Chapter 1. Large Load Electric Demands

Introduction

This book was created to support the developers of large industrial loads and the developers offering electric power generation solutions in navigating a critical and emerging challenge. The next 25 years are predicted to have tremendous growth for electrical demand in the United States, a significant change from the last 15 years which experienced almost no growth (Figure 1.1). Most of the new growth comes from data centers for Artificial Intelligence (AI) and crypto mining facilities. Electrification of refineries and new manufacturing also has significant additions. Conversion of transportation to electric vehicles (EVs) makes a smaller impact assuming a gradual transition. This rapid load growth presents particular challenges for generation to support it. Particularly in regard to Large Loads with significant electricity demand coming online at once from a single interconnection. This text presents the opportunity, the challenges, and potential solutions for generating power to support new large loads, particularly those in the hundreds of MW to GW range.

The rapid growth of large electric loads presents a complex set of challenges for power providers and planners. These facilities demand massive amounts of electricity on accelerated timelines, often outpacing traditional grid expansion schedules. Ensuring reliability while integrating such loads stresses existing infrastructure and requires significant operability enhancements. At the same time, stakeholders must balance sustainability goals, such as reducing carbon emissions, with the cost of new generation and transmission investments. Community acceptance can also be a hurdle, as local opposition to new substations or transmission lines may delay or derail projects. Additionally, navigating policy and regulatory frameworks—often not designed for the speed and scale of modern load growth—adds further complexity. Together, these factors create a multifaceted challenge that demands innovative, collaborative, and forward-looking solutions.

To meet the growing electricity demands of large new loads like data centers, a diverse mix of innovative and flexible power solutions is being deployed. To address schedule constraints, many developers are turning to on-site generation and microgrids, which can be deployed faster than traditional grid upgrades. These systems often incorporate renewable energy sources such as solar and wind, supported by energy storage to manage intermittency and enhance reliability. Natural gas and diesel gensets provide dispatchable backup power, while emerging technologies like small modular reactors (SMRs) and geothermal energy offer promising long-term, low-carbon baseload options. This hybrid approach not only improves reliability and operability but also supports sustainability goals and can reduce costs over time. Moreover, localized solutions can enhance community acceptance by minimizing the need for large-scale transmission infrastructure, while aligning with evolving policy frameworks that encourage clean, resilient energy systems.

Understanding Large Loads

Before exploring solutions for powering large electrical loads, it is essential to first understand the nature of these loads and the specific requirements they impose on infrastructure and energy systems. Clear insights into their operational demands, scalability needs, and temporal usage patterns form the foundation for designing effective and resilient power strategies. North American Electric Reliability Corporation (NERC) defines a large load as “Any commercial or industrial individual load facility or aggregation of load facilities at a single site behind one or more point(s) of interconnection that can pose reliability risks to the Bulk Power System (BPS) due to its demand, operational characteristics, or other factors.” [1]

Forecasted Growth

Understanding the growth in electricity demand hinges on two key metrics: peak coincident power demand (measured in megawatts, MW) and total energy consumption over time (measured in megawatt-hours, MWh). Power demand reflects the scale of infrastructure—such as power plants and transmission lines—required to meet instantaneous load during peak periods. In contrast, energy consumption speaks to the quantity of resources needed over time, including fuel inputs like natural gas and supporting technologies such as energy storage systems. While both metrics are critical, this book primarily emphasizes power demand, with energy consumption informing the economic viability and utilization efficiency of the supporting infrastructure.

NERC produces national forecasts by aggregating data from utility operators across the country. According to its 2024 report, coincident peak demand in the U.S. is projected to reach approximately 880 GW by 2025. The forecast anticipates an increase of 68 GW—an 8% rise—over the following five years to 2030, and a 132 GW jump (15%) by 2035. However, since the release of that report, several regional authorities have issued updated projections that reveal even steeper growth trajectories—in some cases more than double the figures cited in the original NERC forecast, as shown in Table 1.1. These revisions underscore the urgency of re-evaluating infrastructure strategies to accommodate accelerating load expansion.

*Only US regions, excluding data for WECC CAN and WECC MEX

A large load generally refers to a substantial and concentrated electrical demand originating from a single user at a specific point on the grid.  Much of the recent and anticipated growth stems from these types of loads—including data centers, cryptocurrency mining operations, advanced manufacturing facilities, refinery electrification, and fleet-scale EV charging hubs. Notably, data centers and crypto operations represent the bulk of this expansion due to their high-power intensity. Although individual electric vehicles do not constitute large loads on their own, facilities that support large-scale EV fleet charging can exert significant demand at a single connection on the grid and are therefore categorized as such. Table 1.2 outlines this growth across various regions and load categories, offering a more granular view of emerging demand drivers.

*Sources in Table 2 same as Table 1

Significance of Large Load Growth

NERC does not give a numerical value for the size or range and leaves the classification and coordination up to the regional grid operators. The electric grid system operators have set rules for large loads, which depend on the size of the load. These sizes are different between the different grid operators. For example:

ERCOT requires loads greater than 20MW and co-located with generation to submit studies for Reactive Power and Sub synchronous Oscillation. They require all loads greater than 25 MW to submit information for system modeling and planning. A large load with a demand of 75 MW or greater, particularly electronic-based facilities like data centers or cryptocurrency mining operations, requires specific interconnection processes and monitoring. They must register through the Large Load Interconnection Process portal, coordinate with the ERCOT Demand Integration Team, and complete transmission & market interconnection requirements. They can participate in markets as Load Resources through a level 3 or 4 Qualified Scheduling Entity (QSE), offering energy or ancillary services. [21]

NYISO requires large loads, such as data centers and industrial facilities, to undergo an interconnection process to ensure grid reliability. This involves submitting a System Impact Study (SIS) request to assess the load’s impact on the transmission system, followed by a Facilities Study if upgrades are needed. Large loads are defined as a) greater than 10 MW connecting at a voltage level of 115 kV or above, or b) 20 MW or more connecting at a voltage level below 115 kV within a Small Transmission, or c) 40 MW or more connecting at a voltage level below 115 kV within a Large Transmission District. [22]

ISO-NE triggers detailed review for loads above 5 to 10 MW and those above 50MW are considered in system planning studies. They mandate that Load Serving Entities (LSEs) procure sufficient capacity to meet the Installed Capacity Requirement (ICR) through the Forward Capacity Market (FCM), which secures capacity three years in advance to cover peak demand plus reserves. Large loads contribute to the ICR, increasing capacity obligations for LSEs.

Visualizing the electric consumption of large loads can be challenging. To provide context, a 100 MW data center demands roughly the same amount of electricity as a city with 75,000 residents. Hyperscale data centers—those exceeding 1 GW (1,000 MW) of demand—magnify this impact significantly, consuming power equivalent to a large metropolitan area.

The pace at which these large loads are being constructed often outstrips the grid’s ability to respond. Bringing new power infrastructure online quickly requires extensive technical studies by grid operators to safeguard stability and reliability. From an operational standpoint, sudden shifts—such as a large load tripping offline or coming online—can create substantial grid management challenges, necessitating dynamic adjustments to power flows.

Infrastructure costs to deliver electricity to large loads are significant. They must demonstrate investment-worthiness, ensuring they contribute fairly to the cost of grid upgrades. Otherwise, these expenses may fall on the general customer base, resulting in higher electricity prices for all.

Every new large load raises critical questions about how to scale generation and transmission infrastructure efficiently. From a developer’s perspective, financing and planning these projects is particularly difficult without a consistent and long-term operational commitment from the load.

Loads characterized by high power requirements, but low energy utilization typically result in underutilized assets, driving up the cost per kilowatt-hour delivered. Aligning peak demand with generation capacity incurs upfront capital costs that must be justified over time through adequate energy consumption since the cost recovery mechanism for utilities is typically through the energy (MWh)consumed.

Historically, generation facilities have not been co-located with large loads, except in specific cases like petrochemical plants, emergency backup systems, or campuses with thermal energy loads where cogeneration solutions are beneficial. In traditional models, electricity flows from remote generation sources across the grid to the point of consumption. However, co-location strategies—where generation is placed behind the grid interconnection alongside the load—are gaining traction; also known as “behind the meter” generation. In such configurations, the grid serves as a backup, while the co-located system provides primary power, improving reliability and potentially reducing infrastructure strain.

Electric Requirements for Large Loads

Because electricity supply is essential to the technical and economic operation of the business for large loads, they have key requirements. This includes:

1. Fast Deployment Schedule
2. Flexible Operational Capabilities
3. High Service Reliability
4. Substantial Environmental Sustainability
5. Appropriate Site Feasibility
6. Stable under Evolving Policy
7. Low Capacity and Energy Cost

1. Deployment Schedule

Accelerating the timeline for energizing large electrical loads can unlock significant economic and strategic benefits.  STL Partners reported $14.2 Million of lost revenue, cost overruns and contractual penalties incurred for every month delayed based on a 60MW Data Center [23].  Likewise, CBRE reported an average data center lease price of $174.06 per kW per month in (H1 2024 in primary North American markets), which means a delayed 100 MW facility foregoes about $17.4 million per month in rental revenue. [24] One of the most critical and persistent barriers to timely deployment is securing a reliable power source. [25]

Ideally, the construction and commissioning of a large load facility follow a two-year schedule, as outlined in Table 1.3. However, more-often projects take much longer.  At the start, early-stage uncertainties—particularly around energy policy, incentives, and regulatory frameworks—can prevent investors from making firm financial commitments.  Once a project moves forward, supply chain limitations further constrain progress: only 20–40% of publicly announced power projects are under active construction, and transformer lead times have stretched to 2–4 years due to material shortages and manufacturing bottlenecks.

Tying into existing transmission also pushes schedules: as of 2024, more than 2,600 GW of energy projects remain stuck in interconnection queues, with average wait times of 3–5 years.  Moreover, developing supporting infrastructure such as new transmission lines take 5–10 years, driven by the enormity of such projects and the complex government, budget and legal processes.

To accommodate growing demand from large loads, it is vital to streamline policies, expedite equipment delivery, and accelerate infrastructure planning cycles. Without targeted action, mounting delays will continue to undermine the economic viability and operational readiness of large-scale projects.[26] [27] [28] [29] 

2. Operational Capabilities

Large loads require generators to have fast dynamic response to match rapid fluctuations in power consumption. Workloads associated with AI data centers, electric vehicle (EV) charging, and industrial manufacturing can vary substantially within short timeframes, requiring energy systems that can scale output in real time—without sacrificing efficiency. However, not all generators possess the ramping capabilities needed to accommodate these swings, leading to grid stress and operational constraints.

For example: Crypto mining loads can turn on and off very rapidly in response to electric prices in the real time market. Manufacturing operations, such as electric arc furnaces, can induce sudden increases in load. These activities can cause power demands to swing by tens of hundreds of megawatts multiple times per day. [30]

The load profile of data centers can oscillate significantly over time as shown in Figure 1.3. [31] Although many recent reports and studies assume continuous operation at rated capacity, preliminary research suggests that actual utilization averages around 50%. This is largely due to fluctuations in computing demand and cooling requirements. However, because next-generation data centers represent emerging technology, their true utilization patterns remain uncertain and will only become clear as operational data accumulates over time. [32]

Grid stability in such scenarios relies on a combination of inertia and fast-ramping generation. Inertia represents the stored kinetic energy in rotating electrical equipment, which acts as a buffer against sudden changes in load. It is quantified by the system’s inertial constant (H), expressed in seconds, which indicates the amount of energy (MW-seconds) available per unit of system capacity (MVA). So, the frequency of a system will respond to a sudden power change according to the formula δF/δt=∆P/2H, where ∆P is the unitless change in power (∆P/2H) represents the MW power change over twice the system MVA rating).

Fast Generator Control is the Primary Frequency Response and Automatic Voltage Regulation contributing to system stability during large load changes. The rate at which a generator can change output is called the “Ramp Rate”, R, expressed as a percentage of its power that can change per unit time. Slow ramping technologies, such as nuclear generators, have ramp rates in % per minute and fast ramping technologies, such as inverter-based BESS, have ramp rates in the % per millisecond range.

To ensure a stable and reliable power supply, the electric generation system must possess both an inertial constant and a ramp rate that are adequate for the anticipated power fluctuations of large loads. While generator controls are a separate and critical aspect of system operation, this book proceeds with the assumption that appropriate control schemes can be implemented, provided the generation equipment itself has the inherent capabilities to respond to these changes. These combined strategies AI-driven demand response, battery storage, and DERs offer robust solutions to manage the growing impact of large loads, ensuring grid stability and efficient power delivery.

3. Service Reliability

Large electrical loads have varying requirements depending on their operational purpose, as illustrated in Table 1.4. Mission-critical services demand the highest levels of reliability and are generally less sensitive to electricity costs. Many existing facilities—such as medical campuses and military bases—are actively exploring new power solutions to meet growing reliability demands. In contrast, most new large loads, including data centers and advanced manufacturing facilities, are highly sensitive to electricity pricing. Their need for reliability may vary, but cost remains a primary driver in their energy strategy.

Power outages in life-critical systems like hospitals and emergency services can have severe financial and societal impacts. A one-hour outage in a hospital can cost between $500,000 and $2 million due to canceled procedures, equipment damage, and reputational harm, as seen in FEMA’s report on Sacred Heart Hospital after Hurricane Ivan. Similarly, 911 call centers in urban areas may incur societal costs of $10,000 to $100,000 per hour from delayed emergency responses. With the value of lost load (VoLL) for critical infrastructure reaching $10,000 to $50,000 per MWh, based on customer rates in the 2023 IEA analysis. [32] To mitigate these risks, hospitals implement robust power reliability measures. A 500-bed hospital typically requires 10–15 MW to support essential systems and must comply with standards like IEEE 446-1995 and the CMS Emergency Preparedness Rule, which mandates generator startup within 10 seconds and a 96-hour fuel reserve. Facilities often use dual-feed power systems, N+1 UPS configurations, and NFPA 99-aligned generator protocols.

For critical economic operations, electric system reliability is a key requirement, where unplanned outages can incur cost between $100,000 and $1 million per minute. Cryptocurrency mining facilities, while often participating in demand response programs to cut energy costs, risk revenue losses of $50,000 to $500,000 per hour during interruptions.[34] Industrial sites face even greater impacts a 2024 Brattle study found that a one-hour outage at a 50 MW petrochemical plant in Houston could result in $500,000 to $5 million in lost revenue, with medium-sized operations averaging $1.78 million in lost revenue. Beyond direct losses, outages can trigger widespread economic ripple effects through disrupted supply chains and reduced market confidence, potentially costing tens of millions for large commercial loads. [35] Amazon’s (AWS) US-EAST-1 data center experienced a software failure in October of 2025 that crippled companies around the world for a better part of a day at an estimate of hundreds of billions in lost revenue and missed transactions.  Although that fault was not due to a power failure, loss of power would produce a similar result. [36]

To ensure reliable power delivery for critical infrastructure such as hospitals, data centers, and manufacturing, several IEEE standards provide a robust framework for system design and operational resilience.  The foundation for industrial power is the IEEE 3000 Standards Collection® for Industrial & Commercial Power Systems, a set of seven standards.   IEEE 3006.5-2014 emphasizes probabilistic risk assessment using metrics such as Loss of Load Probability (LOLP) and Expected Unserved Energy (EUE), recommending that backup systems be sized for peak demand, with initial restoration efforts limited to less than 5% of synchronized capacity. It prioritizes critical loads through automated load-shedding and redundant configurations such as N+1 or 2N. IEEE 3006.7-2013 addresses continuous power systems for large facilities, introducing a tier classification (I–IV) and underscoring the importance of Failure Modes and Effects Analysis (FMEA), cooling load considerations, and real-time monitoring. [37]

Data centers and cryptocurrency facilities have adopted these standards to achieve high reliability, often targeting “five nines” (99.999%) availability. Table 1.5 outlines the characteristics of data center tiers, detailing redundancy, maintenance protocols, availability, and failure probabilities over five years. Data centers typically maintain voltage within ±5% and frequency within ±1 Hz, while cryptocurrency operations prioritize voltage stability within ±10% and employ high-capacity feeders. Both sectors implement 2N or N+1 redundancy for UPS and generator systems, achieving up to 99.982% uptime in line with Uptime Institute Tier standards. Poor power quality remains a significant risk, with surges potentially damaging servers and cooling systems—resulting in losses ranging from $100,000 to over $1 million per event—and industrial machinery repairs costing $50,000 to $500,000 per outage. Aligning regulatory frameworks with these standards enables policymakers to support resilient power systems that meet the stringent reliability demands of critical and high-load facilities.

At the other end of the spectrum, flexible loads can play a valuable role in supporting lower electricity prices and enhancing grid reliability for all consumers. As the electric market evolves, new rules are being established to integrate flexible loads systems capable of rapidly increasing or decreasing demand in response to real-time pricing or ancillary service signals. When coordinated effectively with utility needs, flexible loads can help stabilize market prices and improve overall grid performance. However, if not properly managed, these loads can introduce volatility by shifting demand in ways that challenge grid stability.

4. Environmental Sustainability

Large electrical loads—such as data centers, advanced manufacturing facilities, and energy-intensive industrial operations—are increasingly navigating a complex landscape of environmental expectations and regulatory demands. Internally, many companies are setting ambitious sustainability goals—including carbon neutrality and full renewable energy sourcing—to meet investor and stakeholder priorities, enhance public trust, and manage long-term risk. Externally, regulatory frameworks continue to evolve, with federal emissions disclosures, state-level greenhouse gas reduction mandates, and local ordinances governing air quality, water usage, and land impact driving deeper scrutiny of operational practices.

Environmental compliance is often a critical path item for securing permits and public support, particularly in urban and densely populated regions where regulatory barriers are highest. In such settings, low- and zero-emission technologies are not only environmentally preferable—they also enable faster and more efficient project approvals. Community concerns over air and noise pollution are shaping permitting outcomes, and large loads must demonstrate responsible siting and mitigation strategies to gain traction.

The lingering environmental impact of fossil-based systems remains a challenge, especially where reliable power is non-negotiable. To bridge sustainability with reliability, many operators are now exploring hybrid configurations, integrating carbon capture technologies, behind-the-meter renewables, and energy storage to reduce emissions while maintaining operational resilience.

5. Site Feasibility

When planning energy delivery for large electrical loads, location plays a central role, not only for the load itself but for the associated power infrastructure. These loads often have strict siting requirements related to proximity to electric transmission lines, natural gas pipelines, fiber networks, logistics corridors, and skilled labor pools. However, their spatial demands and operating intensity also transfer significant constraints onto the electrical system, requiring tailored strategies for generation siting, transmission routing, and regulatory compliance.

In urban and high-density regions, minimizing the land footprint of generation is a critical challenge. Space-constrained environments often make high power density technologies—such as combined-cycle gas turbines (CCGTs)—more attractive, given their ability to deliver substantial output from a compact footprint. Conversely, low-density renewables like solar and wind, though environmentally favorable, struggle to compete in these settings due to their large spatial requirements, making them less viable.

Community safety and environmental standards further shape siting decisions. Public concern over nuclear energy—driven by perceptions of risk, waste handling, and proximity to population centers—can restrict deployment even when technically feasible. Additionally, many urban areas remain out of compliance with federal and state air-quality standards, which significantly limits the use of fossil-fueled generators. Localized opposition to visual obstructions—such as turbine blades, exhaust stacks, and steam plumes—adds another layer of complexity, especially in scenic or residential zones where aesthetics influence public support and permitting success.

Energy resource availability is also geographically uneven, especially for renewable energy. Wind, solar, and geothermal potential varies considerably across regions, and effective sourcing must balance resource quality with load proximity. Co-locating generation with load—when possible—can reduce transmission congestion and enhance reliability, but only if the local resource is sufficient to meet demand.

In many cases, planners must make trade-offs between site feasibility, integration costs, and regulatory barriers, in choosing electric supply technologies that best match both the physical and policy landscape of the chosen location for the large load.

6. Evolving Policy

Stable energy policy is fundamental to the viability of large-load projects. Power-intensive developments require long-term certainty in energy availability, cost structures, and access to grid infrastructure. Without reliable policy signals, investors and developers face increased risk, unpredictable returns, and costly delays in execution.

The Inflation Reduction Act (IRA) of 2022 played a pivotal role in accelerating low-carbon electricity supply by offering robust federal incentives for clean energy and renewable infrastructure. Its provisions subsidized new opportunities for energy storage deployment, emissions monitoring, and carbon reduction technologies, helping large loads align with sustainability targets. However, this momentum was disrupted by the 2025 Big Beautiful Bill, which scaled back many of the IRA’s incentives for renewables and pushed for fossil fuels and nuclear power. The policy switch underscored a critical vulnerability in long-term planning: the economic assumptions underpinning large-load investments can rapidly shift, affecting project feasibility and timeline. [39]

Efforts by the Federal Energy Regulatory Commission (FERC) have aimed to mitigate these risks by streamlining grid access. Through Order No. 2023, FERC reformed interconnection processes to reduce backlogs and prioritize shovel-ready projects. These reforms introduced standardized study timelines and transparent cost allocation mechanisms, giving large-load developers a clearer pathway through interconnection queues and lowering uncertainty around grid integration. [40]

Energy storage policy has also become a vital component of infrastructure planning. With federal tax credits extended to standalone storage, and states like California and Michigan setting multi-gigawatt targets, large-load facilities have gained a flexible tool to manage peak demand, improve reliability, and buffer renewable intermittency. Still, regulatory alignment across states and markets remains fragmented, requiring developers to navigate complex compliance landscapes. [41]

Another cornerstone of feasibility is tariff and rate stability. Transmission and distribution (T&D) charges must be predictable and fairly allocated to ensure cost transparency. Reforms introduced by FERC and state regulators have helped standardize pricing structures and reduce cross-subsidization, allowing large-load users to forecast long-term operating costs more accurately. Without such mechanisms, energy-intensive facilities risk being exposed to sudden rate hikes or absorbing disproportionate infrastructure costs.

Ultimately, stable policy environments shape capital planning, site selection, operational resilience, and investor confidence. When aligned and predictable, energy policy becomes a strategic enabler—unlocking innovation, accelerating deployment, and reducing economic risk for large-load consumers.

7. Capacity & Energy Cost

For operations with large electrical loads, the cost of electric power and back-up power systems is a critical factor in both financial planning and operational resilience. These facilities often operate continuously and consume gigawatt-hours of electricity, meaning that even minor changes in utility rates or demand charges can translate into millions of dollars annually. In fact, electrical power infrastructure—including primary and backup systems—can account for 10% to 40% of a facility’s capital expenditures (CAPEX), while electricity costs may comprise up to 75% of operating expenses, as shown in Table 1.6. For high-consumption facilities, electricity can represent nearly half of the total lifecycle cost of the business.

Sources:  [42] [43] [44] [45] [46] [47] [48]

This economic reality makes access to low-cost electricity essential for maintaining product competitiveness and overall business viability. Companies often choose facility locations based on electricity pricing, and even a modest increase in retail electricity rates over the business’s lifespan can threaten financial stability. In addition to low cost, these companies depend on stable electricity pricing. For instance, consider a large industrial load where electricity costs constitute 50% of operating expenses and the entity maintains a healthy operating margin of 30%. If electricity costs increase by 50%—from $70/MWh to $105/MWh—the operating margin would drop to just 12.5%. For energy-intensive enterprises, unpredictable electricity pricing is a direct risk to the bottom line.

Equally important is the cost and reliability of back-up power systems. In data centers, uninterruptible power supplies (UPS), diesel generators, and increasingly, battery energy storage systems (BESS) are deployed to ensure uptime during primary source outages. The capital investment in these systems is substantial, and ongoing maintenance, fuel, and testing add to operational costs. However, the cost of downtime—measured in lost revenue, data integrity, and customer trust—can far exceed the investment in robust backup infrastructure. In manufacturing, a power interruption can halt production lines, damage in-process materials, and disrupt supply chains, making backup systems not just a safety net but a strategic necessity.

As energy markets evolve and sustainability goals become more prominent, many large-load operators are also exploring renewable integration and demand response programs to manage costs and enhance resilience. Thus, the economics of both primary and backup power are not just operational concerns—they are central to long-term strategic planning.

References

[1] Large Loads Task Force, “Characteristics and Risks of Emerging Large Loads,” North American Electric Reliability Corporation, July 2025. [Online]. Available:  https://www.nerc.com/comm/RSTCReviewItems/3_Doc_White%20Paper%20Characteristics%20and%20Risks%20of%20Emerging%20Large%20Loads.pdf [Accessed 19 June 2025].

[2] Office of Energy Policy and Innovation, “2024 State of the Market,” Federal Energy Regulatory Commission, 25 March 2025. [Online]. Available: https://www.ferc.gov/sites/default/files/2025-03/25_State-of-the-Market_0320_1200.pdf. [Accessed 21 May 2025].

[3] North American Electric Reliability Corporation (NERC), “2024 Long-Term Reliability Assessment,” North American Electric Reliability Corporation (NERC), 15 July 2025. [Online]. Available: https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/NERC_Long%20Term%20Reliability%20Assessment_2024.pdf. [Accessed 22 May 2025].

[4] Electric Reliability Council of Texas, “ERCOT System Planning 2025 Long Term Hourly Peak Demand and Energy Forecast,” Electric Reliability Council of Texas, 8 April 2025. [Online]. Available: https://www.ercot.com/files/docs/2025/04/08/2025-LTLF-Report.pdf. [Accessed 21 May 2025].

[5] Electric Reliability Council of Texas Staff, “2024 Long-Term System Assessment (LTSA) Update,” Electric Reliability Council of Texas , 16 May 2024. [Online]. Available: https://www.ercot.com/files/docs/2023/05/12/2024_LTSA_update_20230516_v1.0.pdf. [Accessed 21 May 2025].

[6] P. Vegas, W. Rickerson and R. l. Scheel, “Item 8.1: Long-Term Load Forecast Update (2025-2031) and Methodology Changes,” Electric Reliability Council of Texas, 8 April 2025. [Online]. Available: https://www.ercot.com/files/docs/2025/04/07/8.1-Long-Term-Load-Forecast-Update-2025-2031-and-Methodology-Changes.pdf. [Accessed 21 May 2025].

[7] New York Independent System Operator (NYISO), “2025 Load & Capacity Data- Gold Book,” New York Independent System Operator (NYISO), 30 April 2025. [Online]. Available: https://www.nyiso.com/documents/20142/2226333/2025-Gold-Book-Public.pdf/088438e1-02f1-5316-211b-dbca17c01b4b?t=1745932590307. [Accessed 21 May 2025].

[8] New York Independent System Operator (NYISO), “2024 Reliability Needs Assessment (RNA),” New York Independent System Operator (NYISO), 19 November 2024. [Online]. Available: https://www.nyiso.com/documents/20142/2248793/2024-RNA-Report.pdf/. [Accessed 21 May 2025].

[9] California Independent System Operator (CAISO), “2025 Summer Loads and Resources Assessment,” California ISO, 5 May 2025. [Online]. Available: https://www.caiso.com/content/summer-loads-resources-assessment/2025/index.html. [Accessed 21 May 2025].

[10] California Independent System Operator (CAISO), “2024-2025 Transmission Planning Process: Draft Transmission Plan,” California ISO, 15 April 2025. [Online]. Available: https://www.caiso.com/documents/revised-draft-2024-2025-transmission-plan.pdf. [Accessed 21 May 2025].

[11] Midcontinent Independent System Operator (MISO), “2024 Regional Resource Assessment,” Midcontinent Independent System Operator (MISO), 11 February 2025. [Online]. Available: https://cdn.misoenergy.org/2024%20RRA%20Report_Final676241.pdf. [Accessed 21 May 2025].

[12] SPP Engineering, “2024 Integrated Transmission Planning Assessment Report,” Southwest Power Pool, 24 January 2025. [Online]. Available: https://www.spp.org/media/2229/2024-itp-assessment-report-v10.pdf. [Accessed 21 May 2025].

[13] A. Salehian, “Future Energy and Resource Needs Study (FERNS) Results,” Southwest Power Pool, 18 March 2025. [Online]. Available: https://www.brattle.com/wp-content/uploads/2025/03/07-2025-03-18-FERNS-Results.pdf. [Accessed 21 May 2025].

[14] K. Schlichting, “ISO New England Power Grid Outlook: CBIA 2024 Energy & Environmental Conference,” ISO New England, 6 June 2024. [Online]. Available: https://www.iso-ne.com/static-assets/documents/100012/iso_new_england_overview_and_regional_update_2024_cbia_wide.pdf. [Accessed 24 May 2025].

[15] PJM Interconnection, “PJM Load Forecast Report,” PJM Interconnection, 1 February 2024. [Online]. Available: https://www.pjm.com/-/media/DotCom/library/reports-notices/load-forecast/2024-load-report.pdf. [Accessed 21 May 2025].

[16] PJM Interconnection, “2024 Regional Transmission Expansion Plan,” PJM Interconnection, 17 April 2025. [Online]. Available: https://www.pjm.com/-/media/DotCom/library/reports-notices/2024-rtep/2024-rtep-report.pdf. [Accessed 21 May 2025].

[17] K. Rogers, J. Snitman and G. Karandikar, “WECC Board Meeting Technical Session – Large Loads,” Western Electricity Coordinating Council (WECC), 11 March 2025. [Online]. Available: https://www.wecc.org/sites/default/files/documents/meeting/2025/March%202025%20Technical%20Session%20Book.pdf. [Accessed 27 May 2025].

[18] Western Electricity Coordinating Council (WECC), “Western Assessment of Resource Adequacy (WARA),” Western Electricity Coordinating Council (WECC), 2024. [Online]. Available: https://feature.wecc.org/wara/. [Accessed 27 May 2025].

[19] M. Montgomery, “SERC Regional Long-Term Assessment & Risk Trend Overview,” SERC Reliability Corporation, May 2024. [Online]. Available: https://eastfuelconf.com/wp-content/uploads/2024/05/Melinda-Montgomery-SERC-LTRA-and-Reliability-Risk-Presentation-EFBC.pdf. [Accessed 27 May 2025].

[20] SERC Reliability Corporation, “2023–2033 SERC Annual Long-Term Reliability Assessment Report,” SERC Reliability Corporation, 2024. [Online]. Available: https://www.serc1.org/docs/default-source/program-areas/reliability-assessment/reliability-assessments/2023-2033-serc-annual-long-term-reliability-assessment-report.pdf. [Accessed 22 May 2025].

[21] Electric Reliability Council of Texas, “ERCOT Nordal Protocol,” Electric Reliability Council of Texas, 1 April 2023. [Online]. Available:
https://www.ercot.com/files/docs/2023/03/29/April%201%2C%202023%20Nodal%20Protocols.pdf. [Accessed 21 May 2025].

[22] C. Alonge, “Large Load Facility Discussion: Data Submission Requirements and Forecasting Methos,” New York Independent System Operator (NYISO), 26 August 2022. [Online]. Available: https://www.nyiso.com/documents/20142/32974180/LargeLoadForecast_Discusson_LFTF_20220826_V3.pdf/b9766465-3c10-a390-44d1-b7853f251e24. [Accessed 2 June 2025].

[23] J. Topp-Mugglestone, “Preventing Multimillion Dollar Data Centre Losses Through Reporting,” STL Partners Executive Briefing, commissioned by Foresight Works, March 2025.

[24] North America Data Center Trends H1 2024. CBRE, 2024. [Online]. Available: https://www.cbre.com/insights/reports/north-america-data-center-trends-h1-2024. [Accessed: 6. 17, 2025].

[25] A Ansar, “Everyday counts on a project: the financial impact on time on data center construction,” Data Center Dynamics, 25 July 2024, [Online]. Available: https://www.datacenterdynamics.com/en/opinions/everyday-counts-on-a-project-the-financial-impact-of-time-on-data-center-construction/. [Accessed 6 Oct 2025].

[26] Energy Technology Policy (ETP) Division , “Global EV Outlook 2024,” International Energy Agency IEA, 23 April 2024. [Online]. Available: https://iea.blob.core.windows.net/assets/a9e3544b-0b12-4e15-b407-65f5c8ce1b5f/GlobalEVOutlook2024.pdf. [Accessed 21 May 2025].

[27] O. Hevia-Koch, B. Wanner and R. Kuwahata, “Electricity Grids and Secure Energy Transitions,” International Energy Agency, 16 October 2023. [Online]. Available: https://iea.blob.core.windows.net/assets/ea2ff609-8180-4312-8de9-494bcf21696d/ElectricityGridsandSecureEnergyTransitions.pdf. [Accessed 22 May 2025].

[28] International Energy Agency, “Critical Minerals Market Review 2023,” International Energy Agency, 10 July 2023. [Online]. Available: https://iea.blob.core.windows.net/assets/c7716240-ab4f-4f5d-b138-291e76c6a7c7/CriticalMineralsMarketReview2023.pdf. [Accessed 22 May 2025].

[29] National Renewable Energy Laboratory, “Explained: Fundamentals of Power Grid Reliability and Clean Electricity,” National Renewable Energy Laboratory, Golden, 2024.

[30] T. H. Norris, T. Profeta, D. Patino-Echeverri and A. Cowie-Haskell, “Rethinking Load Growth: Assessing the Potential for Integration of Large Flexible Loads in US Power Systems,” Nicholas Institute for Energy, Environment & Sustainability, Duke University, Durham, 2025.

[31] E. Choukse, et. al., “Power Stabilization for AI Training Datacenters,” arXiv:2508.14318v2, 2025. [Online]. https://doi.org/10.48550/arXiv.2508.14318.

[32] T. Norris, “The Puzzle of Low Data Center Utilization Rates”, Power & Policy, August 7, 2025. [Online]. Available: https://www.powerpolicy.net/p/the-puzzle-of-low-data-center-utilization. [Accessed 22 August 2025].

[33] Analytical framework for electricity security” International Energy Agency series on Electricity Security 2021, IEA 2021. [Online] https://www.iea.org/reports/analytical-frameworks-for-electricity-security.

[34] U.S. Energy Information Administration, “Tracking Electricity Consumption from U.S. Cryptocurrency Mining Operations,” U.S. Energy Information Administration, 1 February 2024. [Online]. Available: https://www.eia.gov/todayinenergy/detail.php?id=61364. [Accessed 2 June 2025].

[35] The Brattle Group, “Review of Value of Lost Load in The ERCOT Market,” Public Utility Commission of Texas, Austin, 2024.

[36] C. Holloman, “AWS Outage: Billions Lost, Multi-Cloud Is Wall Street’s Solution,” Forbes, 20 Oct 2025. [Online] https://www.forbes.com/sites/christerholloman/2025/10/20/aws-outage-billions-lost-multi-cloud-is-wall-streets-solution/. [Accessed 24 Oct 2025].

[37] “IEEE 3000 Standards Collection for Industrial & Commercial Power Systems”, IEEE Standards Association, [Online] https://standards.ieee.org/products-programs/ieee-3000/ [Accessed 24 Aug 2025].

[38] Uptime Institute, “Tier Classification System,” Uptime Institute, 2025. [Online]. Available: https://uptimeinstitute.com/tiers. [Accessed 21 May 2025].

[39] J. A. Cavanaugh, A.-S. Corbeau, T. Moerenhout, M. Bowen, A. Finan and N. Kaufman, “Assessing the Energy Impacts of the One Big Beautiful Bill Act,” Center for Global Energy Policy at Columbia, 14 July 2025. [Online]. Available: https://www.energypolicy.columbia.edu/assessing-the-energy-impacts-of-the-one-big-beautiful-bill-act/. [Accessed 18 July 2025].

[40] Federal Energy Regulatory Commission (FERC), “Improvements to Generator Interconnection Procedures and Agreements, Docket No. RM22-14-000, Order No. 2023,” Federal Energy Regulatory Commission (FERC), 23 January 2025. [Online]. Available: https://www.ferc.gov/explainer-interconnection-final-rule#. [Accessed 18 July 2025].

[41] S. Newell, A. Levitt, A. Thompson, S. Patel, E. Snyder and A. Yan, “Energy Storage Market Design Roadmap,” The Brattle Group, April 2025. [Online]. Available: https://www.brattle.com/wp-content/uploads/2025/04/Energy-Storage-Market-Design-Roadmap.pdf. [Accessed 21 May 2025].

[42] M. Smolaks, “Data Center Costs Set to Rise and Rise,” Uptime Institute, 29 March 2023. [Online]. Available: https://journal.uptimeinstitute.com/data-center-costs-set-to-rise-and-rise/. [Accessed 22 April 2025].

[43] Technology Engagement Center, “Data Centers: Jobs and Opportunities in Communities Nationwide,” U.S. Chamber of Commerce, 1 May 2017. [Online]. Available: https://www.uschamber.com/assets/archived/images/ctec_datacenterrpt_lowres.pdf. [Accessed 22 May 2025].

[44] Stream Data Centers, “Data Center Cost,” Stream Data Centers, 2025. [Online]. Available: https://www.streamdatacenters.com/resource-library/glossary/data-center-cost/. [Accessed 21 May 2025].

[45] Thunder Said Energy, “Data-centers: the economics,” Thunder Said Energy, Austin, 2023.

[46]  J. Howeel, “Breaking Down Data Center Cost: Building vs. Outsourcing,” ENCOR Advisors, 22 October 2024. [Online]. Available: https://encoradvisors.com/data-center-cost/. [Accessed 6 June 2025].

[47] American Hospital Association, “The Cost of Caring: Challenges Facing America’s Hospitals in 2025,” April 2025. [Online]. Available: https://www.aha.org/costsofcaring. [Accessed 21 May 2025].

[48] K. Garside, “The Role of Energy Management Systems in Controlling Healthcare Energy Expenses,” Integrity Energy, 29 September 2023. [Online]. Available: https://www.integrityenergy.com/blog/the-role-of-energy-management-systems-in-controlling-healthcare-energy-expenses/. [Accessed 23 May 2025].