Data centers are driving a new gas power buildout—while virtual power and green roofs stall, raising grid and climate stakes

Across the country, dozens of gas-burning power plants are reportedly being built specifically to supply electricity to data centers, with emissions projected to rival the annual warming impact of roughly half of all passenger vehicles in the country. The claim, flagged as an urgent climate risk, centers on the idea that the power demand from compute growth is being met with fossil generation rather than cleaner capacity. Separately, in the United States, thousands of households that installed batteries under a federal scheme are not enrolling as “virtual power plants,” meaning stored energy is not being aggregated and sold back to the grid. In the Netherlands, early nationwide figures suggest that almost all rooftops remain gray rather than green, despite benefits for cooling, moisture retention, and urban nature—largely because installation costs are high and collective ownership structures complicate adoption. Geopolitically, the cluster points to a broader energy-security and industrial-policy tension: governments and markets are trying to accelerate digital infrastructure while managing emissions, grid reliability, and public acceptance. If data-center load is met through new gas plants, it locks in higher-carbon generation and can intensify political pressure for stricter climate rules, potentially colliding with near-term power reliability goals. The virtual power plant under-enrollment in the US highlights a governance and trust problem—households may fear complexity, unclear compensation, or performance risk, reducing flexibility that could otherwise defer expensive grid upgrades. Meanwhile, the Dutch green-roof adoption gap illustrates how urban policy design and cost-sharing mechanisms can slow distributed climate adaptation, increasing cooling demand and indirectly stressing electricity systems during heat waves. Market and economic implications are likely to concentrate in power generation and grid services. A gas-led buildout for data centers can lift demand expectations for natural gas, power-plant construction, and ancillary services, while increasing exposure to carbon pricing regimes and future regulatory tightening; the direction is toward higher emissions and potentially higher long-run compliance costs. In the US, low virtual power plant participation can reduce the monetization of residential storage and weaken the revenue stack for battery aggregators, potentially slowing investment in distributed energy resources even when hardware deployment is already underway. In Europe, the slow uptake of green roofs can keep urban cooling loads higher, which tends to support peak power demand and can raise volatility in electricity markets during summer extremes; the magnitude is difficult to quantify from the articles, but the direction is toward greater peak stress rather than demand smoothing. What to watch next is whether policymakers respond with demand-side governance (for example, a data-center moratorium or stricter permitting) and whether grid operators can unlock flexibility from residential storage. Key indicators include enrollment rates into virtual power plant programs, changes in compensation terms, and any regulatory guidance that addresses household mistrust. For the Netherlands, monitoring subsidy adjustments, mandatory green-roof requirements, and new financing models for collective housing will show whether adoption barriers are being reduced. Escalation would occur if gas plant construction accelerates while distributed flexibility remains underutilized, tightening the link between digital growth and fossil emissions; de-escalation would be signaled by faster virtual power participation, clearer market rules for batteries, and policy packages that make cooling and stormwater benefits from green infrastructure financially feasible.

Geopolitical Implications

• 01

  Digital infrastructure growth is colliding with decarbonization and grid-reliability strategies, potentially shifting energy policy toward stricter demand governance.

• 02

  If fossil generation expands while distributed flexibility underperforms, governments may face higher political costs from emissions and future regulatory tightening.

• 03

  Trust-based barriers in energy markets (virtual power enrollment) can undermine resilience and increase the likelihood of more centralized, capital-intensive capacity additions.

• 04

  Urban adaptation lag (e.g., green roofs) can amplify heat-driven electricity demand, strengthening the case for integrated climate-energy planning.

Key Signals

• —Virtual power plant participation rates and changes to household incentives/contract terms
• —Permitting timelines and capacity approvals for gas plants tied to data-center load
• —Carbon policy signals that could raise the cost of new fossil generation
• —Green-roof subsidy reforms, mandatory requirements, and financing mechanisms for collective housing