Data Centers

Objective

Forecasting data center loads in medium- (10-year) and long-term (30-year) planning horizons. Data centers’ massive and growing energy needs place high demands on reliable energy. Utilities must balance support for economically beneficial loads with the risk of stranded assets.

Approach

A categorization process based on three dimensions to represent 15 data center industry segments, including business model, location preferences, and growth aggressiveness. Our multi-factor methodology ensures credibility, transparency, and traceability, with a proven record of technical use. It analyzes and forecasts load and growth trends by amount and location.

Our process is based on deep research and a data-driven, reverse-engineered model of a segment’s business model, location preferences, and growth aggressiveness.

Results

Key takeaways include:

• Future data center loads will likely develop near existing facilities.
• New data center clusters may emerge based on available transmission capacity, incentives, and other factors.
• Forecast indicates continued robust growth of data center loads for the next 10+ years.
• Interconnection queue does not adequately capture future new load after year five.

Data Center Modeling in Load Forecasts

We use a multi-factor approach based on three basic dimensions to track and model 15 slightly- to very-different data center business models.

1. Purpose/market served.
2. Function and design.
3. Type of owner/business model