Who Benefits When Grid Connections Take Years?

The new bottleneck is the power delivery stack

A large data center cannot simply install a generator and bypass the electricity system. Onsite generation creates its own miniature power network.

At the front end are gas turbines from companies such as GE Vernova, reciprocating generator sets from Caterpillar and Cummins, and solid oxide fuel cells from Bloom Energy. Storage introduces another branch through battery cells, inverters, containers, thermal systems, and controls, with companies such as Tesla and Fluence participating in different parts of that stack.

But generation is only the first layer. Electricity still has to be transformed, protected, switched, synchronized, and distributed around the site. That creates demand for medium-voltage switchgear, transformers, breakers, busways, power conversion equipment, controls, and microgrid management systems. Eaton and Schneider Electric therefore sit in a different position from the generation suppliers: they can participate regardless of whether a project chooses turbines, engines, fuel cells, batteries, or some combination.

This distinction matters for investors because alternative power technologies compete with one another, while much of the electrical balance-of-plant is complementary to all of them.

A data center replacing a delayed grid connection with gas turbines may benefit GE Vernova. A project choosing fuel cells instead may benefit Bloom Energy. A solar and storage configuration shifts spending toward batteries and inverters. Yet nearly every architecture still needs switchgear, protection systems, distribution equipment, and controls.

The broadest exposure to the "grid takes too long" problem may therefore sit one layer downstream from generation.

Gas turbines benefit, but their own supply chain is constrained

Gas generation appears like the obvious solution because it can provide dispatchable power at the scale required by hyperscale computing. Microsoft, Chevron, and Engine No. 1, for example, agreed in 2026 to work exclusively on power generation and supply arrangements tied to AI infrastructure. Chevron and Engine No. 1 had already planned adjacent gas plants using GE Vernova turbines.[3]

The problem is that gas turbines are becoming a bottleneck themselves.

Reuters reported that average gas turbine lead times had reached roughly five years in 2025 as data center demand collided with limited manufacturing capacity.[2] That changes the economics of the alternative-power market. A turbine may bypass a transmission constraint, but it does not solve the customer's problem if the equipment arrives near the date when the grid connection would have been available anyway.

GE Vernova can still benefit because scarcity strengthens demand for manufacturing slots, services, generators, and related grid equipment. But turbine scarcity also creates room for technologies whose primary advantage is deployment speed rather than lowest long-run generation cost.

Bloom Energy is the clearest public example. Its fuel cells generate electricity onsite in modular installations without requiring a conventional central power plant. For a customer whose alternative is leaving billions of dollars of computing infrastructure idle, a higher electricity cost can be acceptable if it advances the operating date materially.

Caterpillar and Cummins occupy another version of the same trade. Large natural gas reciprocating engines historically served industrial generation, backup power, and microgrids. The grid constraint gives these products a new economic function: they can become bridge generation for large loads that eventually expect a utility connection.

That makes deployment time itself part of equipment value.

Storage and switchgear may be the more durable beneficiaries

Battery storage is sometimes treated as a competitor to conventional generation, but grid-constrained sites increasingly make the distinction less useful. Storage can be paired with solar, utility power, engines, or turbines to reduce peak capacity requirements and stabilize a microgrid.

Reuters reported that long gas turbine waits are already encouraging developers to use solar and storage configurations that can be deployed faster.[2] For investors, the important point is not that batteries replace gas universally. It is that connection delays increase the number of projects willing to assemble hybrid systems instead of waiting for one ideal source of power.

That broadens the addressable equipment stack.

The same logic strengthens the position of Eaton and Schneider Electric. More distributed generation means more points at which electricity must be controlled and isolated. A campus containing grid power, batteries, onsite generation, and backup systems can require a more complicated electrical architecture than a conventional grid-connected facility.

Switchgear therefore benefits twice. New grid infrastructure requires it, and attempts to bypass delayed grid infrastructure also require it.

This is a useful supply-chain asymmetry. Generation vendors compete for the same megawatt. Electrical distribution suppliers can sell into several competing generation architectures.

Microgrid controls add another layer. Once a facility has several power sources, software and control hardware must decide when to operate generators, charge batteries, import grid electricity, shed loads, or island the site. Schneider Electric, Eaton, Siemens, and other electrical automation suppliers are positioned around this integration problem rather than around a single fuel.

The second-order trade is infrastructure duplication

The deeper investment implication is that slow grid connections can increase total equipment intensity.

If transmission and utility capacity arrived exactly when customers needed it, some data centers could rely primarily on grid electricity plus conventional backup systems. A five-year connection problem can instead produce two infrastructure builds: an onsite system constructed to get the facility operating, followed later by the permanent grid connection.

That can mean turbines or engines, storage, switchgear, transformers, controls, and fuel infrastructure are installed before the utility completes another set of substations, lines, transformers, and switchgear.

Not every temporary system will remain temporary. Once a company has paid for onsite generation, the equipment can provide resilience, peak shaving, backup capacity, or participation in electricity markets after the grid connection arrives.

This is why the grid-delay theme extends beyond utilities and power producers. The economic value created by scarcity can migrate into the manufacturers of the physical equipment required to work around that scarcity.

Conclusion

When grid connections take years, the immediate beneficiaries are the companies selling time.

GE Vernova gains exposure through large-scale gas turbines and electrical equipment, Bloom Energy through rapidly deployable onsite fuel cells, Caterpillar and Cummins through reciprocating generation, and Tesla and Fluence through storage. Eaton and Schneider Electric may occupy an even broader position because switchgear, distribution equipment, and microgrid controls are required across several competing power architectures.

The key bottleneck is also moving. Grid delays first created demand for alternative generation. Turbine shortages then made the generation equipment itself scarce. The next layer of investment analysis should therefore focus on which suppliers can actually deliver complete, energizable systems fastest.

For investors, "who produces electricity?" is becoming a less useful question than "who provides the equipment that lets a customer stop waiting for electricity?"

Scope and limitations

This article maps public companies to functional positions in the alternative-power and electrical balance-of-plant stack using the supplied draft, Berkeley Lab research, and Reuters reporting. It does not establish that every named company is a supplier to every data center project, nor does it assign project-specific revenue, market share, or customer concentration.

The roughly five-year gas turbine lead-time figure is a reported market observation and can vary by manufacturer, model, project specification, and order timing. It should not be treated as a universal delivery schedule.

The Microsoft, Chevron, and Engine No. 1 arrangement was described by Reuters as an exclusivity agreement for power generation and supply, with no definitive commercial agreement finalized at the time of the report. The cited 2,500 MW West Texas project was reported as a proposed project linked to that broader arrangement.

The infrastructure-duplication framework is conceptual. Some bridge-power assets can remain useful after grid service arrives, while other projects may choose different development paths entirely.

No Altsets relationship percentages are used because the supplied article does not provide quantified company relationships that would improve the analysis without introducing unrelated data.

For evidence limits and relationship methodology, see the Altsets methodology.

Sources

1. Lawrence Berkeley National Laboratory, "Speed to Power: Solutions for Accelerating Large Load Connections," June 2026. https://emp.lbl.gov/publications/speed-power-solutions-accelerating

2. Reuters, "INVESTOR VIEW: 'Time to power' is top priority for US capital," June 30, 2026. https://www.reuters.com/business/energy/investor-view-time-power-is-top-priority-us-capital--reeii-2026-06-30/

3. Reuters, "Microsoft, Chevron and Engine No. 1 sign exclusive deal for power supply," March 31, 2026. https://www.reuters.com/business/energy/microsoft-chevron-engine-no-1-sign-exclusive-deal-power-supply-2026-03-31/

How to Cite This

For research inquiries or data access: press@altsets.com

Go Deeper