The AI infrastructure boom has created unprecedented demand for data center capacity. Hyperscalers are committing hundreds of billions in capital expenditures. Private equity firms are acquiring data center portfolios at record valuations. Developers are racing to bring new facilities online.

But beneath the headlines about AI compute demand and semiconductor shortages lies a more fundamental constraint: power.

Specifically, the ability to deliver reliable, continuous power to facilities consuming 50-500 megawatts—and the institutional friction that determines whether that power can actually be procured, financed, and delivered within the timelines capital requires.

This isn’t an engineering problem. It’s an execution risk problem.

And it’s killing deals that looked viable on paper.

II. What Execution Risk Means in Data Center Context

Execution risk refers to the gap between theoretical permissibility and institutional feasibility.

A transaction can be:

Economically attractive (strong demand, favorable pricing)
Legally permissible (zoning approved, permits in hand)
Technically feasible (engineering validated, site suitable)

And still fail because institutional pathways cannot be reconciled within acceptable governance thresholds.

In sanctions-exposed commodity markets, we see transactions collapse when:

Banking declines settlement due to AML/correspondent risk
Insurance withdraws coverage via P&I sanctions exclusions
Routing opacity triggers compliance escalation
Internal committees cannot defend residual risk to audit functions

The pattern is identical in data center development.

Projects fail not because demand is weak or sites are unsuitable, but because:

Grid interconnection extends years beyond project timelines
PPA structures contain exposure that breaks economic models
Project finance requires certainty that utilities cannot guarantee
Regulatory processes compress feasibility windows

Legal site control is insufficient. Power deliverability is the binding constraint.

III. Grid Interconnection: The 7-Year Queue Problem

Grid interconnection—the process of connecting a new load to the electricity transmission system—has become the primary execution barrier for data center development.

Current Reality:

Average interconnection timeline in major markets: 24-48 months
PJM (mid-Atlantic): 1,000+ projects in queue totaling 200+ GW
MISO (Midwest): 5-7 year average study completion time
CAISO (California): Competing load from manufacturing and EV charging

Why This Matters:

Data center developers typically operate on 18-24 month construction timelines. Private equity firms model 3-5 year hold periods. Hyperscalers plan capacity 2-3 years forward.

A 48-month interconnection delay doesn’t just push timelines. It kills IRR assumptions, breaks financing covenants, and renders projects uneconomical.

Execution Failure Points:

Transmission Constraints
Existing transmission infrastructure was designed for gradual load growth. AI-driven data center demand creates step-function increases that exceed planned capacity.
Substation Capacity Limits
Adding 200-500 MW to a local grid requires substation upgrades costing $100M-$300M. Utilities must secure regulatory approval and financing—processes measured in years.
Competing Load
Manufacturing reshoring, EV charging infrastructure, and residential electrification create competing demands on constrained capacity.
Study Backlog
Grid operators are overwhelmed. Interconnection studies that once took 12 months now take 36-48 months—and that’s before construction begins.

Execution Impact:

Projects die in queue before capital is deployed. Developers commit to land acquisition and engineering without certainty that power can be delivered. By the time interconnection infeasibility is confirmed, millions in sunk costs are unrecoverable.

IV. PPA Structure Risk: When “Guaranteed Power” Isn’t Guaranteed

Power Purchase Agreements (PPAs) are the contractual mechanism through which data centers secure electricity supply. But not all PPAs are created equal—and the difference determines execution viability.

Curtailment Exposure in Renewable PPAs:

Many data centers pursue renewable energy PPAs to meet sustainability commitments. But renewable generation profiles (solar peaks midday, wind varies by weather) create shape risk when matched against 24/7 data center load.

PPA language like “subject to availability” or “curtailable during grid stress events” creates economic exposure:

Data center requires 100 MW continuous
PPA provides average 100 MW but with 20-30% variability
During low-generation periods, data center must purchase backup power at spot rates
Spot power during grid stress = 5-10x contracted rates

Economic Model Breakdown:

A data center project modeled at $40/MWh PPA rate can see effective costs spike to $80-$120/MWh when curtailment and backup procurement are factored.

This breaks:

Operating cost assumptions
Profit margin targets
Debt service coverage ratios
Investor return expectations

Firmness vs Interruptibility:

“Firm power” means guaranteed delivery regardless of grid conditions. “Interruptible power” means supply can be curtailed during high-demand periods.

Many PPAs contain hybrid language that appears firm but includes carve-outs for:

Force majeure events
Grid operator directives
Transmission unavailability
Renewable generation shortfalls

Execution Impact:

Deals close based on PPA headline rates. Six months into operations, curtailment events occur. Economic model breaks. Equity partners demand renegotiation or exit. Project distress follows.

V. Project Finance: Banking Requires Certainty That Doesn’t Exist

Lenders financing $200M-$1B data center projects require bankable certainty around power delivery. Standard covenant packages include:

Typical Lender Requirements:

Utility Commitment Letter
Binding commitment from utility to provide interconnection by specific date
Completed Interconnection Study
Engineering validation that grid can accommodate load
PPA with Investment-Grade Counterparty
Power purchase agreement with utility or power provider rated BBB or higher
Performance Guarantees
Financial penalties if power delivery fails
Reserve Accounts
Escrowed funds to cover cost overruns or delays

The Problem:

Many utilities cannot provide the certainty lenders require:

Interconnection timelines are estimates, not guarantees
Transmission upgrades depend on regulatory approval
Substation construction faces permitting and supply chain risks
Utility credit ratings may not meet investment-grade thresholds

Execution Impact:

Developers reach financial close milestones, then discover:

Utility won’t provide binding commitment letter
Lender won’t fund without firm delivery guarantee
Circular dependency: can’t secure PPA without financing, can’t secure financing without PPA

Project stalls. Capital markets window closes. Deal dies.

VI. Regulatory Approval: Multi-Year Timelines in a Fast-Moving Market

Even when grid capacity exists and utilities are willing, regulatory approval processes create execution risk.

Approval Requirements:

Tariff Structures and Rate Cases
New large loads require utility rate cases to determine cost allocation. Residential ratepayers may oppose rate increases to fund data center infrastructure.

Environmental Review Processes
NEPA reviews, endangered species consultations, water usage permits, air quality assessments—each adds 6-18 months.

Community Opposition and Permitting
Local communities increasingly oppose data centers due to:

Water consumption (cooling systems)
Noise (backup generators, cooling equipment)
Grid strain (“why should we subsidize tech companies?”)
Land use (industrial facilities in residential areas)

Execution Impact:

Virginia (historically favorable): approval timelines extending from 12 to 24+ months
California: community opposition adding 18-36 months to permitting
Oregon: legislative proposals to restrict data center water usage

Projects that penciled at 24-month timelines now face 48-60 months. IRR collapses. Capital reallocates.

VII. Case Study Framework: Three Sites, Three Outcomes

To illustrate how these execution risks compound, consider three hypothetical data center sites evaluated simultaneously by a private equity firm:

SITE A — Northern Virginia

Grid Capacity: Existing capacity available, 18-month interconnection timeline
Utility: Dominion Energy (A- credit rating), proven track record on large data center loads
PPA Structure: 15-year firm power agreement, no curtailment provisions
Regulatory: Established approval pathway, supportive local government
Project Finance: Multiple lenders competed for deal, favorable terms

Execution Probability: 85% (GO)

Decision: Proceed to acquisition

SITE B — Ohio

Grid Capacity: 36-month interconnection queue, requires $150M substation upgrade
Utility: Requires board approval for capex, timeline uncertain
PPA Structure: Renewable-heavy (60% wind/solar), shape risk present
Regulatory: Environmental review required, 12-18 month timeline
Project Finance: Lenders require firmer utility commitments than currently available

Execution Probability: 55% (ESCALATE)

Decision: Hold, require mitigation plan before proceeding

SITE C — California

Grid Capacity: Limited availability, competing load from manufacturing electrification
Utility: Transmission upgrades required, no funded plan
PPA Structure: Power costs 40% above pro forma assumptions
Regulatory: Community opposition organized, permitting timeline 24-36 months
Project Finance: Multiple lenders declined, citing execution uncertainty

Execution Probability: 15% (NO-GO)

Decision: Pass, reallocate capital

VIII. The Execution Risk Intelligence Gap

Traditional due diligence processes are not designed to assess institutional execution risk.

What M&A Advisors Cover:

Valuation and purchase price
Legal structure and documentation
Tax optimization

What Engineering Firms Cover:

Technical feasibility
Construction cost estimation
Equipment specifications

What Legal Counsel Covers:

Contract review
Regulatory compliance
Title and permitting

What No One Systematically Covers:

Grid interconnection queue dynamics
Utility financial capacity and timeline reliability
PPA structure risk and curtailment exposure
Project finance bankability under realistic scenarios
Regulatory approval probability and timeline variance

This is the execution risk intelligence gap.

And it’s costing institutional investors hundreds of millions in failed deals, delayed timelines, and unrecoverable sunk costs.

IX. How We Map Execution Probability

Our methodology for assessing data center execution risk integrates:

1. Grid Capacity Analysis

Interconnection queue position review
Transmission constraint mapping
Substation capacity assessment
Competing load evaluation

Primary Sources:

ISO/RTO interconnection queue data (FERC Form 715)
Utility integrated resource plans
Transmission planning studies
Regional load forecasts

2. PPA Structure Risk Evaluation

Curtailment exposure quantification
Generation profile vs load requirement matching
Firmness and availability guarantee review
Backup power cost modeling

Primary Sources:

PPA templates from comparable projects
Utility tariff structures
Regional power market pricing data
Renewable generation profiles

3. Project Finance Bankability Assessment

Lender covenant requirement review
Utility creditworthiness evaluation
Reserve and guarantee structure analysis
Historical close rate for similar profiles

Primary Sources:

Lender term sheets from comparable financings
Utility credit ratings and financial disclosures
Project finance market intelligence

4. Regulatory Approval Timeline Mapping

Permitting pathway identification
Historical approval timeline analysis
Community opposition risk assessment
Regulatory bottleneck identification

Primary Sources:

State public utility commission filings
Environmental review databases
Local government meeting minutes
Media coverage of community responses

Output:

Go / Escalate / No-Go Assessment with Probability Scoring

GO (>70% execution probability): Power deliverable within project timeline, acceptable risk profile
ESCALATE (40-70% execution probability): Significant risks require mitigation, proceed conditionally
NO-GO (<40% execution probability): Execution infeasible under current pathway, capital better deployed elsewhere

Delivery: 10-15 business days, board-ready PDF report

X. Conclusion: Execution Risk as a Category of Infrastructure Intelligence

The AI infrastructure boom is real. Data center demand is real. The capital committed is real.

But demand doesn’t guarantee execution.

And execution is where deals die.

Grid interconnection queues measured in years. PPA structures containing hidden exposure. Project finance requiring certainty utilities can’t provide. Regulatory processes compressing feasibility windows.

These aren’t edge cases. They’re the new normal.

Institutional investors deploying $50M-$500M in data center acquisitions or developments need execution risk intelligence—not just traditional due diligence.

Because discovering you can’t get power after you close the deal isn’t due diligence.

It’s negligence.

Global Assessment provides fixed-scope execution risk intelligence for:

Data center site selection and acquisition
Energy infrastructure feasibility
Manufacturing facility siting
Power-intensive facility development

Services:

Execution Feasibility Screen: $3,500 (5 business days)
Energy Infrastructure Feasibility: $25,000 (10-15 business days)
Comprehensive Execution Mapping: $15,000+ (3-4 weeks)

Learn more: globassessment.com/execution-risk