Comparing Water Intensity Fairly

A defensible water comparison fixes the metric, the boundary, and the life-cycle scope before any technology is ranked.

Water comparisons between generation technologies are frequently misleading because the ground rules shift between them. To compare fairly, three things must be pinned down first, and stated with the result.

The three ground rules

• Metric: withdrawal or consumption? They can differ by an order of magnitude for the same plant.
• Boundary: operational cooling only, or the full life cycle including fuel extraction and manufacturing?
• Basis: per unit of energy delivered, over what capacity factor and lifetime?

Change any one and the ranking can flip. A once-through plant looks water-hungry on withdrawal and modest on consumption. A biofuel or a mined-fuel chain can carry large upstream water that operational figures hide. An intermittent source has low operational water but its energy-basis figure depends on capacity factor.

How Kronos applies this

We make an architecture claim for the burner — direct energy conversion removes the steam cycle's condenser water — rather than publishing a water-intensity number for an unbuilt machine. Where the breeder (Hyperion) is discussed, we tie water to its cooling choice. When we compare to other sources, we use the same metric and boundary on both sides, and we flag any figure that is illustrative rather than measured. A rejected internal comparison that relied on a scale mismatch is a reminder of how easily this goes wrong.

A further discipline is to distinguish nameplate from delivered energy. A source that runs part of the time delivers less energy per unit of water or land than its nameplate implies, so an energy-basis comparison must state the capacity factor assumed. Firm, high-availability sources and intermittent sources cannot be compared on nameplate alone without distorting the result.

Applied here, this means we do not quote a per-unit water figure for the burner, because doing so would imply a measured capacity factor and a measured cooling load that no unbuilt machine has. The architecture claim survives this scrutiny; a bare number would not.