Key takeaways
- The meter is the dividing line between the customer’s system and the utility’s.
- Behind-the-meter resources serve the load directly and reduce what the grid must deliver.
- Front-of-the-meter resources connect to the grid and reach the data center through it.
- Many campuses combine both: grid service plus on-site generation or storage.
- Rules for behind-the-meter arrangements, especially with large generators, vary by state, utility and market.
01What “behind the meter” and “front of the meter” mean
The utility revenue meter marks where the utility’s system ends and the customer’s begins. Everything on the utility side, including transmission lines, substations owned by the utility and power plants selling into the grid, is front of the meter. Everything on the customer side is behind the meter.
For a data center, behind-the-meter usually means generation or storage inside the campus electrical system: gas turbines or reciprocating engines, fuel cells, batteries, solar, or in some proposals a power plant next door feeding the campus directly. Front-of-the-meter means power that comes through the grid, whether from the utility’s own plants, market purchases or a contracted generator somewhere else on the system.
Backup generators are technically behind the meter, but they are not what people usually mean. The term is mostly used for resources that run routinely, either to serve the load or to cut what the grid has to deliver.
02How each arrangement works for a data center
| Factor | Behind the meter | Front of the meter |
|---|---|---|
| Where the resource connects | Customer side of the meter, inside the campus system | Utility or transmission system |
| Who usually owns and operates | Customer or a third party under contract | Utility, independent generator or market participant |
| Effect on grid demand | Reduces net demand the grid must serve | None directly; grid delivers the full load |
| Interconnection path | May still need utility review, and a generator interconnection if it can export | Load interconnection for the data center; separate generator interconnection for the resource |
| Main permits | Air permits, fuel supply, local land use for the plant | Handled by the resource owner, often at another site |
| Reliability backstop | Grid service if retained; otherwise on-site redundancy | Grid reliability planning standards |
| Siting impact | Needs land, fuel access and emissions headroom on or near the campus | Needs strong transmission access at the campus |
In practice the line blurs. A campus may take firm grid service for most of its load and run on-site generation to cover a gap. A plant next to a campus may be front of the meter on paper, selling into the grid, while also supplying the campus under a contract. How the arrangement is classified depends on metering, ownership and the rules of the utility and market.
03Why developers consider behind-the-meter power
- Speed. Where grid upgrades would take years, on-site generation can sometimes start serving load sooner, as bridge power or as a long-term supply.
- Smaller grid request. Serving part of the load on site can reduce the upgrades the grid needs and may shorten studies.
- Cost control. Owning generation can hedge energy prices, though it adds fuel price exposure and operating cost.
- Resilience. On-site resources can carry load during grid events, in addition to backup systems.
- Flexibility credit. A load that can switch to on-site power during grid stress may qualify for flexible service where utilities and markets offer it.
Natural gas is the most common fuel for large behind-the-meter generation today, which ties these projects to pipeline access. Our guides to on-site generation and natural gas and gas pipeline access cover the fuel side.
04The tradeoffs of behind-the-meter supply
Behind-the-meter power moves responsibilities from the utility to the project. Those responsibilities are real and they shape the site.
- Air permitting. Generators that run routinely face stricter permitting than emergency-only units. Emissions limits can cap how much can run.
- Fuel supply. A gas plant needs firm pipeline capacity, and sometimes on-site fuel backup.
- Redundancy. If the plant is the main supply, the campus needs enough units to cover outages and maintenance, which means installing more capacity than the load.
- Land and noise. Generation yards, fuel equipment and their setbacks take acreage and affect neighbors.
- Regulatory treatment. Some states limit who may sell power to a customer, and a third-party-owned plant serving one customer may raise those questions. Rules vary by state.
- Grid charges. Utilities may apply standby or backup charges to customers that self-supply but keep grid service.
05Front-of-the-meter options beyond standard grid service
Front-of-the-meter does not only mean buying from the utility under its standard tariff. Data centers also use the grid to reach resources they help bring online.
- Power purchase agreements with generators elsewhere on the grid, physical or financial depending on the market.
- Utility programs that dedicate new resources to a large customer’s load, where offered.
- Market purchases in regions with retail choice, through a competitive supplier.
- Co-location next to an existing plant with power delivered through the grid, which keeps transmission planning and charges in the picture.
These approaches keep the grid’s reliability as the backstop but leave the project dependent on the load interconnection and on transmission upgrades. Which options exist depends heavily on whether the site is in a vertically integrated utility territory or a competitive market.
06How the choice affects site selection
The decision is less about which model is better than which constraints the site can meet. A useful way to frame it is to start with the grid and treat behind-the-meter supply as a tool for gaps in capacity, timing or cost. Our decision guide on grid power vs. on-site generation walks through that choice.
- 01Establish what the grid can deliver at the site, and when.
- 02Size the gap between that and the project’s ramp.
- 03Check whether the site has gas access, air permitting room and land for a plant.
- 04Confirm how the utility and state treat self-supply, standby service and third-party ownership.
- 05Decide whether on-site generation is a bridge, a permanent share of supply or a resilience layer.
When we screen a site, we record both profiles: what it offers for grid service and what it offers for on-site generation. Sites that score well on both keep the most options open as a project’s needs change.
Common questions
What does behind-the-meter mean for a data center?
Behind-the-meter refers to generation or storage on the data center’s side of the utility meter, serving the campus directly. Examples include on-site gas turbines or engines, fuel cells, batteries and solar. It reduces the net demand the grid has to serve. Emergency backup generators are technically behind the meter, but the term usually refers to resources that run routinely.
What is front-of-the-meter power?
Front-of-the-meter power comes from resources connected to the utility or transmission system, on the grid side of the meter. A data center receives it through its utility service, whether the energy comes from the utility’s plants, market purchases or a power purchase agreement with a generator elsewhere. The grid delivers the full load, and grid reliability planning backs it up.
Can a data center run entirely behind the meter, off the grid?
It is technically possible but demanding. A fully islanded campus needs enough generation to cover its peak load plus outages and maintenance, a firm fuel supply, air permits for routine operation and its own controls. Most projects keep a grid connection as a backstop, even if on-site generation carries much of the load. State rules on self-supply and third-party sales also vary.
Is behind-the-meter generation faster than waiting for grid power?
Sometimes. Where grid upgrades would take years, on-site generation may start serving load sooner. But it has its own timeline: air permits, gas pipeline capacity, equipment lead times and local approvals. Whether it is faster depends on the site’s fuel access, emissions limits and the utility’s rules for connecting a customer that also generates.
Have a site in mind?
Get a straight answer on your land.
Send a parcel number, an address, a map pin or a target load. We’ll tell you what it can support and what it would take.
This guide is general information about data center site selection. It is not engineering, legal, tax or investment advice. Requirements vary by state, utility and county, so confirm the specifics for any site with the relevant authorities and advisors.
Related guides
More in Power & interconnection
- How Much Power Does a Data Center Need?
- Transmission Voltage and Data Centers: 69 kV to 765 kV
- Substation Proximity and Capacity for Data Center Sites
- How the Large-Load Utility Interconnection Process Works
- Power Timelines: Why “When” Matters as Much as “Where”
- ISOs, RTOs and Utility Territories in Data Center Siting
- What Is Powered Land?
- Load Ramp Schedules: How to Phase Power for a Data Center Campus
- Redundant Utility Feeds and Dual Substations for Data Centers
- Large-Load Tariffs and Electric Service Agreements, Explained
- Transmission Upgrades for Data Centers: Who Pays for Them?
- Nuclear and Small Modular Reactors: Siting Considerations for Data Centers
- Renewable Energy Procurement and How It Affects Data Center Siting
