Northern Virginia · capacity / resilience / supply chain

What does a data-center buildout ask of the grid?

A transparent planning experiment: demand pathways become capacity decisions, outage tests, and material consequences. The results below are representative-cycle optimization outputs—not an annual forecast or a utility planning commitment.

central peak MWcentral transmission MWcentral outage starts fully servedPython / GAMSPy reduced-case validation

From demand to consequence

The model follows one auditable chain: public demand evidence becomes an investment decision, an outage test, and a capacity-linked supply-chain estimate.

01Demand evidenceEIA · NOVEC/PJM · Cleanview
02Capacity choiceimport · transmission · solar · storage · gas
03Resilience test24 independent four-hour outage starts
04Material shadowsteel · copper · concrete · lithium

Three pathways into 2030

Low follows the benchmark trajectory; central adds a staged project ramp; high is a stress case. Cleanview capacity supplies project context, not direct demand.

Loading model snapshot...
lowcentralhigh

Capacity follows the peak

Select a pathway to inspect the least-cost capacity bundle under reserve and resilience constraints.

Selected pathway: loading

When the import path disappears

Each point is an independent four-hour outage-start test. High growth reaches the model's resource limits and leaves critical load unserved in some starts.

Loading outage tests...

What the bundle costs

The cost breakdown is an optimization accounting view across the representative cycle; it is not an annualized utility revenue or rate forecast.

Infrastructure has a material shadow

These are capacity-linked accounting estimates using analyst intensity ranges, not facility bills of materials. Quantities are tonnes; supply limits are illustrative scenario thresholds, not verified shortages.

Evidence, not false precision

Inputs

  • Observed / derived EIA regional load
  • NOVEC/PJM forecast benchmark
  • Public Cleanview project status
  • Published planning context

Boundaries

  • Representative 24-hour cycle
  • Analyst technology and material ranges
  • Independent outage survivability tests
  • No facility bill of materials
The reduced central hour-17 formulation matches in Python and GAMSPy within documented tolerance. GAMS remains an offline modeling environment; this page publishes only sanitized outputs.