AI Power & Site Readiness
Company: Applied Digital
Verdict
Phase 1 appears already in service (realized milestone), but deliverable capacity is below the peer cohort median under the curtailment-risk proxy. Carbon intensity remains worse than the pinned peer cohort. Renewable content is shown as a state grid‑mix proxy. The final score is a weighted blend of schedule, deliverability, cost, carbon, and reliability views.
Top drivers (why)
- 1Schedule outlook (time-to-power)In-service hit probability is 1.0 and expected slippage is 0 days (realized Phase 1 milestone), but schedule-chain inputs are proxy-heavy and marked stale; validate current execution details before external distribution.
- 2Power deliverability strengthDeliverable capacity by target is below the peer cohort median under the curtailment-risk proxy; validate node-level deliverability and curtailment exposure for the 50 MW tranche.
- 3Carbon intensity headwindExpected emissions intensity is worse than the peer cohort median, while renewable content (grid‑mix proxy) is above the top quartile; validate contracted renewables posture before external distribution.
KPI snapshot (median / top quartile benchmarks)
| KPI | Current | Benchmark (Median / Top quartile) | Confidence |
|---|---|---|---|
Site Readiness Index Stale | 86%index ProxyBetter than median | 45% / 54% median / top quartile | 71% proxy-heavy |
Probability of Hitting In‑Service Date Stale | 100% InferredBetter than median | 30% / 46% median / top quartile | 70% |
Schedule Slippage (Days) Stale | 0days InferredBetter than median | 341 / 456 median / top quartile | 70% |
Deliverable Capacity by Target (MW) Stale | 37.137MW ComputedWorse than median | 44.73 / 68.89 median / top quartile | 79% |
Blended Effective Power Rate ($/MWh) Current | $85USD/MWh Computed | — / — median / top quartile | 81% |
Expected Emissions Intensity (kgCO₂e/MWh) Current | 420.299kgCO2e/MWh ProxyWorse than median | 348.95 / 458.42 median / top quartile | 58% proxy-heavy |
Reliability Risk Index Stale | 13%index ComputedBetter than median | 44% / 64% median / top quartile | 55% |
Total Delay Cost Exposure ($) Stale | $0 Proxy | — / — median / top quartile | 79% proxy-heavy |
Renewable Content (%) Current | 26% ComputedBetter than median | 11% / 23% median / top quartile | 95% |
Moves that close the gaps
This playbook turns the current posture into the moves that close the gaps, strengthen readiness, and sharpen the next decision.
- 1Secure a firmer power-delivery path for the next 50 MW tranche
Owner: Owner / developer and the utility-facing project team
The latest capacity view sits below the peer median under the curtailment-risk proxy, so the next tranche still looks softer than the headline score suggests.
Readiness effect: Turns the next tranche from capacity-on-paper into a more commitment-backed power path.
- 2Lock the next utility, procurement, and construction milestones
Owner: Owner / developer, EPC, and key suppliers
Phase 1 appears in service, but the next schedule posture still leans on proxy-heavy, stale inputs rather than clearly locked milestones.
Readiness effect: Moves the schedule story from plausible to more commitment-based for pricing, sequencing, and allocation decisions.
- 3Choose and contract the energy-supply posture for the next phase
Owner: Procurement lead, power-architecture decision maker, and delivery leadership
Carbon intensity remains worse than the peer median, and the current renewable number is still a grid-mix proxy rather than contracted coverage.
Readiness effect: Reduces uncertainty around how the next phase will actually be powered and how credible the readiness posture is under scrutiny.
- 4Do not treat the next 50 MW tranche as fully committed
Owner: Commercial owner, delivery team, and counterparties allocating effort or capacity
The current score is strong, but not strong enough to justify treating the next phase as fully de-risked.
Readiness effect: Keeps pricing, sequencing, and resource allocation tied to what is actually locked today.
Power delivery, schedule, and energy supply still need to be locked.
