Site Against Failure

The second rung — place a renewable portfolio to reduce correlated low output rather than to maximise the capacity factor on a brochure.

Stage 2 of 4 · Weeks 3–6

Due

18 September 2026

Weight

20% of the course

Takes as input

an operative definition of the event

Produces

a portfolio with a correlation profile

What you are producing#

A generation portfolio — what technology, how much, and where — together with its correlation profile during your event. The portfolio is the artefact; the correlation profile is the argument.

What earlier work this uses#

Your A1 definition, unchanged. If you revise it here, say so and say why; a moving definition is allowed, a silent one is not.

How the canonical definition constrains this#

The event is a simultaneous low across wind and solar. That is what makes siting hard: a portfolio optimised for annual yield tends to concentrate where the resource is best, and the resource is best in places whose weather often arrives together. Weeks 3 to 6 are the machinery for this — correlation, spatial extent, depth below cut-in, and the hours when the wind–solar complement stops working.

What this feeds#

The 72-Hour Dispatch operates the system you build here. A portfolio that looks strong on paper and collapses in hour 40 of dispatch is a more useful submission than one that is never tested.

What counts as a defensible answer#

Evidence that you sited against a failure mode instead of against a mean. The question is not “how much does this generate?” but “when this portfolio is short, is it all short at once?”