The electricity conversation in America has become increasingly focused on one question: How do we meet rising demand?
The surge in artificial intelligence, data centers, electrification, and domestic manufacturing is placing unprecedented pressure on the grid. The industry’s response has largely centered on adding supply, such as building more generation, expanding transmission, and accelerating new infrastructure projects. Those investments will be necessary, but they take many years to decades to implement and impact only part of the equation.
COMMENTARY
One of the largest untapped opportunities to increase available grid capacity lies within existing buildings, many of which were never designed to minimize their strain on the grid.
For utilities, grid flexibility is not simply a matter of generating more electricity, it is the ability to maximize the value of the electricity already available. Today, a substantial portion of that capacity remains effectively stranded, held in reserve to satisfy peak demand created by inefficient building envelopes and outdated building systems.
The opportunity is hiding in plain sight. By reducing demand in existing buildings, utilities can unlock capacity, strengthen grid resilience, and defer costly infrastructure investments. As electricity demand accelerates, the built environment may prove to be one of the largest and most underutilized energy resources available.
Why the Supply Side Became the Default
The push toward new generation is not irrational. AI-driven electricity demand is accelerating and concentrated in ways that stress specific areas of the grid rather than distributing load evenly. Utilities have a mandate to keep the lights on, and when demand rises, building new capacity is the most straightforward response.
Utility skepticism toward demand-side solutions was understandable. For decades, the most visible efficiency measures improved building performance but rarely delivered results meaningful enough to influence grid planning LED lighting retrofits were among the most successful efficiency initiatives of the past several decades, often reducing lighting energy consumption by 20% to 60%. Since lighting represents only a small fraction of total building energy use, the resulting impact on grid capacity was negligible. The second wave of energy efficiency improvements, building management systems and variable frequency drives (VFDs) had an even smaller impact than LED lighting.
The conclusion utilities drew was reasonable given the options available. There was no demand-side solution capable of delivering grid-scale impact, so supply-side investment was the only real option. The question worth asking now is whether that conclusion still holds, or whether the technology environment has changed enough to reopen it.
Where Grid Stress Actually Originates
Operational flexibility is a peak-load problem. Grid capacity must be allocated and held in reserve for peak demand events, even when that capacity sits idle the vast majority of hours in a year.
The scale of that problem is significant. According to the IEA, grids are built to serve peak demand but carry substantial unused capacity during non-peak periods, and over 2,500 GW of renewable, large-load and storage projects are currently stalled in grid queues worldwide as a result. That stranded capacity represents a real cost embedded in infrastructure, rate structures, and long-term capital planning.
The geography of peak demand matters here. Grid pressure is highest in dense urban areas, and those are also the places where the existing commercial building stock is oldest and most energy intensive. Buildings account for roughly 40% of total U.S. electricity consumption. The dominant driver of HVAC load within those buildings, and therefore the dominant driver of peak electricity draw, is the building envelope, specifically the windows due to their lack of sufficient insulation.
Most people in the built environment know that approximately 50% to 60% of heat loss and heat gain in a commercial building moves through the glazing. What that means for grid operators is that the peak demand events utilities are building capacity to serve are, in large part, a function of building envelopes that were designed 40, 50, or 60 years ago with low-performing single pane or double pane glass and have never been meaningfully upgraded.
The Existing Building Stock as Grid Infrastructure
New commercial construction is increasingly efficient. Better glazing specifications, tighter codes, and modern mechanical systems mean that buildings coming online today are not the problem. The inefficiency lives in the existing stock, and that is also where the grid impact opportunity is greatest.
The scale of the opportunity is substantial. Commercial buildings represent billions of square feet of inefficiently ventilated space, much of it operating with aging envelopes that bleed energy and mechanical systems that drive unnecessary peak demand.
The barrier has always been assumed to be the cost and disruption of meaningful retrofits. That assumption was built on a previous generation of available solutions. Deep envelope upgrades historically meant invasive window replacement, extended shutdowns, and capital costs that pushed payback periods beyond what most property owners would accept. That math is changing.
A New Category of Demand-Side Solution
Recent advances in building-envelope technology are beginning to change the economics of demand reduction. Transparent Insulation, for example, enables existing windows to achieve wall-like thermal performance without the disruption traditionally associated with deep envelope retrofits.
The implications extend beyond individual buildings. Using Department of Energy data, we estimate that retrofitting just 10% of the U.S. commercial building stock with Transparent Insulation could free generation-equivalent capacity sufficient to power approximately 1,000 medium-sized data centers. Technologies that make deep retrofits practical at-scale are what make those kinds of grid-level outcomes conceivable.
For context, LED lighting upgrades freed roughly 2 kilowatt hours per square foot of floor space. Early data from Transparent Insulation installations is tracking closer to 12 kilowatt hours per square foot. That is a different order of magnitude, and it is the difference between a building efficiency measure and a grid infrastructure measure.
For utility planners, the distinction is important. Transparent Insulation eliminates existing load. High-performance glazing holds a conditioned space at temperature, which means that when ambient temperatures spike on the hottest days of the year, the HVAC surge that would otherwise show up as peak demand on the grid does not materialize. Those are precisely the electrons utilities are currently holding in reserve.
The NoMad Tower at 1250 Broadway in Manhattan is an early example of what this looks like at-scale. Measured outcomes at that property are tracking ahead of initial energy reduction projections, with peak demand avoidance emerging as one of the most significant grid-relevant metrics.
Transparent Insulation is not a standalone solution. Duct sealing technologies address a separate but related source of building energy waste, reducing the load that HVAC systems carry through inefficient distribution. Window-mounted heat pump systems perform better when the glazing they are conditioning against has higher thermal resistance. When these technologies are deployed together, their combined grid impact exceeds what any single intervention delivers on its own. That kind of stacking is where demand-side solutions start to look like genuine infrastructure rather than isolated efficiency projects.
What Utility Companies Can Do
Some utility companies are already drawing the right conclusions. Programs that offer serious demand-side rebates for envelope upgrades reflect an understanding that incentivizing a retrofit in a dense urban building can be cheaper on a dollar-per-kilowatt basis than procuring or building new generation capacity. Con Edison has issued individual rebates on single properties where the measured demand reduction justified the investment. They are doing it because the grid math works.
The challenge is that this thinking is not yet uniform across the industry. Utilities leading on demand-side investment are doing so largely in isolation, and the lessons they are learning are not transferring quickly enough to the rest of the sector.
Two shifts would accelerate progress. The first is reframing demand-side incentives as capital allocation decisions rather than customer programs. The relevant question is at what dollar-per-kilowatt threshold incentivizing a retrofit outperforms building or procuring new capacity. In high-density urban markets, where peak demand is greatest and existing building stock is most concentrated, that threshold is increasingly favorable to the demand side.
The second is moving toward incentive structures that reward combined retrofits at a rate that reflects their compounding grid benefit. If a building owner deploys Transparent Insulation alongside duct sealing and a heat pump upgrade, the combined peak demand reduction is greater than the sum of the parts. A rebate structure that recognizes that multiplier effect would accelerate the kind of deployments that deliver the most grid value.
The Mindset Shift
New generation will be part of the answer to AI-driven load growth. There is no version of the next decade where utilities do not need to add capacity.
The argument is that operational flexibility depends on how efficiently you can use what you have, and that means taking the demand side seriously in a way the industry has not been structured to do.
For decades, the honest answer was that demand-side solutions could not deliver impact at grid scale. That shaped infrastructure planning, capital allocation, and the way utility leaders and grid planners approach every capacity decision. It was a reasonable answer given what was available.
The technology landscape has changed. Demand-side solutions are no longer limited to incremental efficiency gains measured at the building level. A new generation of technologies can reduce energy consumption and peak demand at a scale relevant to utility planning.
The utility companies best positioned to navigate the next decade of load growth will be those that view existing buildings not simply as consumers of electricity, but as opportunities to unlock capacity already embedded within the grid. Generation and transmission upgrades will remain essential; however, the most cost-effective megawatt is often the one that never has to be generated.
The grid does not only need more electrons. It needs fewer wasted ones.
—Scott Thomsen is CEO and founder of LuxWall.
