Summary
- The dispute concerns the Peru Shelf and South Valley developments, which together are planned for more than 2GW.
- NV Energy alleges that Tract has not accepted the full cost of infrastructure required to serve the sites.
- Tract denies seeking a ratepayer subsidy and says it has committed substantial capital to Nevada’s electricity network.
NV Energy has filed a lawsuit against data centre park developer Tract over who should bear the infrastructure costs required to supply two proposed Nevada developments with more than 2GW of combined capacity.
The dispute centres on Tract’s Peru Shelf and South Valley projects. Peru Shelf is planned across 686 acres with an expected capacity of 810MW, while South Valley covers 1,534 acres and is intended to reach 1.215GW.
Tract has previously indicated energisation targets of September 2027 for South Valley and February 2028 for Peru Shelf. It is not yet clear whether the litigation will affect those dates.
NV Energy alleges that Tract has failed to accept costs associated with reserving and delivering the power required by the developments, creating a risk that expenses could be transferred to other electricity customers.
Tract rejects that characterisation. The developer says it has already spent more than $50m and committed nearly $1bn towards infrastructure in Nevada, and that it has not sought preferential treatment or a subsidy from other ratepayers.
The claims remain contested, and the court has not determined their merits. The dispute nevertheless exposes the contractual pressure created when a proposed data centre load is comparable with the demand of a large industrial region.
Connection agreements carry project risk
Data centre projects are often described by their planned megawatts, but those megawatts require generation, transmission, substations, distribution infrastructure, and operational reserves. The commercial allocation of those costs can determine whether a site remains viable.
Utilities must protect existing customers from investments made for speculative projects that fail to materialise. Developers, meanwhile, need confidence that payments and commitments will result in capacity being delivered on an agreed timetable.
The problem becomes more acute where a developer prepares a large park for multiple future occupiers rather than building a single facility for a named customer. The park may require the utility to plan for extensive load before every tenant has signed a lease or fixed its final requirements.
Security deposits, minimum-demand charges, cancellation payments, construction contributions, and phased energisation agreements are intended to distribute that risk. Their precise terms become commercially sensitive when grid reinforcement is expensive and lead times extend across several years.
The combined 2.025GW planned at Peru Shelf and South Valley places those questions at exceptional scale. Delivering that demand could require new generation and network infrastructure well beyond the boundaries of the two sites.
The Nevada dispute has parallels in European markets, where regulators and governments are considering how large-load customers should enter grid queues and demonstrate that projects are sufficiently mature to justify network investment.
Connection queues can be distorted when developers reserve more capacity than they eventually use or hold positions for projects that have not secured land, financing, planning consent, or customers. Requiring stronger financial commitments can remove speculative demand, but charges set too early or too high can prevent credible projects from progressing.
The allocation also affects public acceptance. Ratepayers are increasingly alert to the possibility that data centre growth may raise electricity bills or divert network investment from homes and other businesses.
Developers counter that large campuses can finance new substations, transmission assets, and generation that strengthen the wider system. Whether those benefits materialise depends on how infrastructure is funded, owned, and used after construction.
Tract’s Nevada projects are designed as master-planned data centre parks, with the developer preparing powered and entitled sites for other companies. That model places grid agreements at the centre of the commercial proposition: without credible power delivery, the land cannot be sold or leased as ready infrastructure.
The lawsuit turns a private contractual disagreement into a wider test of large-load cost allocation. The eventual outcome may influence not only Tract’s timetable but the conditions under which other multi-gigawatt developments seek service in Nevada.
