Cracking the speed to power code

For decades, the process of connecting new power to the American grid was steady and predictable. Regional transmission organizations like PJM Interconnection—America’s largest and most complex grid—were historically built to review a manageable flow of centralized power projects each year. Ten years ago, PJM fielded 389 requests to interconnect to its grid, representing only 36 gigawatts.

But the era of flat energy demand is over. In April of this year, PJM fielded a massive influx of 811 generation requests in the first cycle of its newly reformed interconnection process. RTOs like PJM are facing historic numbers of applications as we push the limits of the current grid amid soaring demand. If we cannot solve the problem of speed to power, we will bottleneck our own economy.

This surging volume of interconnection requests is the result of compounding trends. The continuing electrification of our economy across transportation, buildings, and manufacturing had already begun pushing legacy grid processes to their limits. Now, the great American data center buildout has added to that trend, with analysts projecting it to represent 2 percent of U.S. GDP this year.

The ongoing investment in data center construction is an undeniable economic reality, one that is driving a massive physical expansion of our power systems. Looking strictly at generation, the 617 projects that have entered the first phase of PJM's new interconnection process represent the potential to add nearly 169 GW of new capacity. To put that sheer scale into perspective, this single study cycle proposes adding an amount of power nearly equal to the entire nameplate generation capacity of PJM’s existing grid, which was built out over almost a century.

The grid is “a single machine,” but it doesn’t yet act like one.

Engineers, planners, and policymakers across the power sector face the scale of this challenge. Each of these interconnection requests requires multiple hours of application review. Engineers have to evaluate each application under a tight timeline against regulatory tariff criteria that can stretch thousands of pages. And this is before running detailed studies and power flow analyses across hundreds of simulations under various operating conditions, a process that is stretching out interconnection queues to as long as seven years.

To its credit, PJM has overhauled its interconnection rules to keep up with today's demand surge. But even the best policy reforms will still hit a wall if they rely on legacy processes and manual workflows to execute them.

For their part, policymakers, including in Congress and at FERC, have acknowledged this shifting landscape. In June, FERC issued orders to the six RTOs it regulates and their respective transmission owners to examine whether changes are needed to their interconnection and planning processes in light of new load growth. The ability to rapidly and accurately model these changes is essential.

The country cannot afford a modest response. While years ago, RTOs lacked the tools to reliably and safely speed interconnection or explore new solutions for studies and models, today innovation has caught up with the grid, with technology that can help securely deliver speed to power.

A prime example of this is HyperQ, an agentic AI solution built by Tapestry that allows RTOs and developers to rapidly evaluate interconnection applications. This isn’t software that was vibecoded overnight. Working directly with PJM’s planning team, we mapped each crucial step in their review to handle the flood of new project requests.

Deployed for the first cycle of PJM’s reformed interconnection process, HyperQ tackles administrative and technical hurdles by simultaneously comparing site control provisions and lease documentation against regional tariffs, some extending to thousands of pages. Instead of manual site control reviews that historically took multiple hours per application, the software completes its initial parsing and synthesis of complex documentation with a median computational runtime of just six minutes. It surfaces clear, cited deficiencies to support PJM's expert reviewers, who then conduct their own analysis and make all final determinations.

We’ve also worked with grid operators around the world for years to build a physics-based AI platform that can optimize their existing operations, because we know the fastest electron is the one already on the grid. This tool gives planners the ability to visualize the impact of grid-enhancing technologies that can increase the efficiency of existing assets, showing them exactly where on the grid upgrades like power flow controls, dynamic line ratings, and reconductoring can unlock the most spare capacity.

This platform can also be used to model co-located generation and load, flexible load behavior, and re-dispatch strategies. Speeding interconnection further and limiting infrastructure overbuild can lower costs for every grid customer.

But perhaps the most vital way software can benefit our grid is by facilitating full coordination and transparency between the various models employed by transmission owners and planners. Transmission planning modeling is a version-control nightmare, relying on manual processes, email coordination, and file-sharing tools. Different studies—like steady-state power flow, transient stability, short-circuit, and electromagnetic transient studies—all require forking base models and manipulating them in distinct software formats.

Worse yet, grid operators and utilities each have different models of the grid, each with input from third parties, who all don’t see the full picture and can disagree about accuracy. As more and more equipment is added to the grid, these small discrepancies can add up to large problems. And many manufacturers of inverter-based resources like batteries, solar panels, and wind turbines employ proprietary code that they do not share with planners and operators. This massive lack of coordination is labor-intensive, error-prone, and serves as a major bottleneck, slowing down almost every aspect of grid and transmission operations.

As we detail in our new white paper released today, the path forward is clear: grid data that is complete, accurate, interoperable and secure underpinning a unified, collaborative workspace for grid participants. This allows engineers and others to sync planning, management, and operation while balancing critical energy infrastructure information and essential cybersecurity protocols and safeguards. What Git did for code, we must do for grid data, creating a single source of truth for all parties that can speed collaboration between developers, utilities, and grid operators.

Rather than viewing our nation’s growing energy demands as an intractable challenge, this surge presents a generational opportunity to modernize our aging infrastructure. Instead of tweaking legacy tools or patiently piloting new solutions, the current environment offers the chance for bold innovation and rapid deployment. New approaches—from automation to interoperability to AI-driven power system modeling—now exist to meaningfully and safely manage this transition. The policymakers, operators, and owners of our grid can seize this moment to bring the 20th century’s greatest, most complex machine into the future.

Read our white paper.