Use allocation based on physical relationships when subdivision and expansion aren’t possible
- Use allocation based on economic value when physical relationships don’t apply
System expansion sits at the top of this hierarchy for good reason. It captures real market effects rather than artificially splitting impacts. But it only works when displacement is real, identifiable, and defensible.
The choice between system expansion and allocation isn’t about mathematical convenience. It’s about representing real-world consequences of production decisions and product markets.
