Transmission Upgrades for Data Centers: Who Pays for Them?

A large data center can add hundreds of megawatts of demand at a single point on the grid. Even where a high-voltage line runs past the site, the system around it was planned for existing loads. Load studies test whether the system can carry the new demand under normal conditions and under contingencies, with a major line or transformer out of service. When it cannot, upgrades are needed.

• New or expanded substations and switching stations.
• New transmission lines, or rebuilding existing lines with larger conductors.
• Additional transformers at existing substations.
• Breakers, protection and control changes at substations miles from the site.
• Voltage support equipment where a large load weakens local voltage.

The upgrades are identified through the large-load interconnection process. Their cost and who carries it are settled in the studies and the electric service agreement.

Most cost allocation starts with one question: who benefits? Facilities that exist mainly to serve one customer are treated differently from upgrades that strengthen the shared system.

The boundaries are not fixed. A switching station built for one campus may be classified as a network facility if it also serves the grid. A line extension may be sole-use today and shared later if other customers connect. These classifications are often negotiated and can be reviewed by regulators.

There are two broad ways to pay for upgrades, and many utilities use a mix.

Participant funding

Under participant funding, the customer that causes the upgrade pays for it, either up front or over time through a dedicated charge. The logic is cost causation: other customers should not carry costs for facilities they did not need. Participant funding is common for direct-assignment facilities and is increasingly applied to network upgrades driven by very large loads.

Rate-based recovery

Under rate-based recovery, the utility builds the upgrade, adds it to its rate base and recovers the cost from all customers over the life of the asset. The logic is that network upgrades serve the whole system and that a large customer pays its share through its own bills over time. This approach has historically been common for network upgrades serving ordinary load growth.

Hybrid approaches are common. A utility may rate-base a network upgrade but require the customer to post collateral or accept a minimum bill large enough to cover the cost if the load does not materialize. Others credit back part of a customer’s upfront payment as its load generates revenue. Our guide to large-load tariffs and electric service agreements explains those contract tools.

Cost allocation for transmission sits at the meeting point of state and federal authority. Retail service to end-use customers, including data centers, is generally regulated by state public utility commissions or, for public power and cooperatives, by their own boards. Transmission rates and planning in interstate commerce fall under FERC, and in regions with an ISO or RTO, regional planning processes set how certain upgrades are planned and shared.

• Vertically integrated utilities typically handle large-load upgrades under their state-approved tariffs and line extension rules.
• In ISO or RTO regions, the transmission owner, the retail utility and the regional operator may all play a role, and the load may be studied in a regional process as well as locally.
• Municipal utilities and cooperatives set their own policies and may depend on a separate transmission provider or generation and transmission cooperative.
• State commissions in many places are revisiting how large-load costs are allocated, so rules can change over the life of a project.

For more on how market structure shapes this, see ISO and RTO markets and data center siting.

Upgrade cost and timing are among the clearest differences between sites that look similar on a map. Two parcels beside the same voltage line can face very different upgrades because one sits near a strong substation and the other at the end of a long, heavily loaded line.

• Distance. Longer line extensions mean more direct-assignment cost and more easements.
• Upstream constraints. A binding limit miles away can add network upgrades the site’s neighbors did not face.
• Competing requests. Several large loads in one area may be studied together and share, or compete for, the same upgrades.
• Timing. Network upgrades needing new lines can take much longer than the customer’s own substation, and often set the power date.
• Redundancy. A second independent source, discussed in our guide to redundant utility feeds, can trigger its own set of upgrades.

1. 01Which facilities will be classified as customer-specific, and which as network upgrades?
2. 02How does the utility recover network upgrade costs for large loads: participant funding, rates or a mix?
3. 03Is any upfront payment refundable or creditable as load grows?
4. 04What collateral or minimum bill backs the utility’s investment?
5. 05Are any needed upgrades already in a utility or regional plan, and when are they due?
6. 06Could other large requests nearby change the upgrades or their cost share?
7. 07Are cost allocation rules under review by the state commission or regional operator?

The answers rarely come all at once. They firm up as studies advance. Asking them early, for every site on a shortlist, keeps a project from committing to a site whose real cost appears only after the final study.