Skip to main content

You are viewing a project that is currently in draft state for the Standards Working Group in the Green Software Foundation. This project should not be considered finished or officially supported in any way by the Green Software Foundation or its members.

Embodied Water

Embodied Water

In short: embodied water (W_M) is water used across the lifecycle of the hardware and infrastructure a software system depends on — raw-material extraction, component and hardware manufacturing, transportation, and end-of-life treatment. It is distinct from operational water, which is used while that hardware is actually running. Embodied water asks: what's your fair share of the water it took to build the machine you're borrowing? You're only charged for the slice of its lifetime you actually used, and the slice of its capacity you actually used — not the water it took to build the whole thing.

If you've worked with the SCI specification's embodied carbon (M) calculation, this will already feel familiar — it is the identical time-share/resource-share allocation mechanism, applied to a water total instead of a carbon total.

Allocating shared hardware​

Most software does not have hardware to itself; it shares it with other workloads. The specification allocates only the fraction of a piece of hardware's total embodied water that this software is actually responsible for.

Shape: total water to build the whole machine × (fraction of its life you used) × (fraction of its capacity you used) = your share. Like splitting a shared taxi fare by both how many stops you rode for and how many people were in the car.

Worked: building the server cost 4,000 L across its lifetime; you used it for 1 month out of an expected 48-month life (a time share of 1/48); you used 4 of its 32 CPUs (a resource share of 4/32); your share = 4,000 × (1/48) × (4/32) ≈ 10.4 L. (illustrative placeholder, not real data)

Precise notation, using the same time-share / resource-share mechanics the SCI specification uses for embodied carbon:

W_M,b = TW_b × TS × RS   (for each basis b: c = consumption, w = withdrawal)

where TW_b is the hardware's total lifecycle water for basis b, TS is the time share of the hardware's expected lifetime reserved for this software, and RS is the resource share of the hardware's total resources reserved for it. Both shares are themselves ratios:

TS = TiR / EL
RS = RR / ToR
  • TiR — time reserved for the software; EL — the hardware's expected operational lifetime.
  • RR — resources (e.g. CPU cores, memory) reserved for the software; ToR — total resources the hardware has available.

Putting it together:

W_M,b = TW_b × (TiR / EL) × (RR / ToR)

What this requires in practice​

The specification requires an embodied water estimate for every hardware component that materially contributes within the software boundary — not just the most obvious one (e.g. the server), but network equipment, storage, and other infrastructure where material. The most granular lifecycle data available should be used; where measured or supplier-specific data isn't available, disclosed modelled estimates may be used instead.

Two practical questions follow from this: where do the lifecycle water totals (TW_b) themselves come from — see Datasets — and how should TS and RS actually be chosen for a real, shared piece of cloud hardware — see Allocation Calculations.


Please submit any comments you have here.