Energy & Electrical Impact

Power consumption, grid stability risk, and utility infrastructure upgrades for IronGate expansion

Overview

Data centers consume massive, continuous power loads. The IronGate facilities' planned 100 MW build-out is equivalent to the electricity use of ~80,000 homes running 24/7. This creates two critical concerns: (1) local grid stability under sudden load swings, and (2) utility cost responsibility during expansion.

NERC Warning (2026): The North American Electric Reliability Corporation (NERC), the continent's top grid-reliability regulator, has flagged data centers as "one of the greatest risks to grid stability" due to rapidly fluctuating power demand that can stress transformers and local equipment.

Power Consumption: IronGate's Load

Current State (Sept 2026):

Facility Live Capacity Under Construction Planned Total Build-Out
401 Bielenberg 10–16 MW — ~50–55 MW ~65–66 MW
500 Bielenberg 4 MW 10 MW 20 MW ~34 MW
COMBINED ~20 MW 10 MW ~70–75 MW ~100 MW

Electricity Equivalent:

Grid Stability: The NERC Risk

Why Data Centers Are a Grid Risk:

NERC quote (Aug 2026): "Rapidly fluctuating data center loads are among the greatest risks to grid stability. Equipment shortages and local infrastructure strain from fast-swinging loads could contribute to cascading failures if not carefully managed."

Utility Upgrades & Cost Responsibility

Minnesota 2025 Law (Large-Load Tariff):

A new state law (effective 2025) now requires data centers to pay for their own incremental grid costs through "large-load tariffs." This means:

Unknown: Will Xcel's cost calculations cover ALL infrastructure strain (including voltage stability measures), or only the most direct upgrades?

Key Questions for Interim Ordinance Study

Grid Stability & Risk:

Utility Upgrades & Timeline:

Resilience & Redundancy:

Related Energy Topics

Backup Generators: Data centers install diesel or gas generators for grid outage backup. These require fuel delivery, maintenance testing (which generates noise and exhaust), and environmental compliance. Test runs can be as loud as jet engines.

Cooling Load: HVAC and chiller systems consume nearly as much power as the compute equipment itself. Total load is thus higher than just the server capacity.

AI Impact: As data center workloads shift toward AI training and inference, power consumption and load volatility are expected to increase, making grid stability even more critical.