IND/01 · HYPERSCALE & AI
Data Centers & AI Infrastructure
Power determines location.
We evaluate whether a site can actually energize the load its capital plan assumes — substation by substation, utility by utility, before the LOI.
Available capacity
180 MW
Forecast load
450 MW
Interconnection
Queue #214
Deliverability
Partial
Time-to-power
38 mo
Transmission
500 kV · 4.1 mi
Utility posture
Cautious
Water
1.8 MGD
Fiber
2 diverse
Land
312 ac
Modeled · illustrative composite · S&M synthesis of public ISO/RTO, EIA and utility filings · As of Oct 2026
02 · 03
Power constraint model
Generation → Transmission → Substation → Campus → Compute.
01 · GENERATION
Firm 1.4 GW
02 · TRANSMISSION
N-1 limit 620 MW
03 · SUBSTATION
Headroom 180 MW
BINDING CONSTRAINT
04 · CAMPUS
Design 450 MW
05 · COMPUTE LOAD
Ramp to 410 MW
A 450 MW campus is only as deliverable as its weakest link. In this illustrative case the substation — not generation — sets the realistic first-phase load at ~180 MW until a 2029 upgrade.
Modeled · illustrative · not a live reading · As of Oct 2026
06
Geographic intelligence
Where AI load is — and is not — being absorbed.
Corridor dossier
Northern Virginia
- TRACKED LOAD
- 11.2 GW
- TIME-TO-POWER
- 48–72 mo
Dominion large-load queue saturated; 2029+ energization typical.
Modeled · illustrative · not a live reading · As of Oct 2026
04
Critical pressure vectors
Nine variables that decide a campus.
Site Readiness
82Zoning, land control, water and fiber verified against energization timing.
Power Availability
91Firm headroom versus queue position and behind-the-meter optionality.
Transmission Access
74Constraint inventory and the upgrade calendar that governs it.
Utility Engagement
68How the utility will actually respond to a large-load study request.
Interconnection Risk
88Study cluster risk, network upgrade allocation, withdrawal cascades.
Load Growth
95Corridor-level absorption of AI training and inference load.
Time-to-Power
86Realistic months-to-energize, not the brochure figure.
Campus Expansion
61Whether phases 3–6 have a credible power path.
Infrastructure Constraints
79Binding constraints that cap phase-one load.
05
Interactive risk model
The intelligence graph behind a campus decision.
Stock & Milster Intelligence Graph
07
Decision framework
Five gates before capital commits.
- G1
Is firm capacity available at the substation inside the window?
PASSProceed to utility engagement.
FLAGModel on-site bridge power or re-site.
- G2
Does the transmission path survive N-1 at full load?
PASSPhase plan holds.
FLAGCap phase-one load; map upgrade calendar.
- G3
Will the utility sign large-load terms we can underwrite?
PASSLock tariff assumptions.
FLAGStress cost-of-power and minimum-take.
- G4
Do water, fiber and land control align to energization?
PASSSite is decision-ready.
FLAGResolve sequencing before LOI.
- G5
Does the corridor support expansion beyond phase two?
PASSUnderwrite full campus.
FLAGValue as a single-phase asset.
08
How Stock & Milster helps
Independent diligence for the capital behind compute.
Decision-grade deliverables
- D01Pre-LOI infrastructure intelligence pack
- D02Substation-level site readiness score
- D03Utility posture and counterparty map
- D04Energization timeline counterfactual
- D05Corridor comparison matrix
Questions leadership brings to us
- Is this substation realistically deliverable inside our window?
- Where else in the corridor is firm capacity available?
- What does the rate-case calendar do to cost-of-power?
Stock & Milster advises on the decision. We do not sell the project. Engagements are scoped to geography, facilities and depth of analysis.
10 · EXECUTIVE ACCESS