As demand for high‑intensity electricity usage rises, driven by industries such as data centers, digital infrastructure, and large‑scale manufacturing, the question of where to locate major power loads becomes increasingly strategic. Regions that “surge” tend to combine robust electricity supply, substation and transmission capacity, fiber‑optic connectivity, favorable regulatory and permitting environments; regions that falter often lack these advantages. This article analyses the drivers behind the uneven geography of power‑intensive site development, and examines whether policymakers should consider slowing expansion to avoid overburdening the grid and putting other sectors at a disadvantage.

A useful lens through which to view power siting strategy is the study of grid interconnection queues. In the United States, active capacity seeking transmission interconnection reached nearly 2,600 GW towards the end of 2023. That is more than twice the installed generating capacity of the existing U.S. power plant fleet [1]. Of this queued capacity, more than 95% is solar, battery storage, and wind, with over 1,080 GW of solar and around 1,030 GW of storage capacity waiting for connection. But a key insight: historically, only a small fraction of queued capacity converts into operating plants. For projects entering the queue between 2000 and 2018, just 14% of capacity has come online, even as queue wait times have grown [1].

Moreover, the time from interconnection request to commercial operation has increased significantly: whereas it was under two years in the early 2000s, by 2023 the median wait was about five years [2]. These delays and backlog reflect structural bottlenecks, in permitting, transmission‑system upgrades, grid studies, and financing, that shape where power‑hungry facilities can realistically locate.

In parallel, for certain grid regions, regulatory reforms have been proposed or adopted to speed up interconnection. For example, PJM Interconnection (PJM), the largest U.S. regional grid operator, has implemented queue reforms in response to delays [3]. At the same time, in 2025, a partnership between Google and PJM began rolling out AI‑assisted tools for interconnection‑queue processing, aiming to reduce approval times [4].

These data highlight that interconnection capacity, and the speed and costs associated with connecting new loads or generation, is a foundational constraint on where energy‑intensive development can occur.

Certain regions become “hot spots” for power‑intensive development. Several factors tend to coincide in these surging locales:

Regions with ample generation capacity, available substations, and transmission infrastructure can absorb added load more easily. In areas where the grid is already near capacity, adding a large new load (e.g., a data center or manufacturing plant) may require costly upgrades, often a disincentive.For example, pundits argue that in PJM the backlog of generation interconnection requests and the resulting long wait times, along with high costs for network upgrades, have stifled supply entry, even as demand surges [3]. Conversely, some regions like parts of the U.S. Midwest or other less‑congested grid zones offer “space” on the grid, making them attractive for new power‑intensive facilities. For instance, the Midcontinent Independent System Operator (MISO) has recently reformed its interconnection queue rules to improve efficiency [5].

Speed and predictability in permitting, both for generation (if being built) and for new loads, matter. When permitting is slow, uncertain, or unpredictable, developers face higher risk. Regions with streamlined regulatory frameworks or “fast‑track” capacities naturally attract more investment. PJM’s recent adoption of reforms, including a cluster‑based interconnection study process, aims to address some of these issues, although challenges remain [3]. Regulatory stance on large loads (e.g., data centers) is also under discussion: as of November 2025, PJM stakeholders failed to agree on new rules for interconnecting large loads, leaving uncertainty in a critical domain [2].

Power alone is not enough for many modern facilities, especially data centers. They also require robust connectivity: fiber‑optic backbone, access to internet exchanges, redundancy, low latency. Regions with strong telecom infrastructure, and proximity to major fiber networks or subsea cables, gain a competitive advantage. An example is Northern Virginia, often cited as a global leader in data‑center deployment. Its connectivity (fiber, network density, proximity to major metro hubs) combined with historic grid capacity has made it a magnet for data centers [6].

Regions that offer favorable economic terms, tax breaks, industrial power rates, subsidies, streamlined development, naturally draw site‑hungry institutions. The presence of such incentives can outweigh other disadvantages or make up for higher upfront costs. Recent analyses confirm that states and localities actively compete for data‑center and large‑load manufacturing investment by offering such incentives, while also leveraging job creation, infrastructure upgrades, and long‑term revenue streams [7].

In some cases, facilities plan to generate power “behind the meter” (on‑site generation) or build a mix of generation + storage to avoid long grid interconnection delays or transmission constraints. Analysts argue this trend may become more common, especially for very large loads [8]. Indeed, as grid interconnection backlogs persist, waiting for grid-based supply may be increasingly impractical for developers needing reliable, low‑latency power.

While there is no publicly available clickable “heat‑map” detailing exactly which states surge for all kinds of power‑intensive developments, evidence from data‑center distribution and energy‑load forecasts points to a few consistent patterns.

• Northern Virginia remains a global hub for data centers, due to a combination of grid capacity, fiber connectivity, and favorable business conditions [6].

• In parts of the U.S. Midwest, under MISO, growing load forecasts for 2024 - 2035 include demands from “Data Centers (149 - 241 TWh)” plus building electrification, manufacturing, and EV adoption [9].

• Regions less burdened by previous build‑out (transmission and generation) and with relatively underutilized infrastructure are increasingly considered by developers seeking a balance between cost, reliability, and growth potential [10].

These examples underscore that the interplay among grid infrastructure, regulatory framework, connectivity, local incentives, and future‑proofing is critical in determining where new power‑hungry facilities are built.

Given the concentration of power‑intensive facilities in certain hot zones, and the pressures on grid capacity and interconnection systems, a valid policy question emerges: should regulators and policymakers slow expansion (or at least distribute it more evenly) to avoid starving other manufacturing or critical sectors of power?

A few arguments support this view:

1. Grid Reliability & Diversity: Over concentration of large loads in certain regions may stress local transmission infrastructure, reduce reliability, and increase blackout risk. By slowing or redirecting expansion, policymakers could help preserve grid stability, especially in regions with older infrastructure or limited spare capacity.

2. Preventing Uneven Economic Development: If power‑hungry industries cluster too much in a few states, other regions may be left behind, even if they have potential. Policies encouraging geographic dispersion could support more balanced economic growth.

3. Fair Access to Power for All Sectors: If data centers or AI campuses dominate power consumption, other energy‑dependent industries (manufacturing, critical infrastructure) might struggle for access. Slowing or pacing expansion could ensure a fairer allocation of power resources.

4. Encouraging Efficient Use of Grid Infrastructure: Rapid accumulation of queued projects, especially speculative ones, may clog the interconnection process and delay needed upgrades. Slower, more deliberate development could lead to more prudent upgrades, efficient transmission planning, and better long-term grid resilience.

However, deliberate slowing raises trade‑offs:

• It could discourage investment in emerging industries (AI, cloud, digital infrastructure), harming competitiveness.

• It may push developers toward self-generation (on‑site generation, storage), which can have environmental, land‑use, and regulatory implications.

• It could create uncertainty and reduce the incentive for innovation in sectors that rely on high-power consumption.

While the idea of slowing growth or steering load to underserved regions might seem appealing, in practice it faces several constraints:

As discussed, the backlog in interconnection requests is substantial, and delays have lengthened materially [2]. For many developers, seeking faster connection means selecting regions with less congestion, which reinforces clustering rather than dispersal.

Moreover, even with reforms (e.g., by PJM and MISO), the economics of bringing new generation online remain challenging, with high network upgrade costs and long lead times [3].

High‑tech companies, hyperscalers, and cloud providers are unlikely to locate in areas that fail to offer both affordable power and high connectivity. The benefits of tax breaks, fiber connectivity, and regulatory friendliness often outweigh higher initial infrastructure costs [7].

Also, when demand for digital services is global, companies prioritize latency, redundancy, and network reach, which tends to cluster around existing hubs rather than disperse.

Building generation, transmission, substations, and grid upgrades takes time, capital, and regulatory alignment. As per a recently published whitepaper, projects currently in queue for PJM have little chance of commissioning before 2030 [11]. In effect, the grid evolves slowly compared to the pace of demand growth; market signals tend to concentrate new demand where capacity appears least constrained, reinforcing regional surges.

As grid-based supply becomes less predictable, many developers may choose on‑site generation plus storage or hybrid energy supply (e.g., co‑location of renewable + battery + load), a choice that shifts load away from shared grid constraints, but can complicate broader planning and further weaken shared infrastructure incentives [8]. In effect, decentralized load growth makes grid planning harder, potentially undermining attempts to manage siting at scale.

If policymakers do consider influencing power‑intensive siting, to avoid over concentration or to favor broader economic development, what strategies could they use? Based on current data and industry dynamics, several approaches emerge:

1. Grid‑wide Long‑Term Planning: Instead of first‑come, first‑served interconnection queues, regulators could adopt long‑term grid expansion plans aligned with projections for load growth (industry, data centers, electrification). This helps avoid ad‑hoc wait times and regional bottlenecks.

2. Zoning and Siting Guidance: Encourage siting of large loads (data centers, manufacturing) in regions with underutilized grid capacity or where planned grid upgrades are feasible. Offer incentives (tax, permits) for siting in “green‑field” or underserved areas.

3. Demand‑side Participation & Load Management: Promote demand‑response, scheduling of load, partial on‑site generation or storage, or “time‑shifting” of high-load operations (e.g., non‑peak computing tasks) to reduce peak grid stress. Encourage hybrid approaches that align load with variable renewable generation or cheaper off-peak power.

4. Transparent, Efficient Interconnection Processes: Continue reforming queue management to reduce delays: streamlined studies, cluster‑based review, AI or automation for planning, accountability for speculative projects, and clear cost‑allocation rules for network upgrades.

5. Cross‑Sector Coordination: Bring together stakeholders from data centers, manufacturing, utilities, regulators, communities, to jointly plan grid upgrades, siting, and long‑term capacity. This holistic approach can help balance competing demands and avoid over concentration.

The evidence from recent interconnection‑queue data, grid‑operator reports, and industry analysis shows that power‑siting is far from random. Regions that surge, such as parts of the U.S. Midwest or data‑center hubs like Northern Virginia, do so because they combine critical assets: grid/substation capacity, transmission infrastructure, regulatory speed, favorable economics, and connectivity. Meanwhile, systemic bottlenecks in queue processing, permitting, and transmission upgrades limit where new power-intensive loads can go, shaping geography by constraint as much as choice.

At present, though, market forces and infrastructure path‑dependency make clustering likely to continue. Without coordinated, long‑term planning, the risk is that power-intensive growth will overload existing hubs, leading to reliability problems, delayed interconnections, high upgrade costs, and further inequality in regional industrial opportunity. Ultimately, a robust power‑siting strategy must balance economic growth, grid reliability, fairness across sectors and regions, and long-term sustainability. As load demands rise and technologies evolve, the decisions about where to site power‑hungry facilities will shape not only local economies but the stability and performance of power systems for decades.

1. Queued Up: Characteristics of Power Plants Seeking Transmission Interconnection as of the End of 2023 | Berkeley Lab (2024) https://emp.lbl.gov/sites/default/files/2024-04/Queued%20Up%202024%20Edition_R2.pdf

1. Grid interconnection queues jumped 27%, to 2.6 TW, in 2023, led by solar, storage | Utility Dive (2024) https://www.utilitydive.com/news/grid-interconnection-queue-berkeley-lab-lbnl/712926/

2. PJM’s Speed to Power Problem and How to Fix It | RMI (2025) https://rmi.org/pjms-speed-to-power-problem-and-how-to-fix-it/

3. Google brings AI to grid teams, slashing US connection times | Reuters (2025) https://www.reuters.com/business/energy/google-brings-ai-grid-teams-slashing-us-connection-times-2025-05-20/

4. MISO Interconnection Queue Reform and Progress Update | MISO / U.S. Department of Energy (2024) https://www.energy.gov/sites/default/files/2024-08/0724_MISO_IQ_reform_and_2023_cycle_policy_note_draft_FINAL.pdf

5. Top 10 U.S. Data Center Markets and Why They Are Hot | CoreSite (2023) https://www.coresite.com/blog/top-10-u-s-data-center-markets-and-why-they-are-hot

6. The Data Center Balance: How US States Can Navigate the Opportunities and Challenges | McKinsey & Company (2025) https://www.mckinsey.com/industries/public-sector/our-insights/the-data-center-balance-how-us-states-can-navigate-the-opportunities-and-challenges

As demand for high‑intensity electricity usage rises, driven by industries such as data centers, digital infrastructure, and large‑scale manufacturing, the question of where to locate major power loads becomes increasingly strategic. Regions that “surge” tend to combine robust electricity supply, substation and transmission capacity, fiber‑optic connectivity, favorable regulatory and permitting environments; regions that falter often lack these advantages. This article analyses the drivers behind the uneven geography of power‑intensive site development, and examines whether policymakers should consider slowing expansion to avoid overburdening the grid and putting other sectors at a disadvantage.

A useful lens through which to view power siting strategy is the study of grid interconnection queues. In the United States, active capacity seeking transmission interconnection reached nearly 2,600 GW towards the end of 2023. That is more than twice the installed generating capacity of the existing U.S. power plant fleet [1]. Of this queued capacity, more than 95% is solar, battery storage, and wind, with over 1,080 GW of solar and around 1,030 GW of storage capacity waiting for connection. But a key insight: historically, only a small fraction of queued capacity converts into operating plants. For projects entering the queue between 2000 and 2018, just 14% of capacity has come online, even as queue wait times have grown [1].

Moreover, the time from interconnection request to commercial operation has increased significantly: whereas it was under two years in the early 2000s, by 2023 the median wait was about five years [2]. These delays and backlog reflect structural bottlenecks, in permitting, transmission‑system upgrades, grid studies, and financing, that shape where power‑hungry facilities can realistically locate.

In parallel, for certain grid regions, regulatory reforms have been proposed or adopted to speed up interconnection. For example, PJM Interconnection (PJM), the largest U.S. regional grid operator, has implemented queue reforms in response to delays [3]. At the same time, in 2025, a partnership between Google and PJM began rolling out AI‑assisted tools for interconnection‑queue processing, aiming to reduce approval times [4].

These data highlight that interconnection capacity, and the speed and costs associated with connecting new loads or generation, is a foundational constraint on where energy‑intensive development can occur.

Certain regions become “hot spots” for power‑intensive development. Several factors tend to coincide in these surging locales:

Regions with ample generation capacity, available substations, and transmission infrastructure can absorb added load more easily. In areas where the grid is already near capacity, adding a large new load (e.g., a data center or manufacturing plant) may require costly upgrades, often a disincentive.For example, pundits argue that in PJM the backlog of generation interconnection requests and the resulting long wait times, along with high costs for network upgrades, have stifled supply entry, even as demand surges [3]. Conversely, some regions like parts of the U.S. Midwest or other less‑congested grid zones offer “space” on the grid, making them attractive for new power‑intensive facilities. For instance, the Midcontinent Independent System Operator (MISO) has recently reformed its interconnection queue rules to improve efficiency [5].

Speed and predictability in permitting, both for generation (if being built) and for new loads, matter. When permitting is slow, uncertain, or unpredictable, developers face higher risk. Regions with streamlined regulatory frameworks or “fast‑track” capacities naturally attract more investment. PJM’s recent adoption of reforms, including a cluster‑based interconnection study process, aims to address some of these issues, although challenges remain [3]. Regulatory stance on large loads (e.g., data centers) is also under discussion: as of November 2025, PJM stakeholders failed to agree on new rules for interconnecting large loads, leaving uncertainty in a critical domain [2].

Power alone is not enough for many modern facilities, especially data centers. They also require robust connectivity: fiber‑optic backbone, access to internet exchanges, redundancy, low latency. Regions with strong telecom infrastructure, and proximity to major fiber networks or subsea cables, gain a competitive advantage. An example is Northern Virginia, often cited as a global leader in data‑center deployment. Its connectivity (fiber, network density, proximity to major metro hubs) combined with historic grid capacity has made it a magnet for data centers [6].

Regions that offer favorable economic terms, tax breaks, industrial power rates, subsidies, streamlined development, naturally draw site‑hungry institutions. The presence of such incentives can outweigh other disadvantages or make up for higher upfront costs. Recent analyses confirm that states and localities actively compete for data‑center and large‑load manufacturing investment by offering such incentives, while also leveraging job creation, infrastructure upgrades, and long‑term revenue streams [7].

In some cases, facilities plan to generate power “behind the meter” (on‑site generation) or build a mix of generation + storage to avoid long grid interconnection delays or transmission constraints. Analysts argue this trend may become more common, especially for very large loads [8]. Indeed, as grid interconnection backlogs persist, waiting for grid-based supply may be increasingly impractical for developers needing reliable, low‑latency power.

While there is no publicly available clickable “heat‑map” detailing exactly which states surge for all kinds of power‑intensive developments, evidence from data‑center distribution and energy‑load forecasts points to a few consistent patterns.

• Northern Virginia remains a global hub for data centers, due to a combination of grid capacity, fiber connectivity, and favorable business conditions [6].

• In parts of the U.S. Midwest, under MISO, growing load forecasts for 2024 - 2035 include demands from “Data Centers (149 - 241 TWh)” plus building electrification, manufacturing, and EV adoption [9].

• Regions less burdened by previous build‑out (transmission and generation) and with relatively underutilized infrastructure are increasingly considered by developers seeking a balance between cost, reliability, and growth potential [10].

These examples underscore that the interplay among grid infrastructure, regulatory framework, connectivity, local incentives, and future‑proofing is critical in determining where new power‑hungry facilities are built.

Given the concentration of power‑intensive facilities in certain hot zones, and the pressures on grid capacity and interconnection systems, a valid policy question emerges: should regulators and policymakers slow expansion (or at least distribute it more evenly) to avoid starving other manufacturing or critical sectors of power?

A few arguments support this view:

1. Grid Reliability & Diversity: Over concentration of large loads in certain regions may stress local transmission infrastructure, reduce reliability, and increase blackout risk. By slowing or redirecting expansion, policymakers could help preserve grid stability, especially in regions with older infrastructure or limited spare capacity.

2. Preventing Uneven Economic Development: If power‑hungry industries cluster too much in a few states, other regions may be left behind, even if they have potential. Policies encouraging geographic dispersion could support more balanced economic growth.

3. Fair Access to Power for All Sectors: If data centers or AI campuses dominate power consumption, other energy‑dependent industries (manufacturing, critical infrastructure) might struggle for access. Slowing or pacing expansion could ensure a fairer allocation of power resources.

4. Encouraging Efficient Use of Grid Infrastructure: Rapid accumulation of queued projects, especially speculative ones, may clog the interconnection process and delay needed upgrades. Slower, more deliberate development could lead to more prudent upgrades, efficient transmission planning, and better long-term grid resilience.

However, deliberate slowing raises trade‑offs:

• It could discourage investment in emerging industries (AI, cloud, digital infrastructure), harming competitiveness.

• It may push developers toward self-generation (on‑site generation, storage), which can have environmental, land‑use, and regulatory implications.

• It could create uncertainty and reduce the incentive for innovation in sectors that rely on high-power consumption.

While the idea of slowing growth or steering load to underserved regions might seem appealing, in practice it faces several constraints:

As discussed, the backlog in interconnection requests is substantial, and delays have lengthened materially [2]. For many developers, seeking faster connection means selecting regions with less congestion, which reinforces clustering rather than dispersal.

Moreover, even with reforms (e.g., by PJM and MISO), the economics of bringing new generation online remain challenging, with high network upgrade costs and long lead times [3].

High‑tech companies, hyperscalers, and cloud providers are unlikely to locate in areas that fail to offer both affordable power and high connectivity. The benefits of tax breaks, fiber connectivity, and regulatory friendliness often outweigh higher initial infrastructure costs [7].

Also, when demand for digital services is global, companies prioritize latency, redundancy, and network reach, which tends to cluster around existing hubs rather than disperse.

Building generation, transmission, substations, and grid upgrades takes time, capital, and regulatory alignment. As per a recently published whitepaper, projects currently in queue for PJM have little chance of commissioning before 2030 [11]. In effect, the grid evolves slowly compared to the pace of demand growth; market signals tend to concentrate new demand where capacity appears least constrained, reinforcing regional surges.

As grid-based supply becomes less predictable, many developers may choose on‑site generation plus storage or hybrid energy supply (e.g., co‑location of renewable + battery + load), a choice that shifts load away from shared grid constraints, but can complicate broader planning and further weaken shared infrastructure incentives [8]. In effect, decentralized load growth makes grid planning harder, potentially undermining attempts to manage siting at scale.

If policymakers do consider influencing power‑intensive siting, to avoid over concentration or to favor broader economic development, what strategies could they use? Based on current data and industry dynamics, several approaches emerge:

1. Grid‑wide Long‑Term Planning: Instead of first‑come, first‑served interconnection queues, regulators could adopt long‑term grid expansion plans aligned with projections for load growth (industry, data centers, electrification). This helps avoid ad‑hoc wait times and regional bottlenecks.

2. Zoning and Siting Guidance: Encourage siting of large loads (data centers, manufacturing) in regions with underutilized grid capacity or where planned grid upgrades are feasible. Offer incentives (tax, permits) for siting in “green‑field” or underserved areas.

3. Demand‑side Participation & Load Management: Promote demand‑response, scheduling of load, partial on‑site generation or storage, or “time‑shifting” of high-load operations (e.g., non‑peak computing tasks) to reduce peak grid stress. Encourage hybrid approaches that align load with variable renewable generation or cheaper off-peak power.

4. Transparent, Efficient Interconnection Processes: Continue reforming queue management to reduce delays: streamlined studies, cluster‑based review, AI or automation for planning, accountability for speculative projects, and clear cost‑allocation rules for network upgrades.

5. Cross‑Sector Coordination: Bring together stakeholders from data centers, manufacturing, utilities, regulators, communities, to jointly plan grid upgrades, siting, and long‑term capacity. This holistic approach can help balance competing demands and avoid over concentration.

The evidence from recent interconnection‑queue data, grid‑operator reports, and industry analysis shows that power‑siting is far from random. Regions that surge, such as parts of the U.S. Midwest or data‑center hubs like Northern Virginia, do so because they combine critical assets: grid/substation capacity, transmission infrastructure, regulatory speed, favorable economics, and connectivity. Meanwhile, systemic bottlenecks in queue processing, permitting, and transmission upgrades limit where new power-intensive loads can go, shaping geography by constraint as much as choice.

At present, though, market forces and infrastructure path‑dependency make clustering likely to continue. Without coordinated, long‑term planning, the risk is that power-intensive growth will overload existing hubs, leading to reliability problems, delayed interconnections, high upgrade costs, and further inequality in regional industrial opportunity. Ultimately, a robust power‑siting strategy must balance economic growth, grid reliability, fairness across sectors and regions, and long-term sustainability. As load demands rise and technologies evolve, the decisions about where to site power‑hungry facilities will shape not only local economies but the stability and performance of power systems for decades.

1. Queued Up: Characteristics of Power Plants Seeking Transmission Interconnection as of the End of 2023 | Berkeley Lab (2024) https://emp.lbl.gov/sites/default/files/2024-04/Queued%20Up%202024%20Edition_R2.pdf

1. Grid interconnection queues jumped 27%, to 2.6 TW, in 2023, led by solar, storage | Utility Dive (2024) https://www.utilitydive.com/news/grid-interconnection-queue-berkeley-lab-lbnl/712926/

2. PJM’s Speed to Power Problem and How to Fix It | RMI (2025) https://rmi.org/pjms-speed-to-power-problem-and-how-to-fix-it/

3. Google brings AI to grid teams, slashing US connection times | Reuters (2025) https://www.reuters.com/business/energy/google-brings-ai-grid-teams-slashing-us-connection-times-2025-05-20/

4. MISO Interconnection Queue Reform and Progress Update | MISO / U.S. Department of Energy (2024) https://www.energy.gov/sites/default/files/2024-08/0724_MISO_IQ_reform_and_2023_cycle_policy_note_draft_FINAL.pdf

5. Top 10 U.S. Data Center Markets and Why They Are Hot | CoreSite (2023) https://www.coresite.com/blog/top-10-u-s-data-center-markets-and-why-they-are-hot

6. The Data Center Balance: How US States Can Navigate the Opportunities and Challenges | McKinsey & Company (2025) https://www.mckinsey.com/industries/public-sector/our-insights/the-data-center-balance-how-us-states-can-navigate-the-opportunities-and-challenges