Visions of an AI-powered future are driving a global race to develop and deploy digital infrastructure and manufacturing facilities at breakneck speed. But this future endeavor is hitting a critical roadblock, as the aging grid struggles to keep up with soaring demand.Â
Over the next three years, the average number of data center builds is expected to increase sixfold, signaling a dramatic acceleration in development. This surge will be primarily driven by AI, with electricity demand from AI-optimized data centers projected to more than quadruple by 2030.1 In ERCOT alone, data center demand is expected to add over 22 GW to the system peak by 2030. Similarly, adjustments were made in PJM’s latest long-term forecast, which accounts for large, unanticipated load changes, to reflect the unprecedented pace of data center construction. The challenge for the energy industry is two-fold: new facilities are large-scale, requiring enough power to serve the equivalent of small cities; and these new loads need power on timescales that are much faster than traditional grid planning and deployment have moved. Â
 PJM’s latest long-term forecast now reflects the unanticipated arrival of new data centers.Â
 When a new facility needs to connect to the grid, an interconnecting customer needs to coordinate with utilities and transmission planners to successfully work through relevant engineering studies and permitting processes. Historically, the process was relatively quick and predictable. Now, due to complexities related to declining grid capacity, tighter air quality limits and permitting hurdles, and the evolving generation mix, interconnections can now take seven years or longer. For industries that move at the speed of innovation and thrive on rapid change and continuous advancement, these delays are simply unacceptable.Â
These lengthy interconnection timelines aren’t just an inconvenience; they represent a serious threat to economic growth, technological progress, and national security. When a new data center or manufacturing site is ready to operate but must wait years for grid capacity, valuable opportunities are lost. Beyond the financial impact on the operators, local economies miss out on jobs, innovation is delayed, and the nation risks falling behind competing nation-states that have unbottled infrastructure deployment at larger scale and speed. Â
Grid operators and planners will need new tools to match the innovation and growth of the customers they are serving, while continuing to meet their core obligations to safety, reliability, and affordability.Â
Flexible onsite power: A smarter, faster alternativeÂ
Flexible onsite power generation presents a compelling alternative to the standard model of centralized power supply or antiquated backup power. By installing dispatchable generation capacity directly at or near the load, developers can bring facilities online much more quickly without waiting for necessary grid upgrades or navigating complex permitting backlogs that can be avoided. Dispatchable capacity can include technologies like natural gas generation, fuel cells, or renewable solutions supported by energy storage.Â
These systems – often referred to as microgrids – can be deployed in months instead of years. Whether temporary or permanent, they offer a reliable source of power that can be operated in various modes from baseload, peaking, or backup power. Many flexible generation assets can also be configured to ramp up or down based on grid signals, allowing large users to play a supportive role in maintaining grid stability when grid interconnection is available.Â
Unlocking speed and certainty in a volatile marketÂ
In industries where time-to-market is a key competitive advantage and financing is constrained, the ability to quickly secure power can be the difference between success and stagnation. Flexible onsite generation can provide the required certainty and speed, while status-quo solutions are stuck in myriad delays. As a result, developers won’t have to gamble on uncertain interconnection and transmission schedules or supply chain slowdowns.Â
This also offers economic value beyond reliability. Projects that avoid delays are less exposed to inflationary pressures, labor shortages, and supply chain volatility. In addition, predictable energy costs, especially from cleaner-burning fuels or renewables, can improve long-term budgeting and reduce exposure to wholesale price spikes.Â
Moving beyond self-sufficiency to support the grid Â
The greatest value from flexible onsite power generation isn’t in helping loads defect from the grid; it’s in supplementing it. During periods of high demand or grid stress, which are typically limited in duration, large energy users can reduce their draw from the grid and shift to onsite generation when directed. This kind of load flexibility helps balance supply and demand across the system, reduces the need for costly centralized infrastructure, and mitigates the risk of outages during extreme weather, unplanned transmission outages, and/or supply constraints.Â
The key to flexibility lies in the on-demand control and dispatchability of onsite resources. When paired with real-time market data and grid conditions, these systems can operate intelligently and autonomously, providing energy when it’s most valuable and reducing emissions when it’s most impactful. When renewables ramp down, flexible generation can ramp up to fill the gap and ensure system reliability. Onsite generation can also be dispatched to manage and reduce transmission congestion and losses, allowing for more efficient utilization of existing transmission networks.Â
The search for speedy grid flexibility solutionsÂ
Grid operators are actively tackling the dual challenges of inadequate generation and speed-to-market. In SPP, the proposed Conditional High Impact Large Load Service (CHILLS) offers a potential bridge to firm power. This service would allow new, large loads to consume grid power on as-available basis, temporarily bypassing the long wait for new grid infrastructure. When the grid cannot accommodate the demand, data centers using this non-firm service can rely on their own flexible, paired generation—which can be located either onsite or elsewhere—to ensure operational continuity. This proposed High Impact Large Load Generator Assessment (HILLGA) process is complementary to CHILLS, and it pairs new loads with their own dedicated, “bring-your-own” generation or microgrid.Â
SPP is fast-tracking new loads onto the grid by creating an interruptible service and a streamlined process for generation projects that are paired with those loads.Â
A cleaner path forwardÂ
While diesel generators have long been and continue to be the standard for onsite backup power, they’re increasingly out of step with today’s emissions standards and the grid’s operational needs. Case in point, diesel generators may only provide up to 50 hours annually of non-emergency runtime for the grid, while the peak needs on utility systems may be anywhere from 350-3,000 hours annually .Â
Fortunately, there are commercially proven alternatives to diesel generation that can provide backup power services, have far lower emissions profiles, and utilize more robust fuel supply chains while unlocking grid support and market participation. For example, generators using low pressure natural gas and renewable natural gas (RNG) can cut local pollutants by up to 96% compared to traditional diesel. RNG, sourced from food waste or agricultural byproducts, can achieve carbon neutrality for facilities looking to meet corporate sustainability goals. These fuels can be tracked and certified, offering transparency and accountability in emissions reporting. Additionally, smart microgrid controllers and automated platforms are making it easier than ever to monitor and optimize dispatch around carbon intensity in real time. Â
Flexibility: A new playbook for powerÂ
New large loads cannot wait for legacy infrastructure timelines. If we want to sustain growth in a fast-moving digital economy and meet ambitious climate targets, we need to rethink how we power new demand and ensure grid stability together. Flexible onsite generation paired with dynamic grid management must become a central strategy for large load project developers, utilities, grid planners, and operators.Â