Capacity Forecast¶
Planning forecast — instantiates Anticipatory Forecasting
Converts a forecast of future load into the resource capacity it will require, then starts the long-lead provisioning so the capacity is in place before the peak arrives.
A demand estimate is not yet a plan; it becomes one only when it is translated into how much of a resource must be standing ready, and by when. Capacity Forecast takes an anticipated load — customers, requests, patients, megawatts of draw — and works out the capacity it implies, then anchors that number to the calendar of a long-lead provisioning decision. Its defining move is the pivot from what will be demanded to what must be built, contracted, or staffed to serve it, sized against the time it actually takes to acquire that capacity. Where a demand forecast answers "how big will the wave be," a capacity forecast answers "how much do we have to have ready, and when do we have to start" — and it exists to make provisioning begin early enough that the resource is present at the moment of need, not scrambled for after.
Example¶
A regional grid operator forecasts that a summer heat wave will push peak electricity demand to roughly 62 gigawatts in July, against a normal-year peak nearer 55. A demand figure alone changes nothing. The capacity forecast converts it: to serve a 62 GW peak with a prudent planning reserve margin, the system must have on the order of 71 GW of firm generation committed, plus enough transmission to move it.[n1] Then it reads the calendar backward. New generation cannot be conjured in July; capacity contracts must be signed, deferred maintenance rescheduled out of the peak window, and demand-response agreements lined up — decisions with lead times measured in months to years. So the forecast is dated: the reserve must be locked by early spring, and the maintenance shuffle by late winter. The operator commits capacity in February for a peak in July, because the alternative — discovering the shortfall in July — is rolling blackouts.
How it works¶
What distinguishes a capacity forecast from the demand estimate feeding it is the translation-and-timing step:
- Convert load to required capacity. Apply the service standard — a reserve margin, a utilization ceiling, a queueing target — to turn expected load into the capacity that must be provisioned to meet it without breaking.
- Work backward from lead time. Identify the slowest-to-acquire piece of that capacity and set the commit date from its lead time, not from when the peak lands.
- Attach the provisioning action. Name the concrete acquisition — contract, hire, build, reschedule — so the forecast terminates in a decision rather than a report.
- Bound the horizon. State the window the capacity must cover, so provisioning is neither too brief to matter nor open-endedly overbuilt.
Tuning parameters¶
- Service standard — the reserve margin or utilization target the capacity must satisfy. Higher standards cut the chance of shortfall but provision more idle headroom.
- Lead-time assumption — how long acquisition is assumed to take. Padding it commits earlier and safer; trimming it frees flexibility but risks starting too late to finish.
- Provisioning granularity — one large commitment versus staged increments. Staging tracks the forecast as it firms up but carries per-step overhead and may miss volume economics.
- Coverage horizon — how far past the peak the provisioned capacity must hold. Longer horizons smooth procurement but risk sizing to a load that never recurs.
- Reversibility bias — preferring capacity that can be released (short contracts, demand response) over irreversible builds when the load estimate is soft.
When it helps, and when it misleads¶
Its strength is that it refuses to let a good demand estimate die as a slide. By dating the provisioning to real acquisition lead times, it is the mechanism that actually gets the resource in place before the wave — and by preferring releasable capacity when the forecast is soft, it can hedge its own error.
Its failure mode is provisioning to a peak that does not arrive, or arrives smaller: capacity acquired on a long lead time is expensive to unwind, so an inflated load estimate becomes stranded cost, and a service standard set too high quietly bakes in permanent overbuild. The classic misuse is treating the reserve margin as a free safety knob and ratcheting it up after every near-miss until the system carries capacity it never uses. The guarding discipline is to size to a stated, defensible standard rather than to nerves, prefer reversible capacity when the load is uncertain, and re-check the commit before the lead-time clock forces it — not after.
How it implements the components¶
Capacity Forecast fills the translate-and-provision side of the archetype — turning an anticipated state into a timed acquisition, not estimating the state itself:
forecast_target— names the future state as a required-capacity figure (firm generation, staffed seats, provisioned cores), the number provisioning is sized to.decision_horizon— bounds the window the capacity must cover, so provisioning matches the period of need.lead_time_requirement— sets the commit date by working backward from the slowest-to-acquire piece of capacity.preparation_action— terminates the forecast in a concrete acquisition: a contract signed, a build funded, a maintenance window moved.
It does not estimate the underlying load — the signal_basis, uncertainty_range, and update_rule that produce the demand figure belong to its nearest twin, Demand Forecasting, which this mechanism consumes. And its provisioning serves one dated forecast, not the cross-future option bundle of Scenario-Informed Preparation.
Related¶
- Instantiates: Anticipatory Forecasting — Capacity Forecast is the step that turns an anticipated load into resource that is standing ready in time.
- Consumes: Demand Forecasting supplies the load estimate this mechanism converts into required capacity.
- Sibling mechanisms: Demand Forecasting · Trend Projection · Reference-Class Forecast · Early Warning Forecast · Forecast Trigger Dashboard · Scenario-Informed Preparation · Forecast After-Action Review · Rolling Forecast Review
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Converts a forecast of future load into the resource capacity it will require, then starts the long-lead provisioning so the capacity is in place before the peak arrives, making its operative form a computation, comparison, model, or analytic representation used to infer, estimate, or choose.
Independent corroboration: The frozen evidence defines Capacity Forecast as 'Converts a forecast of future load into the resource capacity it will require, then starts the long-lead provisioning so the capacity is in place before the peak arrives', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Convergent development
Present-day reach: Multi-domain
Rationale: Operations planning established translating forecast load, service standards, and reserve margin into timed resource provision.
Related originating lineages:
- Economics & Finance — Investment appraisal contributes demand scenarios and the cost of too-early or too-late expansion.
- Engineering & Design — Infrastructure engineering contributes long-horizon capacity and commissioning constraints.
- Logistics & Supply Chain Management — Inventory and distribution planning contribute seasonal and lead-time-aware demand forecasts.
Review resolution: Operations research is the agreed primary lineage because forecasting demand against service capacity is a core planning problem. Logistics, engineering, and economics independently contribute seasonal demand, physical expansion lead times, and investment horizons, so the record is convergent.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The planning reserve margin is the firm generating capacity a power system holds above forecast peak demand, expressed as a percentage, to cover forecast error, outages, and extreme weather. It is the service standard that converts a demand peak into a required-capacity target, and setting it is a genuine trade-off between reliability and the cost of idle capacity. ↩