Require Data Centers to Pay Their Fair Share
The Issue
REQUIRE DATA CENTERS TO PAY THEIR FAIR SHARE
Modernize America’s grid without making taxpayers, utility customers, and communities subsidize unchecked data-center growth.
TL;DR — START
America needs data centers. Our hospitals, banks, governments, businesses, cybersecurity systems, cloud platforms, communications networks, scientific research, entertainment, phones, and everyday technology depend on them. I am not anti-data center. My own career depends on them. I work professionally as a Senior Cybersecurity Engineer focused heavily on cloud and endpoint security. I regularly work with Terraform and infrastructure-as-code, and many of the systems I help secure depend on data centers every day.
Before working professionally in cybersecurity, I also worked as an electrician, and I am an amateur radio operator. Those backgrounds heavily influence how I see this issue. Cybersecurity tells me why we need computing infrastructure. Electrical work taught me that electricity does not magically appear at an outlet. Generation, transformers, substations, transmission lines, distribution equipment, switchgear, protection systems, backup systems, and physical infrastructure all have to exist somewhere and be paid for by someone. Amateur radio constantly reminds me that communications technology is still governed by physics. Spectrum and bandwidth are finite. Antennas have limitations. Signals experience path loss and interference. Power has to come from somewhere, and heat has to go somewhere. Technology does not eliminate physical constraints. It changes which constraints we have to manage.
Right now, we are building computing infrastructure faster than many of the physical systems required to support it. Lawrence Berkeley National Laboratory estimates that data centers could account for approximately 11.8% of all U.S. electricity consumption by 2030 under its central scenario, with modeled scenarios ranging from approximately 9.5% to 15.3%.
https://eta.lbl.gov/publications/united-states-data-center-energy-2025
At the same time, America’s electrical grid already needed enormous investment. We were already dealing with aging infrastructure, transportation and industrial electrification, manufacturing growth, transmission constraints, transformer shortages, resilience concerns, storage, increasing electrical demand, physical security, and cybersecurity. The data-center boom did not create those problems, but adding enormous new industrial loads makes solving them much more urgent.
We need to modernize that grid, but ordinary homeowners, small businesses, taxpayers, and utility customers should not be expected to finance extraordinary infrastructure primarily required by some of the wealthiest corporations in the world. If a data center requires a new substation, it should pay its fair share. If it requires new transformers, transmission capacity, generation, storage, pipelines, water infrastructure, wastewater infrastructure, roads, or other major improvements, those costs need to be reflected honestly in the economics of the project.
Federal regulators are already confronting this issue. FERC has opened a proceeding examining how very large new electrical loads should connect to the interstate transmission system and how the costs and risks should be allocated. FERC has also ordered regional grid operators under its jurisdiction to justify or reform their rules for connecting major new loads while protecting ratepayers.
Ohio is already demonstrating that ratepayer protections are possible. The Public Utilities Commission of Ohio approved a special structure for large new data-center customers in AEP Ohio territory requiring substantial financial commitments intended to reduce the risk that existing customers are left paying infrastructure costs created for those developments.
https://puco.ohio.gov/wps/portal/gov/puco/utilities/electricity/resources/data-centers
Congress is moving in this direction too. H.R. 9340, the bipartisan Ratepayer Protection Act, introduced by Republican Rep. Gabe Evans and Democratic Rep. Kathy Castor, would require state regulators to consider protections for non-residential customers requesting 100 megawatts or more. The proposed standard would seek recovery of the full incremental costs of generation, transmission, and distribution upgrades needed to serve those loads and would require financial assurances intended to protect other customers. The House Energy and Commerce Committee advanced the bill by a unanimous 52-0 vote on July 21, 2026.
This petition supports the Ratepayer Protection Act and calls on Congress to go further. Electrical cost shifting is only one part of the issue. Responsible data-center policy must also address speculative capacity reservations, stranded infrastructure costs, grid modernization, transformer shortages, water consumption, residential siting, continuous and low-frequency noise, dedicated generation, tax subsidies, environmental impacts, decommissioning, storage, distributed generation, and long-term resource planning.
This is not a petition to stop building data centers. It is a petition to build them responsibly.
Build the data centers we need. Modernize the American electrical grid. Build a diverse, interconnected, decentralized, resilient energy system capable of supporting the future. Protect taxpayers, ratepayers, water, and communities. Make extraordinary private demand carry extraordinary private responsibility.
TL;DR — END
DETAILED CASE — START
WHY STRONGER RULES ARE NECESSARY
The short version above explains what this petition is asking for. The rest explains why.
My perspective sits at the intersection of cybersecurity, cloud infrastructure, electrical infrastructure, and communications. The lesson those areas share is simple: digital systems do not exist separately from the physical world.
A facility consuming hundreds of megawatts cannot simply be “plugged in” because somebody purchased land and ordered servers. The generation, transmission, substations, transformers, distribution infrastructure, cooling, water, communications, backup systems, and supporting infrastructure all have to exist. If they do not, somebody has to build them.
The real policy question is not whether we need data centers. We do. The question is who should pay for the extraordinary infrastructure required to support them.
When infrastructure broadly benefits the public, public investment makes sense. When extraordinary infrastructure is required primarily because an enormous private industrial customer has arrived, that customer should carry the appropriate share of the cost.
Extraordinary private demand should carry extraordinary private responsibility.
THE SCALE HAS FUNDAMENTALLY CHANGED
This is not another office building being connected to the grid. The scale of modern data-center development is dramatically different.
The Department of Energy reported that U.S. data centers consumed approximately 4.4% of total U.S. electricity in 2023.
Lawrence Berkeley National Laboratory’s updated modeling estimates that by 2030, data centers could consume approximately 11.8% of U.S. electricity under its central scenario, with modeled outcomes ranging from approximately 9.5% to 15.3%.
https://eta.lbl.gov/publications/united-states-data-center-energy-2025
Regardless of exactly where the final number lands, this represents an extraordinary amount of new electrical demand concentrated into a relatively short period. That demand is also not distributed evenly. A giant facility can suddenly create enormous demand inside one utility territory or transmission region, creating needs for generation, transmission, transformers, substations, pipelines, backup generation, water systems, roads, and other infrastructure that did not previously exist.
This is why the economics of a proposed data center cannot stop at “How much does the building cost?” The real question is: What infrastructure must exist throughout the surrounding region for this facility to operate, and who is paying for it?
AMERICA’S ELECTRICAL GRID WAS ALREADY OVERDUE FOR MODERNIZATION
The data-center boom did not create America’s electrical infrastructure problems. Many were already here.
We were already asking difficult questions about how the grid would handle transportation electrification before hyperscale AI facilities became such a major concern. As electric vehicles become more common, energy demand shifts from petroleum infrastructure onto the electrical system. That means charging infrastructure, local distribution capacity, transformers, generation, and load management.
Electric heating and heat pumps create additional demand. Manufacturing creates demand. New housing creates demand. Industrial electrification creates demand. Economic growth creates demand. Aging infrastructure needs replacement regardless of whether a data center ever arrives.
Those were legitimate planning challenges before the current data-center boom. Now we are attempting to address them while simultaneously connecting facilities whose electrical demands can be measured in hundreds or even thousands of megawatts.
That does not mean we should stop electrifying transportation, stop building homes and factories, or stop building data centers. It means we need to acknowledge reality: the United States needs a major modernization of its electrical infrastructure.
We need additional generation, stronger transmission, more substations, more transformers, modernized distribution systems, better regional interconnection, storage, modern monitoring and control systems, physical security, cybersecurity, redundancy, and greater resilience against severe weather and other emergencies. We need to stop waiting for infrastructure to fail before replacing or expanding it.
THE GRID EQUIPMENT ITSELF IS BECOMING A BOTTLENECK
It is not only a question of generating enough electricity. The physical equipment needed to move and transform that electricity has its own supply constraints.
Transformers, switchgear, breakers, conductors, generators, and other heavy electrical equipment cannot always be ordered today and delivered next month. Some equipment has manufacturing lead times measured in years.
Reuters reported in July 2026 that surging data-center demand is contributing to shortages of critical grid equipment, with some transformer lead times extending beyond 160 weeks.
That matters far beyond data centers. Utilities need transformers. Factories need them. Neighborhoods need them. Hospitals need them. Schools need them. Housing developments need them. Critical infrastructure needs them. Communities rebuilding after disasters need them.
When extraordinarily wealthy companies can reserve enormous quantities of grid equipment years in advance, everyone else is competing inside the same constrained supply chain. Infrastructure planning needs to occur before giant projects advance.
MODERNIZE THE GRID, AND MAKE DATA CENTERS PART OF THE INVESTMENT
The data-center boom could make America’s grid problems dramatically worse, or it could help solve them.
There is an unprecedented amount of private capital flowing into computing infrastructure. America already needs major electrical infrastructure investment. Those two facts should be connected.
If data-center developers want extraordinary access to America’s electrical system, some of that investment should help expand and modernize the system necessary to support them. If a data center requires a new substation, it should pay its appropriate share. If it requires additional transmission, the incremental costs associated with that demand should be reflected in the project. The same principle should apply to transformers, generation, storage, pipelines, water systems, wastewater systems, roads, and other major infrastructure.
There is a difference between replacing infrastructure that already broadly serves the public and constructing extraordinary infrastructure primarily because one giant private customer has arrived.
If a utility replaces an aging transformer serving an existing neighborhood, that is normal utility investment. If one hyperscale campus suddenly requires a completely new substation, major transmission work, multiple transformers, and hundreds of megawatts of additional capacity, that is different.
The incremental cost should follow the incremental demand.
Data centers should become part of America’s grid-modernization investment, not another reason ordinary Americans pay more for electricity.
PASS THE RATEPAYER PROTECTION ACT AND GO FURTHER
This idea is not starting from zero.
H.R. 9340, the Ratepayer Protection Act, was introduced by Republican Rep. Gabe Evans of Colorado and Democratic Rep. Kathy Castor of Florida. On July 21, 2026, the House Committee on Energy and Commerce advanced it by a unanimous 52-0 vote.
The legislation would require state regulators to consider a large-load standard for non-residential customers with a peak demand of 100 megawatts or more. The proposed standard would seek recovery of the full incremental costs of generation, transmission, and distribution upgrades necessary to serve the load while also requiring financial assurances intended to protect other customers.
That is exactly the kind of policy direction this petition supports.
Congress should pass the Ratepayer Protection Act, but it should treat it as a floor, not the finish line.
Ratepayer protection addresses one of the most urgent parts of this problem, but responsible data-center policy must also address grid modernization, speculative capacity reservations, stranded investments, equipment shortages, water use, zoning, continuous and low-frequency noise, subsidies, environmental impacts, storage, distributed generation, on-site generation, and decommissioning.
Protecting ratepayers is the starting point. Making data-center development sustainable for the electrical grid and the communities hosting it is the larger goal.
PROTECT RATEPAYERS FROM STRANDED INFRASTRUCTURE COSTS
What happens if a utility builds infrastructure and the data center never shows up?
A proposed data center can influence years of planning. Utilities may order transformers, construct substations, expand transmission, reserve generation, acquire land, and make other long-term investments based on expected demand.
But markets change. Companies change strategies. AI architectures change. Projects get delayed, downsized, relocated, or canceled.
Utilities and regulators therefore need to ensure that companies requesting enormous amounts of capacity make meaningful financial commitments before the public takes on the risk. That can include long-term contracts, minimum-demand commitments, deposits or other financial assurances where warranted, and exit obligations if a project triggers major infrastructure investment and then disappears.
Ordinary customers should not become involuntary investors in speculative data-center development.
STOP SPECULATIVE CAPACITY HOARDING
Electrical capacity itself has value. A developer asking for 500 megawatts, one gigawatt, or more is not simply filling out another service application. Those requests affect utility forecasts, transmission planning, generation decisions, infrastructure reservations, and other customers waiting for connections.
FERC has specifically raised concerns around how major new loads are forecast, connected, and financially committed.
Developers requesting extraordinary quantities of electricity should demonstrate real site control, financing, credible construction schedules, realistic demand estimates, utility plans, water plans, and meaningful financial commitment.
Scarce electrical capacity should not become something companies can casually reserve across multiple speculative sites while deciding later which project they actually intend to build.
INFRASTRUCTURE BEFORE HARDWARE
There is something fundamentally backwards about acquiring computing hardware faster than the infrastructure required to operate it can be built.
Companies are competing for GPUs, memory, storage, servers, networking equipment, switchgear, transformers, generators, and cooling infrastructure while electrical equipment can take years to procure.
The electrical planning has to come first. The transmission planning has to come first. The water planning has to come first. Projects need to demonstrate that they are technically viable as complete systems, not simply that somebody can afford the servers.
When enormous quantities of computing hardware are being purchased or committed to while adequate power infrastructure still does not exist, something about the planning process has gone wrong.
We should not be creating data centers full of hardware with no power available to run the hardware.
CONSUMERS ARE GETTING HIT THROUGH THE TECHNOLOGY SUPPLY CHAIN TOO
The resource competition extends beyond electricity.
The AI data-center buildout has created extraordinary demand for memory, storage, GPUs, and other components. In June 2026, industry groups representing automakers, retailers, electronics companies, and other sectors warned that AI data-center demand for memory chips was contributing to shortages and potentially significant consumer price increases.
Later that month, Apple raised prices on MacBooks and iPads as memory and storage costs surged amid the AI infrastructure boom.
The point is not that every increase in every phone or computer can be blamed directly on data centers. The point is that massive AI infrastructure demand is competing in the same supply chains used by consumers, schools, businesses, corporate IT departments, researchers, developers, repair shops, homelab users, electronics manufacturers, and automakers.
There is something deeply backwards about purchasing enormous amounts of computing hardware faster than the supporting electrical infrastructure can be constructed while everyone else simultaneously competes for increasingly constrained components.
Innovation should make computing more accessible. We should question an infrastructure arms race when it starts accomplishing the opposite.
MY OWN EXPERIENCE IS WHY COST SHIFTING MATTERS TO ME
This is not only an abstract policy concern for me.
Over the past several years, my own household electric bill has increased by more than 50%. During roughly that same period, efficiency improvements I made reduced my average electricity consumption by approximately 25%.
I am not claiming that data centers alone caused that increase in my particular bill. Electricity rates are complicated. Fuel prices, generation, transmission, storm restoration, capital projects, regulation, inflation, and many other factors affect rates.
But my experience is exactly why I care about what happens next.
Consumers are constantly told to conserve energy, buy efficient appliances, improve insulation, replace lighting, upgrade HVAC systems, manage thermostats, and reduce consumption. Many of us have spent our own money doing exactly that.
Ordinary customers should not make those investments only to have the financial benefits swallowed by infrastructure costs created by enormous new private industrial loads.
If a project creates extraordinary new demand, the costs specifically associated with that demand should primarily follow the project.
“JUST BUILD MORE RENEWABLES” IS NOT A COMPLETE GRID PLAN
Renewable energy absolutely belongs in America’s future energy strategy. But saying “just build more renewables” is not a complete engineering answer.
Generation is only one component of an electrical system. Solar generates electricity. Wind generates electricity. Nuclear generates electricity. Geothermal generates electricity. Hydroelectric facilities generate electricity. Natural-gas plants generate electricity.
Regardless of generation source, electricity still has to be generated, balanced, transmitted, distributed, protected, controlled, stored when appropriate, and delivered reliably when customers need it.
The International Energy Agency notes that as variable renewable resources such as solar and wind become larger portions of generation, power systems need greater flexibility through tools such as storage, flexible generation, flexible demand, stronger grids, and improved control systems.
https://www.iea.org/reports/electricity-grids-and-secure-energy-transitions/executive-summary
https://www.iea.org/reports/electricity-2026/flexibility
The argument is not that renewable energy cannot work. The stronger and more accurate point is that variable renewable generation alone is not a complete grid architecture.
THERE IS NO SINGLE MAGIC ENERGY SOURCE
America should stop looking for one technological savior.
Solar is modular, distributed, and increasingly affordable, but output varies by time, weather, season, and geography. Wind has enormous potential but is similarly location- and weather-dependent. Hydroelectricity provides valuable generation and flexibility but depends on geography and hydrology. Geothermal can provide firm generation, and emerging enhanced-geothermal technologies may expand where it is viable.
Nuclear energy can provide substantial firm generation but carries capital costs, construction timelines, regulatory requirements, fuel-cycle considerations, and waste-management responsibilities. Natural gas can provide dispatchable power but brings fuel-price, infrastructure, emissions, and environmental considerations. Storage can solve important balancing and resilience problems, but different storage technologies serve different timescales and have different costs and material requirements.
Research into advanced reactor designs, alternative nuclear fuel cycles including thorium-based approaches, enhanced geothermal systems, fusion, long-duration storage, and other emerging technologies should continue.
No technology should be treated as a magic word that makes engineering constraints disappear.
The goal should be a diverse, affordable, resilient, maintainable, scalable, and secure electrical system.
URUGUAY SHOWS WHAT DELIBERATE LONG-TERM ENERGY PLANNING CAN ACCOMPLISH
The United States does not have to invent every policy lesson from scratch.
According to the U.S. International Trade Administration, Uruguay generated approximately 99% of its electricity from renewable sources in 2024, primarily using hydropower, wind, and biomass, with solar also contributing.
https://www.trade.gov/country-commercial-guides/uruguay-energy
The lesson is not that America should copy Uruguay’s exact energy mix. Our geography, population, industries, resources, and electrical demand are dramatically different.
The useful lesson is: have an energy strategy instead of an energy patchwork.
Uruguay made a deliberate long-term transition involving multiple generation technologies and coordinated policy. America needs the same willingness to decide what kind of electrical system we want decades from now and deliberately build toward it.
A MODERN GRID SHOULD BE STRONGER NATIONALLY AND MORE DECENTRALIZED LOCALLY
The traditional electrical model relies heavily on large centralized generating facilities, long-distance transmission, substations, distribution systems, and customers at the edge.
That architecture is not going away, nor should it. America needs more transmission capacity between regions, not less.
But centralized infrastructure does not need to be our only tool. We should also expand distributed generation, rooftop solar, community solar, commercial solar, home and community storage, microgrids, virtual power plants, flexible demand, and other distributed energy resources.
The Department of Energy describes microgrids as localized systems capable of coordinating generation, storage, and loads and, when appropriately designed, disconnecting from the larger grid to operate independently.
https://www.energy.gov/oe/microgrid-systems
A twenty-first-century electrical system can therefore be more interconnected nationally while simultaneously becoming more decentralized locally.
Those are not contradictory ideas. Strong regional transmission allows areas to support one another. Distributed local resources reduce strain on distant generation and transmission. Microgrids provide resilience when parts of the larger system fail.
THINK OF THE GRID AS A RESILIENT NETWORK
My cybersecurity background strongly influences how I think about this.
In cybersecurity and network engineering, unnecessary single points of failure are bad architecture. We design redundant systems, alternate routes, failover, segmentation, distributed services, local resilience, backups, and recovery plans.
Our energy system can benefit from similar thinking.
Imagine communities with local storage, hospitals operating resilient microgrids, emergency facilities with local generation, municipal buildings capable of supporting critical services during outages, homes with distributed solar and batteries, businesses capable of reducing demand during grid emergencies, and data centers capable of supporting or relieving the grid instead of only consuming from it.
Then connect those systems through strong regional infrastructure capable of providing support when local resources are insufficient.
The goal is not to disconnect everybody from the grid. The goal is to build a stronger network where more parts of the system can support one another and fewer failures cascade unnecessarily.
STORAGE SHOULD BECOME INFRASTRUCTURE, NOT AN AFTERTHOUGHT
Energy storage needs to become a normal part of electrical planning, and storage should not mean building one enormous battery somewhere.
We should increasingly think about cells of storage capacity throughout the grid: home batteries, neighborhood and community storage, commercial and industrial systems, data-center storage, microgrid storage, utility-scale batteries, thermal storage, pumped hydroelectric storage where geography permits, vehicle-to-grid systems where technically appropriate, long-duration storage, and emerging technologies as they mature.
Storage can absorb electricity when generation exceeds immediate demand, make energy available during demand peaks, support local systems during outages, provide grid services, help integrate variable generation, reduce short-duration peaks, and provide another layer of resilience.
Different technologies will fit different applications. The point is not to choose one battery chemistry. The point is to treat storage as a normal component of grid infrastructure.
DATA CENTERS SHOULD BECOME GRID PARTICIPANTS, NOT SIMPLY GIANT LOADS
At a certain scale, a data center should no longer be treated like an ordinary customer plugging another appliance into the wall.
If a facility consumes electricity on the scale of a city, it has become part of the electrical system whether policymakers acknowledge that or not.
The question should not only be “How quickly can we connect you?” It should also be “What are you contributing to the system you are asking to use?”
Can the facility build or finance storage? Can it operate a microgrid? Can it provide demand flexibility? Can non-time-sensitive workloads shift away from extreme grid peaks? Can some backup resources provide grid support where technically and environmentally appropriate? Can the project finance generation or transmission? Can excess heat be reused where practical? Can infrastructure built for the project also benefit the surrounding community?
If companies want extraordinary access to America’s electrical infrastructure, we should expect extraordinary participation in strengthening it.
MAKE SMALL-SCALE ENERGY GENERATION EASIER FOR ORDINARY AMERICANS TOO
If America desperately needs more generation and distributed resources, why is it still so difficult and expensive for ordinary people to generate even modest amounts themselves?
Traditional rooftop solar can involve major equipment costs, professional installation, permitting, engineering, utility interconnection, inspections, financing, contractor overhead, and substantial administrative complexity. For many households, the realistic choices become spending tens of thousands of dollars or doing nothing.
There should be more options in between.
Germany’s federal network regulator reported approximately 430,000 new plug-in balcony solar installations in 2025 alone, adding roughly 0.5 gigawatts of capacity.
https://www.bundesnetzagentur.de/SharedDocs/Pressemitteilungen/EN/2026/20260108_EEG.html
America should examine how we can create similarly accessible options.
Because I used to work as an electrician, I want to be very clear: this does not mean allowing people to plug random homemade solar equipment into outlets. Anti-islanding protection matters. Overcurrent protection matters. Equipment certification matters. Grounding matters. Conductor ratings matter. Connector design matters. Utility-worker safety matters.
But safe does not have to mean needlessly inaccessible.
Federal regulators, standards organizations, manufacturers, states, utilities, and code bodies should establish a clear pathway for appropriately engineered, certified, current-limited, protected small-scale plug-in solar and storage systems.
Renters should have options. Apartment residents should have options. Low-income households should have options. Homeowners should have inexpensive options. Small businesses should have options.
A few hundred watts from one balcony will not power a data center. That is not the point. A few hundred watts across millions of buildings becomes meaningful distributed generation.
WATER CANNOT BE TREATED AS AN UNLIMITED INDUSTRIAL RESOURCE
Electricity is not the only major resource involved.
Depending on design, cooling technology, and climate, data centers can consume substantial quantities of water. DOE states that data-center water consumption varies widely based on cooling technology and local conditions but can reach millions of gallons per day, while newer cooling approaches can dramatically reduce water requirements.
The argument should not be that every data center wastes millions of gallons of water. That would be inaccurate.
The stronger argument is that water use depends heavily on engineering decisions, so responsible engineering and disclosure should be mandatory.
Projects should disclose expected water demand before approval and actual consumption once operating. Developers should identify whether the source is municipal drinking water, groundwater, surface water, reclaimed wastewater, or another supply. Closed-loop and lower-water cooling should be prioritized where technically appropriate, and reclaimed or non-potable water should be considered where feasible.
Facilities proposed in drought-prone regions, areas dependent on stressed aquifers, or communities with limited water capacity deserve additional scrutiny.
The question should not simply be “Can the company afford this water?” It should be “Can the community sustainably provide this water?”
A corporation being able to purchase millions of gallons does not create millions of gallons of additional water.
If a project requires additional wells, treatment facilities, pipes, pumps, sewer capacity, or wastewater systems, those costs should be included in the economics of the project.
DATA CENTERS ARE INDUSTRIAL INFRASTRUCTURE. ZONE THEM ACCORDINGLY.
A hyperscale data center is not an office building. It is industrial infrastructure.
Cooling systems can operate continuously. Transformers operate continuously. Fans and pumps operate. Backup generators require testing. Electrical systems create sound. Construction and expansion can continue for years.
Virginia’s Joint Legislative Audit and Review Commission found that constant low-frequency data-center noise has created problems when facilities are located near residential areas and noted that conventional noise rules do not always address this type of sound well.
https://rga.lis.virginia.gov/Published/2025/RD206
That is exactly why siting matters.
Large data centers should be subject to appropriate industrial zoning, meaningful setbacks from homes, buffers, independent acoustic modeling, low-frequency noise analysis, nighttime standards, post-construction testing, generator-noise analysis, environmental review, and meaningful enforcement.
Residential zoning should mean something. A family should not buy or rent a home in an established residential community only to discover that their neighborhood has effectively become an industrial district because a developer found enough acreage nearby.
CONTINUOUS AND LOW-FREQUENCY NOISE DESERVES SPECIAL TREATMENT
Traditional noise ordinances often address parties, vehicles, temporary construction, lawn equipment, dogs, and other intermittent sound.
A data center can produce something different: continuous, relatively stable, low-frequency or tonal sound twenty-four hours a day.
Virginia’s JLARC found that residents can experience quality-of-life problems even where data-center sound does not necessarily violate conventional noise ordinances.
https://rga.lis.virginia.gov/Published/2025/RD206
Policymakers therefore need measurement standards appropriate for the source. Rules should consider continuous sound, low-frequency components, tonal characteristics, nighttime background levels, cumulative noise from multiple buildings, transformers, cooling systems, generators, and continued campus expansion.
STOP GIVING BLANK CHECKS TO SOME OF THE WEALTHIEST COMPANIES ON EARTH
According to the National Conference of State Legislatures, 38 states offer dedicated tax incentives for data centers, including sales-tax exemptions, use-tax exemptions, property-tax abatements, and other programs.
https://www.ncsl.org/fiscal/subsidizing-servers-how-states-are-competing-to-attract-data-centers
Economic-development incentives are not automatically bad. There may be circumstances where incentives produce genuine public benefit.
But they should not become an automatic bidding war where communities compete to offer increasingly favorable treatment to companies already capable of financing enormous projects themselves.
Any major subsidy should include a public cost-benefit analysis, clear employment and investment commitments, transparent infrastructure costs, expiration dates, periodic review, measurable performance requirements, and enforceable clawbacks if promised benefits do not materialize.
A data center can represent billions of dollars of capital investment while employing significantly fewer permanent workers than many other forms of industrial development. That does not mean data centers provide no economic benefit. It means the benefits need to be evaluated honestly.
If a project receives major tax breaks while ratepayers finance grid expansion and taxpayers finance supporting infrastructure, public officials should demonstrate why the arrangement actually benefits the public.
Private investment should not mean private profit and public infrastructure bills.
ON-SITE GENERATION CANNOT BECOME A LOOPHOLE
Some developers are responding to limited grid capacity by considering or constructing dedicated natural-gas turbines, reciprocating engines, generator installations, pipelines, and islanded power systems.
The issue is not simply that natural gas exists. The issue is whether moving generation behind the fence becomes a way to bypass responsibilities that would otherwise apply.
EPA maintains specific Clean Air Act resources for data centers and relevant stationary sources.
https://www.epa.gov/stationary-sources-air-pollution/clean-air-act-resources-data-centers
If a data center needs dedicated generation, the full system still needs to be accounted for: pipelines, fuel, emissions controls, noise, equipment, maintenance, safety, environmental impacts, and eventual decommissioning.
Running out of grid capacity should not mean creating an entirely separate power system and pretending its impacts do not exist.
DATA-CENTER DEVELOPERS SHOULD HAVE DECOMMISSIONING RESPONSIBILITIES
Computing infrastructure changes quickly. Facilities become obsolete. Companies consolidate. Buildings become economically undesirable. Equipment reaches the end of its useful life.
Communities should not be left with abandoned industrial sites and expensive infrastructure built around facilities that no longer operate.
Large projects should therefore have credible decommissioning and remediation plans, potentially including financial assurance appropriate to the size and environmental risks of the project.
If specialized pipelines, electrical infrastructure, cooling systems, fuel systems, batteries, industrial equipment, or contaminated materials need to be removed or remediated later, taxpayers should not automatically inherit those costs.
The company benefiting from the facility during its useful life should bear responsibility for responsibly closing it too.
PUTTING DATA CENTERS IN SPACE DOES NOT MAKE PHYSICS DISAPPEAR
There is growing interest in orbital computing and the idea of moving substantial computing infrastructure into space.
Space computing may eventually make sense for specialized applications. I am not arguing that research should stop.
But “put the data centers in space” is not a magic escape hatch from terrestrial infrastructure constraints.
As an amateur radio operator, the communications problem is one of the first things that jumps out at me.
A computer in orbit still needs to communicate. You still need spectrum, antennas, ground infrastructure, uplinks, downlinks, potentially inter-satellite links, link budgets, redundancy, and enormous amounts of network capacity. You still have path loss, interference, finite bandwidth, latency, orbital geometry, and handoffs.
Moving enormous amounts of computation into orbit means figuring out how enormous amounts of data move into, out of, and between those systems.
Physics does not disappear because the server rack is above the atmosphere.
SPACE IS COLD, BUT COOLING IN SPACE IS NOT SIMPLE
One of the most misleading intuitive ideas surrounding orbital data centers is “Space is cold, so cooling will be easy.”
Vacuum does not work like cold air.
On Earth, data centers can move heat through air and liquid systems and reject that heat into the environment. A spacecraft cannot use the surrounding vacuum as an enormous convection cooling system.
Waste heat has to be transported through the spacecraft and ultimately rejected primarily through radiation. NASA treats thermal control and radiative heat rejection as fundamental spacecraft engineering requirements.
https://www.nasa.gov/smallsat-institute/sst-soa/thermal-control/
That does not mean orbital computing is physically impossible. It means moving computation into space trades one thermal problem for another.
Engineering still has to close the thermal budget.
MAINTENANCE AND HARDWARE REPLACEMENT BECOME RADICALLY HARDER
On Earth, a failed server can be pulled from a rack and repaired or replaced. GPUs can be upgraded. SSDs and RAM can be replaced. Networking equipment can be upgraded. Cooling systems can be modified. Obsolete equipment can be removed and recycled.
Space changes that equation dramatically.
Every piece of hardware must remain reliable for its intended lifetime, have redundancy, be remotely or robotically serviceable, or eventually be replaced by launching new equipment.
Computing hardware also becomes obsolete quickly. A GPU that makes economic sense today may be considerably less attractive several years from now.
That creates a fundamental lifecycle question: How frequently will orbital data centers need to be replaced to remain economically competitive, and what happens to the old hardware?
WE RISK TURNING E-WASTE INTO AEROSPACE WASTE
Terrestrial computing already creates an enormous electronic-waste problem. Moving large amounts of rapidly depreciating computing hardware into orbit adds another layer.
If the operating model becomes launch, use, deorbit, replace, repeat, then environmental accounting needs to include the entire lifecycle: manufacturing, spacecraft materials, launch vehicles, propellants, launch effects, replacement hardware, reentry, and disposal.
That is not an argument to ban research. It is an argument to count all of the costs before presenting space as an easy environmental solution.
ORBITAL DEBRIS IS ALREADY A REAL REGULATORY ISSUE
Orbit is not an infinite garbage dump.
The FCC already maintains orbital-debris mitigation requirements and end-of-life rules for satellite operators.
https://www.fcc.gov/document/fcc-adopts-new-5-year-rule-deorbiting-satellites-0
This matters tremendously if orbital computing eventually involves very large constellations. More objects mean more collision-avoidance requirements, tracking requirements, failure scenarios, debris risks, replacement missions, and end-of-life operations.
The lifecycle has to be part of the economics from the beginning.
THE NIGHT SKY AND RADIO ASTRONOMY MATTER TOO
There is also a public resource that rarely appears on a corporate balance sheet: the night sky and radio environment.
Large satellite constellations already create concerns for optical astronomy and radio astronomy. The International Astronomical Union’s Centre for the Protection of the Dark and Quiet Sky works specifically on the impacts of large constellations.
Researchers have also detected unintended electromagnetic radiation from large satellite constellations affecting radio-astronomy observations.
As a ham radio operator, this aspect particularly concerns me. Spectrum and the radio environment are shared resources.
We should not wait until enormous orbital computing infrastructure exists before asking how it affects communications, astronomy, spectrum management, interference, orbital congestion, and the night sky.
Those questions should be addressed before deployment at enormous scale.
SPACE COMPUTING MAY HAVE A FUTURE, BUT IT REQUIRES FULL LIFECYCLE ACCOUNTING
Specialized orbital computing may eventually make sense. Processing space-generated sensor data closer to where it originates may make sense. Some workloads may make sense in orbit. Research should continue.
But orbital data centers should be evaluated under the same principle as terrestrial data centers: account for the complete system, not merely the server.
For orbital computing, that means power generation, storage, thermal rejection, launch requirements, spacecraft mass, spectrum, ground stations, inter-satellite communication, space-to-ground bandwidth, latency, radiation, hardware lifetime, servicing, redundancy, replacement cycles, debris, collision avoidance, reentry, astronomy, radio interference, manufacturing impacts, and end-of-life disposal.
You cannot simply move a server into space and declare that its terrestrial environmental footprint disappeared.
You cannot launch bad infrastructure policy into orbit and call the problem solved.
THE LARGER OPPORTUNITY
The frustrating thing about all of this is that the data-center boom could actually become an enormous opportunity.
America already needs major electrical infrastructure investment. We need more transmission, generation, transformers, substations, storage, regional interconnection, resilient distribution, sophisticated control systems, distributed resources, cybersecurity, physical resilience, and emergency capability.
There is also an unprecedented amount of private money flowing into computing infrastructure.
That investment could help accelerate the modernization we already needed.
Instead of asking only, “How do we find enough electricity for data centers?” we should also ask, “How do we use this investment to strengthen the electrical system for everybody?”
A transmission project required by a data-center campus might also increase regional capacity. A new substation might account for future community growth. A giant facility could finance storage. A data center could operate as a microgrid. Flexible computing workloads could participate in demand-response programs. Projects could finance additional generation. Communities could receive resilience improvements.
Data-center investment could also help expand domestic manufacturing of transformers, switchgear, conductors, and other critical electrical equipment.
There is an opportunity to turn an infrastructure crisis into an infrastructure modernization effort, but that only happens if policy requires it.
WHAT WE ARE ASKING POLICYMAKERS TO DO
DEMAND 01: PASS THE RATEPAYER PROTECTION ACT AND BUILD UPON IT. Congress should pass H.R. 9340, the Ratepayer Protection Act, and establish strong protections ensuring that extraordinarily large electrical customers pay the incremental generation, transmission, distribution, and other infrastructure costs created by their demand.
DEMAND 02: PROTECT RESIDENTIAL AND SMALL-BUSINESS RATEPAYERS FROM COST SHIFTING. Utilities and regulators should use appropriate large-load tariffs, minimum-demand commitments, long-term contracts, and financial assurances so existing customers are not forced to subsidize infrastructure built primarily for giant private loads.
DEMAND 03: PROTECT CUSTOMERS FROM STRANDED INFRASTRUCTURE COSTS. If a data-center developer causes a utility to invest in major infrastructure and later cancels, relocates, delays, or substantially downsizes the project, the developer should remain financially responsible for the appropriate costs.
DEMAND 04: REQUIRE PROJECT READINESS BEFORE MASSIVE CAPACITY RESERVATIONS. Developers seeking extraordinary electrical capacity should demonstrate site control, financing, realistic demand estimates, credible construction schedules, utility plans, and meaningful financial commitment before tying up scarce grid resources.
DEMAND 05: STOP SPECULATIVE CAPACITY HOARDING. Companies should not be allowed to reserve enormous quantities of electrical capacity indefinitely or across multiple speculative sites without meaningful financial consequences.
DEMAND 06: CREATE A NATIONAL GRID-MODERNIZATION STRATEGY. America should accelerate investment in generation, transmission, substations, transformers, distribution systems, regional interconnection, storage, grid-enhancing technologies, cybersecurity, physical security, monitoring, controls, and resilience.
DEMAND 07: MAKE DATA-CENTER INVESTMENT PART OF GRID MODERNIZATION. The larger the new electrical load, the greater the project’s responsibility should be for financing infrastructure required to accommodate it.
DEMAND 08: PURSUE A DIVERSE, TECHNOLOGY-NEUTRAL ENERGY STRATEGY. Solar, wind, geothermal, hydroelectricity, nuclear energy, storage, emerging reactor technologies, alternative fuel cycles, and other resources should be evaluated based on reliability, safety, environmental effects, cost, geography, scalability, construction time, and engineering reality.
DEMAND 09: EXPAND DISTRIBUTED GENERATION AND MICROGRIDS. Communities, hospitals, campuses, businesses, critical facilities, and neighborhoods should have greater ability to generate, store, and manage electricity locally while remaining connected to a stronger regional grid.
DEMAND 10: BUILD DISTRIBUTED CELLS OF ENERGY STORAGE. Home, community, commercial, industrial, microgrid, utility-scale, thermal, pumped-hydro, vehicle-to-grid, long-duration, and emerging storage technologies should be encouraged where technically appropriate.
DEMAND 11: TREAT ENORMOUS DATA CENTERS AS ACTIVE GRID PARTICIPANTS. Regulators should evaluate storage, demand flexibility, microgrid capability, generation investment, grid-support services, and infrastructure contributions proportional to the scale of a project’s demand.
DEMAND 12: CREATE A SAFE PATHWAY FOR INEXPENSIVE SMALL-SCALE PLUG-IN SOLAR AND STORAGE. Federal regulators, standards organizations, utilities, manufacturers, states, and electrical-code bodies should establish safe standards allowing appropriately certified small systems to be deployed with significantly less cost and bureaucracy than full rooftop installations.
DEMAND 13: PLAN DATA CENTERS AND OTHER LARGE NEW LOADS TOGETHER. Data centers, transportation electrification, industrial electrification, housing growth, manufacturing, and other major electrical demands compete for the same generation, transmission, transformers, equipment, and labor and should not be planned as unrelated problems.
DEMAND 14: REQUIRE WATER-USE TRANSPARENCY AND SUSTAINABLE WATER PLANNING. Developers should disclose expected and actual consumption, identify water sources, demonstrate long-term availability, prioritize efficient cooling and reclaimed water where practical, and pay for infrastructure their demand requires.
DEMAND 15: PROTECT DRINKING WATER AND GROUNDWATER. Projects in drought-prone regions, areas with stressed aquifers, or communities with limited municipal capacity should receive additional scrutiny.
DEMAND 16: KEEP NEW HYPERSCALE DATA CENTERS OUT OF ESTABLISHED RESIDENTIAL NEIGHBORHOODS. Local governments should require appropriate industrial zoning, setbacks, buffers, environmental review, acoustic analysis, and meaningful public participation.
DEMAND 17: ESTABLISH ENFORCEABLE DATA-CENTER NOISE STANDARDS. Rules should account for continuous sound, low-frequency and tonal noise, cooling equipment, transformers, backup generation, nighttime conditions, cumulative impacts, and post-construction verification.
DEMAND 18: END BLANK-CHECK SUBSIDIES. Data-center tax incentives should require transparent evidence of genuine net public benefit, measurable commitments, expiration dates, periodic review, and enforceable clawbacks.
DEMAND 19: REQUIRE PUBLIC DISCLOSURE OF MAJOR RESOURCE COMMITMENTS AND INCENTIVES. Communities deserve to know anticipated electricity demand, water consumption, generation plans, emissions, infrastructure requirements, tax incentives, and other public commitments before projects are approved.
DEMAND 20: DO NOT ALLOW ON-SITE GENERATION TO BECOME A REGULATORY LOOPHOLE. Dedicated natural-gas or other power systems should remain subject to appropriate environmental, air-quality, safety, noise, public-review, infrastructure-cost, and decommissioning requirements.
DEMAND 21: REQUIRE DEVELOPERS TO FUND DECOMMISSIONING AND REMEDIATION. Communities should not inherit abandoned facilities, obsolete infrastructure, contaminated property, or cleanup expenses when data centers close.
DEMAND 22: REQUIRE FULL LIFECYCLE ANALYSIS FOR ORBITAL DATA CENTERS. Any proposal to move computing infrastructure into orbit at large scale should account for communications, spectrum, ground infrastructure, thermal management, power, radiation, servicing, replacement, launches, orbital debris, astronomy, radio interference, reentry, and end-of-life disposal.
THIS IS NOT ANTI-TECHNOLOGY. IT IS RESPONSIBLE TECHNOLOGY.
I believe in cloud computing, cybersecurity, artificial intelligence research, scientific computing, modern communications infrastructure, electrification, renewable energy, nuclear research, energy storage, microgrids, distributed generation, and exploring technologies we have not perfected yet.
I believe in innovation.
But innovation is not a magic word that makes engineering constraints disappear.
My cybersecurity background tells me why we need this infrastructure. My electrical background tells me that infrastructure has physical requirements. My amateur-radio experience reminds me that even our most advanced communications networks are still governed by physics.
Power has to come from somewhere. Heat has to go somewhere. Water has to come from somewhere. Signals have to travel somehow. Equipment has to be manufactured. Infrastructure has to be maintained. Somebody ultimately has to pay for all of it.
The question is whether those costs are honestly included in the economics of a project or quietly transferred onto everybody else.
If your project requires a new substation, help pay for the substation. If it requires new transmission, help pay for the transmission. If it requires additional generation, help pay for the generation. If it requires storage, invest in storage. If it requires new transformers, account for them before construction. If it requires new water infrastructure, pay for the infrastructure your project requires. If it needs millions of gallons of water, demonstrate that the community can sustainably provide it. If it creates industrial noise, do not put it next to people’s homes. If it requires new pipelines or on-site generation, account for those costs and impacts. If it asks for hundreds of megawatts of electricity, prove the capacity exists or finance the infrastructure necessary to create it.
If utilities build major infrastructure because of your project and you later abandon that project, do not leave ordinary utility customers with the bill.
And if a project’s economics only work because taxpayers subsidize it while ordinary ratepayers absorb its infrastructure costs, then the project is not actually paying for itself. The public is simply paying part of the bill.
WE CAN BUILD BOTH FUTURES
America needs data centers. America also needs a dramatically better electrical grid. Those goals do not have to compete.
We can modernize transmission, expand generation, develop a diverse energy portfolio, build storage throughout the grid, establish resilient microgrids, expand distributed generation, make small-scale solar more accessible, strengthen local distribution, improve regional interconnection, invest in grid cybersecurity, and build infrastructure capable of supporting electric vehicles, manufacturing, housing, businesses, critical infrastructure, and data centers together.
And we can require companies creating extraordinary new demand to become part of the investment required to make that possible.
The answer is not “No data centers.”
The answer is responsible data centers.
Plan them before building them. Power them before filling them with hardware. Account for their water. Account for their noise. Account for their environmental impacts. Put industrial facilities in appropriate locations. Stop speculative projects from monopolizing scarce grid resources. Stop giving blank checks to some of the world’s wealthiest corporations. Protect residential and small-business ratepayers. Pass the Ratepayer Protection Act and build upon it. Modernize the grid. Build storage. Build local resilience. Strengthen national transmission. Support diverse generation. Continue energy research. Make distributed energy easier for ordinary Americans. And make private companies pay the fair cost of the extraordinary private demand they create.
Build the data centers we need. Modernize the American electrical grid. Build a diverse, interconnected, decentralized, resilient energy system capable of supporting the future. Protect our water. Protect our communities. Protect ordinary taxpayers and ratepayers. And make the companies creating extraordinary new demand pay their fair share of what it takes to support it.
DETAILED CASE — END
32
The Issue
REQUIRE DATA CENTERS TO PAY THEIR FAIR SHARE
Modernize America’s grid without making taxpayers, utility customers, and communities subsidize unchecked data-center growth.
TL;DR — START
America needs data centers. Our hospitals, banks, governments, businesses, cybersecurity systems, cloud platforms, communications networks, scientific research, entertainment, phones, and everyday technology depend on them. I am not anti-data center. My own career depends on them. I work professionally as a Senior Cybersecurity Engineer focused heavily on cloud and endpoint security. I regularly work with Terraform and infrastructure-as-code, and many of the systems I help secure depend on data centers every day.
Before working professionally in cybersecurity, I also worked as an electrician, and I am an amateur radio operator. Those backgrounds heavily influence how I see this issue. Cybersecurity tells me why we need computing infrastructure. Electrical work taught me that electricity does not magically appear at an outlet. Generation, transformers, substations, transmission lines, distribution equipment, switchgear, protection systems, backup systems, and physical infrastructure all have to exist somewhere and be paid for by someone. Amateur radio constantly reminds me that communications technology is still governed by physics. Spectrum and bandwidth are finite. Antennas have limitations. Signals experience path loss and interference. Power has to come from somewhere, and heat has to go somewhere. Technology does not eliminate physical constraints. It changes which constraints we have to manage.
Right now, we are building computing infrastructure faster than many of the physical systems required to support it. Lawrence Berkeley National Laboratory estimates that data centers could account for approximately 11.8% of all U.S. electricity consumption by 2030 under its central scenario, with modeled scenarios ranging from approximately 9.5% to 15.3%.
https://eta.lbl.gov/publications/united-states-data-center-energy-2025
At the same time, America’s electrical grid already needed enormous investment. We were already dealing with aging infrastructure, transportation and industrial electrification, manufacturing growth, transmission constraints, transformer shortages, resilience concerns, storage, increasing electrical demand, physical security, and cybersecurity. The data-center boom did not create those problems, but adding enormous new industrial loads makes solving them much more urgent.
We need to modernize that grid, but ordinary homeowners, small businesses, taxpayers, and utility customers should not be expected to finance extraordinary infrastructure primarily required by some of the wealthiest corporations in the world. If a data center requires a new substation, it should pay its fair share. If it requires new transformers, transmission capacity, generation, storage, pipelines, water infrastructure, wastewater infrastructure, roads, or other major improvements, those costs need to be reflected honestly in the economics of the project.
Federal regulators are already confronting this issue. FERC has opened a proceeding examining how very large new electrical loads should connect to the interstate transmission system and how the costs and risks should be allocated. FERC has also ordered regional grid operators under its jurisdiction to justify or reform their rules for connecting major new loads while protecting ratepayers.
Ohio is already demonstrating that ratepayer protections are possible. The Public Utilities Commission of Ohio approved a special structure for large new data-center customers in AEP Ohio territory requiring substantial financial commitments intended to reduce the risk that existing customers are left paying infrastructure costs created for those developments.
https://puco.ohio.gov/wps/portal/gov/puco/utilities/electricity/resources/data-centers
Congress is moving in this direction too. H.R. 9340, the bipartisan Ratepayer Protection Act, introduced by Republican Rep. Gabe Evans and Democratic Rep. Kathy Castor, would require state regulators to consider protections for non-residential customers requesting 100 megawatts or more. The proposed standard would seek recovery of the full incremental costs of generation, transmission, and distribution upgrades needed to serve those loads and would require financial assurances intended to protect other customers. The House Energy and Commerce Committee advanced the bill by a unanimous 52-0 vote on July 21, 2026.
This petition supports the Ratepayer Protection Act and calls on Congress to go further. Electrical cost shifting is only one part of the issue. Responsible data-center policy must also address speculative capacity reservations, stranded infrastructure costs, grid modernization, transformer shortages, water consumption, residential siting, continuous and low-frequency noise, dedicated generation, tax subsidies, environmental impacts, decommissioning, storage, distributed generation, and long-term resource planning.
This is not a petition to stop building data centers. It is a petition to build them responsibly.
Build the data centers we need. Modernize the American electrical grid. Build a diverse, interconnected, decentralized, resilient energy system capable of supporting the future. Protect taxpayers, ratepayers, water, and communities. Make extraordinary private demand carry extraordinary private responsibility.
TL;DR — END
DETAILED CASE — START
WHY STRONGER RULES ARE NECESSARY
The short version above explains what this petition is asking for. The rest explains why.
My perspective sits at the intersection of cybersecurity, cloud infrastructure, electrical infrastructure, and communications. The lesson those areas share is simple: digital systems do not exist separately from the physical world.
A facility consuming hundreds of megawatts cannot simply be “plugged in” because somebody purchased land and ordered servers. The generation, transmission, substations, transformers, distribution infrastructure, cooling, water, communications, backup systems, and supporting infrastructure all have to exist. If they do not, somebody has to build them.
The real policy question is not whether we need data centers. We do. The question is who should pay for the extraordinary infrastructure required to support them.
When infrastructure broadly benefits the public, public investment makes sense. When extraordinary infrastructure is required primarily because an enormous private industrial customer has arrived, that customer should carry the appropriate share of the cost.
Extraordinary private demand should carry extraordinary private responsibility.
THE SCALE HAS FUNDAMENTALLY CHANGED
This is not another office building being connected to the grid. The scale of modern data-center development is dramatically different.
The Department of Energy reported that U.S. data centers consumed approximately 4.4% of total U.S. electricity in 2023.
Lawrence Berkeley National Laboratory’s updated modeling estimates that by 2030, data centers could consume approximately 11.8% of U.S. electricity under its central scenario, with modeled outcomes ranging from approximately 9.5% to 15.3%.
https://eta.lbl.gov/publications/united-states-data-center-energy-2025
Regardless of exactly where the final number lands, this represents an extraordinary amount of new electrical demand concentrated into a relatively short period. That demand is also not distributed evenly. A giant facility can suddenly create enormous demand inside one utility territory or transmission region, creating needs for generation, transmission, transformers, substations, pipelines, backup generation, water systems, roads, and other infrastructure that did not previously exist.
This is why the economics of a proposed data center cannot stop at “How much does the building cost?” The real question is: What infrastructure must exist throughout the surrounding region for this facility to operate, and who is paying for it?
AMERICA’S ELECTRICAL GRID WAS ALREADY OVERDUE FOR MODERNIZATION
The data-center boom did not create America’s electrical infrastructure problems. Many were already here.
We were already asking difficult questions about how the grid would handle transportation electrification before hyperscale AI facilities became such a major concern. As electric vehicles become more common, energy demand shifts from petroleum infrastructure onto the electrical system. That means charging infrastructure, local distribution capacity, transformers, generation, and load management.
Electric heating and heat pumps create additional demand. Manufacturing creates demand. New housing creates demand. Industrial electrification creates demand. Economic growth creates demand. Aging infrastructure needs replacement regardless of whether a data center ever arrives.
Those were legitimate planning challenges before the current data-center boom. Now we are attempting to address them while simultaneously connecting facilities whose electrical demands can be measured in hundreds or even thousands of megawatts.
That does not mean we should stop electrifying transportation, stop building homes and factories, or stop building data centers. It means we need to acknowledge reality: the United States needs a major modernization of its electrical infrastructure.
We need additional generation, stronger transmission, more substations, more transformers, modernized distribution systems, better regional interconnection, storage, modern monitoring and control systems, physical security, cybersecurity, redundancy, and greater resilience against severe weather and other emergencies. We need to stop waiting for infrastructure to fail before replacing or expanding it.
THE GRID EQUIPMENT ITSELF IS BECOMING A BOTTLENECK
It is not only a question of generating enough electricity. The physical equipment needed to move and transform that electricity has its own supply constraints.
Transformers, switchgear, breakers, conductors, generators, and other heavy electrical equipment cannot always be ordered today and delivered next month. Some equipment has manufacturing lead times measured in years.
Reuters reported in July 2026 that surging data-center demand is contributing to shortages of critical grid equipment, with some transformer lead times extending beyond 160 weeks.
That matters far beyond data centers. Utilities need transformers. Factories need them. Neighborhoods need them. Hospitals need them. Schools need them. Housing developments need them. Critical infrastructure needs them. Communities rebuilding after disasters need them.
When extraordinarily wealthy companies can reserve enormous quantities of grid equipment years in advance, everyone else is competing inside the same constrained supply chain. Infrastructure planning needs to occur before giant projects advance.
MODERNIZE THE GRID, AND MAKE DATA CENTERS PART OF THE INVESTMENT
The data-center boom could make America’s grid problems dramatically worse, or it could help solve them.
There is an unprecedented amount of private capital flowing into computing infrastructure. America already needs major electrical infrastructure investment. Those two facts should be connected.
If data-center developers want extraordinary access to America’s electrical system, some of that investment should help expand and modernize the system necessary to support them. If a data center requires a new substation, it should pay its appropriate share. If it requires additional transmission, the incremental costs associated with that demand should be reflected in the project. The same principle should apply to transformers, generation, storage, pipelines, water systems, wastewater systems, roads, and other major infrastructure.
There is a difference between replacing infrastructure that already broadly serves the public and constructing extraordinary infrastructure primarily because one giant private customer has arrived.
If a utility replaces an aging transformer serving an existing neighborhood, that is normal utility investment. If one hyperscale campus suddenly requires a completely new substation, major transmission work, multiple transformers, and hundreds of megawatts of additional capacity, that is different.
The incremental cost should follow the incremental demand.
Data centers should become part of America’s grid-modernization investment, not another reason ordinary Americans pay more for electricity.
PASS THE RATEPAYER PROTECTION ACT AND GO FURTHER
This idea is not starting from zero.
H.R. 9340, the Ratepayer Protection Act, was introduced by Republican Rep. Gabe Evans of Colorado and Democratic Rep. Kathy Castor of Florida. On July 21, 2026, the House Committee on Energy and Commerce advanced it by a unanimous 52-0 vote.
The legislation would require state regulators to consider a large-load standard for non-residential customers with a peak demand of 100 megawatts or more. The proposed standard would seek recovery of the full incremental costs of generation, transmission, and distribution upgrades necessary to serve the load while also requiring financial assurances intended to protect other customers.
That is exactly the kind of policy direction this petition supports.
Congress should pass the Ratepayer Protection Act, but it should treat it as a floor, not the finish line.
Ratepayer protection addresses one of the most urgent parts of this problem, but responsible data-center policy must also address grid modernization, speculative capacity reservations, stranded investments, equipment shortages, water use, zoning, continuous and low-frequency noise, subsidies, environmental impacts, storage, distributed generation, on-site generation, and decommissioning.
Protecting ratepayers is the starting point. Making data-center development sustainable for the electrical grid and the communities hosting it is the larger goal.
PROTECT RATEPAYERS FROM STRANDED INFRASTRUCTURE COSTS
What happens if a utility builds infrastructure and the data center never shows up?
A proposed data center can influence years of planning. Utilities may order transformers, construct substations, expand transmission, reserve generation, acquire land, and make other long-term investments based on expected demand.
But markets change. Companies change strategies. AI architectures change. Projects get delayed, downsized, relocated, or canceled.
Utilities and regulators therefore need to ensure that companies requesting enormous amounts of capacity make meaningful financial commitments before the public takes on the risk. That can include long-term contracts, minimum-demand commitments, deposits or other financial assurances where warranted, and exit obligations if a project triggers major infrastructure investment and then disappears.
Ordinary customers should not become involuntary investors in speculative data-center development.
STOP SPECULATIVE CAPACITY HOARDING
Electrical capacity itself has value. A developer asking for 500 megawatts, one gigawatt, or more is not simply filling out another service application. Those requests affect utility forecasts, transmission planning, generation decisions, infrastructure reservations, and other customers waiting for connections.
FERC has specifically raised concerns around how major new loads are forecast, connected, and financially committed.
Developers requesting extraordinary quantities of electricity should demonstrate real site control, financing, credible construction schedules, realistic demand estimates, utility plans, water plans, and meaningful financial commitment.
Scarce electrical capacity should not become something companies can casually reserve across multiple speculative sites while deciding later which project they actually intend to build.
INFRASTRUCTURE BEFORE HARDWARE
There is something fundamentally backwards about acquiring computing hardware faster than the infrastructure required to operate it can be built.
Companies are competing for GPUs, memory, storage, servers, networking equipment, switchgear, transformers, generators, and cooling infrastructure while electrical equipment can take years to procure.
The electrical planning has to come first. The transmission planning has to come first. The water planning has to come first. Projects need to demonstrate that they are technically viable as complete systems, not simply that somebody can afford the servers.
When enormous quantities of computing hardware are being purchased or committed to while adequate power infrastructure still does not exist, something about the planning process has gone wrong.
We should not be creating data centers full of hardware with no power available to run the hardware.
CONSUMERS ARE GETTING HIT THROUGH THE TECHNOLOGY SUPPLY CHAIN TOO
The resource competition extends beyond electricity.
The AI data-center buildout has created extraordinary demand for memory, storage, GPUs, and other components. In June 2026, industry groups representing automakers, retailers, electronics companies, and other sectors warned that AI data-center demand for memory chips was contributing to shortages and potentially significant consumer price increases.
Later that month, Apple raised prices on MacBooks and iPads as memory and storage costs surged amid the AI infrastructure boom.
The point is not that every increase in every phone or computer can be blamed directly on data centers. The point is that massive AI infrastructure demand is competing in the same supply chains used by consumers, schools, businesses, corporate IT departments, researchers, developers, repair shops, homelab users, electronics manufacturers, and automakers.
There is something deeply backwards about purchasing enormous amounts of computing hardware faster than the supporting electrical infrastructure can be constructed while everyone else simultaneously competes for increasingly constrained components.
Innovation should make computing more accessible. We should question an infrastructure arms race when it starts accomplishing the opposite.
MY OWN EXPERIENCE IS WHY COST SHIFTING MATTERS TO ME
This is not only an abstract policy concern for me.
Over the past several years, my own household electric bill has increased by more than 50%. During roughly that same period, efficiency improvements I made reduced my average electricity consumption by approximately 25%.
I am not claiming that data centers alone caused that increase in my particular bill. Electricity rates are complicated. Fuel prices, generation, transmission, storm restoration, capital projects, regulation, inflation, and many other factors affect rates.
But my experience is exactly why I care about what happens next.
Consumers are constantly told to conserve energy, buy efficient appliances, improve insulation, replace lighting, upgrade HVAC systems, manage thermostats, and reduce consumption. Many of us have spent our own money doing exactly that.
Ordinary customers should not make those investments only to have the financial benefits swallowed by infrastructure costs created by enormous new private industrial loads.
If a project creates extraordinary new demand, the costs specifically associated with that demand should primarily follow the project.
“JUST BUILD MORE RENEWABLES” IS NOT A COMPLETE GRID PLAN
Renewable energy absolutely belongs in America’s future energy strategy. But saying “just build more renewables” is not a complete engineering answer.
Generation is only one component of an electrical system. Solar generates electricity. Wind generates electricity. Nuclear generates electricity. Geothermal generates electricity. Hydroelectric facilities generate electricity. Natural-gas plants generate electricity.
Regardless of generation source, electricity still has to be generated, balanced, transmitted, distributed, protected, controlled, stored when appropriate, and delivered reliably when customers need it.
The International Energy Agency notes that as variable renewable resources such as solar and wind become larger portions of generation, power systems need greater flexibility through tools such as storage, flexible generation, flexible demand, stronger grids, and improved control systems.
https://www.iea.org/reports/electricity-grids-and-secure-energy-transitions/executive-summary
https://www.iea.org/reports/electricity-2026/flexibility
The argument is not that renewable energy cannot work. The stronger and more accurate point is that variable renewable generation alone is not a complete grid architecture.
THERE IS NO SINGLE MAGIC ENERGY SOURCE
America should stop looking for one technological savior.
Solar is modular, distributed, and increasingly affordable, but output varies by time, weather, season, and geography. Wind has enormous potential but is similarly location- and weather-dependent. Hydroelectricity provides valuable generation and flexibility but depends on geography and hydrology. Geothermal can provide firm generation, and emerging enhanced-geothermal technologies may expand where it is viable.
Nuclear energy can provide substantial firm generation but carries capital costs, construction timelines, regulatory requirements, fuel-cycle considerations, and waste-management responsibilities. Natural gas can provide dispatchable power but brings fuel-price, infrastructure, emissions, and environmental considerations. Storage can solve important balancing and resilience problems, but different storage technologies serve different timescales and have different costs and material requirements.
Research into advanced reactor designs, alternative nuclear fuel cycles including thorium-based approaches, enhanced geothermal systems, fusion, long-duration storage, and other emerging technologies should continue.
No technology should be treated as a magic word that makes engineering constraints disappear.
The goal should be a diverse, affordable, resilient, maintainable, scalable, and secure electrical system.
URUGUAY SHOWS WHAT DELIBERATE LONG-TERM ENERGY PLANNING CAN ACCOMPLISH
The United States does not have to invent every policy lesson from scratch.
According to the U.S. International Trade Administration, Uruguay generated approximately 99% of its electricity from renewable sources in 2024, primarily using hydropower, wind, and biomass, with solar also contributing.
https://www.trade.gov/country-commercial-guides/uruguay-energy
The lesson is not that America should copy Uruguay’s exact energy mix. Our geography, population, industries, resources, and electrical demand are dramatically different.
The useful lesson is: have an energy strategy instead of an energy patchwork.
Uruguay made a deliberate long-term transition involving multiple generation technologies and coordinated policy. America needs the same willingness to decide what kind of electrical system we want decades from now and deliberately build toward it.
A MODERN GRID SHOULD BE STRONGER NATIONALLY AND MORE DECENTRALIZED LOCALLY
The traditional electrical model relies heavily on large centralized generating facilities, long-distance transmission, substations, distribution systems, and customers at the edge.
That architecture is not going away, nor should it. America needs more transmission capacity between regions, not less.
But centralized infrastructure does not need to be our only tool. We should also expand distributed generation, rooftop solar, community solar, commercial solar, home and community storage, microgrids, virtual power plants, flexible demand, and other distributed energy resources.
The Department of Energy describes microgrids as localized systems capable of coordinating generation, storage, and loads and, when appropriately designed, disconnecting from the larger grid to operate independently.
https://www.energy.gov/oe/microgrid-systems
A twenty-first-century electrical system can therefore be more interconnected nationally while simultaneously becoming more decentralized locally.
Those are not contradictory ideas. Strong regional transmission allows areas to support one another. Distributed local resources reduce strain on distant generation and transmission. Microgrids provide resilience when parts of the larger system fail.
THINK OF THE GRID AS A RESILIENT NETWORK
My cybersecurity background strongly influences how I think about this.
In cybersecurity and network engineering, unnecessary single points of failure are bad architecture. We design redundant systems, alternate routes, failover, segmentation, distributed services, local resilience, backups, and recovery plans.
Our energy system can benefit from similar thinking.
Imagine communities with local storage, hospitals operating resilient microgrids, emergency facilities with local generation, municipal buildings capable of supporting critical services during outages, homes with distributed solar and batteries, businesses capable of reducing demand during grid emergencies, and data centers capable of supporting or relieving the grid instead of only consuming from it.
Then connect those systems through strong regional infrastructure capable of providing support when local resources are insufficient.
The goal is not to disconnect everybody from the grid. The goal is to build a stronger network where more parts of the system can support one another and fewer failures cascade unnecessarily.
STORAGE SHOULD BECOME INFRASTRUCTURE, NOT AN AFTERTHOUGHT
Energy storage needs to become a normal part of electrical planning, and storage should not mean building one enormous battery somewhere.
We should increasingly think about cells of storage capacity throughout the grid: home batteries, neighborhood and community storage, commercial and industrial systems, data-center storage, microgrid storage, utility-scale batteries, thermal storage, pumped hydroelectric storage where geography permits, vehicle-to-grid systems where technically appropriate, long-duration storage, and emerging technologies as they mature.
Storage can absorb electricity when generation exceeds immediate demand, make energy available during demand peaks, support local systems during outages, provide grid services, help integrate variable generation, reduce short-duration peaks, and provide another layer of resilience.
Different technologies will fit different applications. The point is not to choose one battery chemistry. The point is to treat storage as a normal component of grid infrastructure.
DATA CENTERS SHOULD BECOME GRID PARTICIPANTS, NOT SIMPLY GIANT LOADS
At a certain scale, a data center should no longer be treated like an ordinary customer plugging another appliance into the wall.
If a facility consumes electricity on the scale of a city, it has become part of the electrical system whether policymakers acknowledge that or not.
The question should not only be “How quickly can we connect you?” It should also be “What are you contributing to the system you are asking to use?”
Can the facility build or finance storage? Can it operate a microgrid? Can it provide demand flexibility? Can non-time-sensitive workloads shift away from extreme grid peaks? Can some backup resources provide grid support where technically and environmentally appropriate? Can the project finance generation or transmission? Can excess heat be reused where practical? Can infrastructure built for the project also benefit the surrounding community?
If companies want extraordinary access to America’s electrical infrastructure, we should expect extraordinary participation in strengthening it.
MAKE SMALL-SCALE ENERGY GENERATION EASIER FOR ORDINARY AMERICANS TOO
If America desperately needs more generation and distributed resources, why is it still so difficult and expensive for ordinary people to generate even modest amounts themselves?
Traditional rooftop solar can involve major equipment costs, professional installation, permitting, engineering, utility interconnection, inspections, financing, contractor overhead, and substantial administrative complexity. For many households, the realistic choices become spending tens of thousands of dollars or doing nothing.
There should be more options in between.
Germany’s federal network regulator reported approximately 430,000 new plug-in balcony solar installations in 2025 alone, adding roughly 0.5 gigawatts of capacity.
https://www.bundesnetzagentur.de/SharedDocs/Pressemitteilungen/EN/2026/20260108_EEG.html
America should examine how we can create similarly accessible options.
Because I used to work as an electrician, I want to be very clear: this does not mean allowing people to plug random homemade solar equipment into outlets. Anti-islanding protection matters. Overcurrent protection matters. Equipment certification matters. Grounding matters. Conductor ratings matter. Connector design matters. Utility-worker safety matters.
But safe does not have to mean needlessly inaccessible.
Federal regulators, standards organizations, manufacturers, states, utilities, and code bodies should establish a clear pathway for appropriately engineered, certified, current-limited, protected small-scale plug-in solar and storage systems.
Renters should have options. Apartment residents should have options. Low-income households should have options. Homeowners should have inexpensive options. Small businesses should have options.
A few hundred watts from one balcony will not power a data center. That is not the point. A few hundred watts across millions of buildings becomes meaningful distributed generation.
WATER CANNOT BE TREATED AS AN UNLIMITED INDUSTRIAL RESOURCE
Electricity is not the only major resource involved.
Depending on design, cooling technology, and climate, data centers can consume substantial quantities of water. DOE states that data-center water consumption varies widely based on cooling technology and local conditions but can reach millions of gallons per day, while newer cooling approaches can dramatically reduce water requirements.
The argument should not be that every data center wastes millions of gallons of water. That would be inaccurate.
The stronger argument is that water use depends heavily on engineering decisions, so responsible engineering and disclosure should be mandatory.
Projects should disclose expected water demand before approval and actual consumption once operating. Developers should identify whether the source is municipal drinking water, groundwater, surface water, reclaimed wastewater, or another supply. Closed-loop and lower-water cooling should be prioritized where technically appropriate, and reclaimed or non-potable water should be considered where feasible.
Facilities proposed in drought-prone regions, areas dependent on stressed aquifers, or communities with limited water capacity deserve additional scrutiny.
The question should not simply be “Can the company afford this water?” It should be “Can the community sustainably provide this water?”
A corporation being able to purchase millions of gallons does not create millions of gallons of additional water.
If a project requires additional wells, treatment facilities, pipes, pumps, sewer capacity, or wastewater systems, those costs should be included in the economics of the project.
DATA CENTERS ARE INDUSTRIAL INFRASTRUCTURE. ZONE THEM ACCORDINGLY.
A hyperscale data center is not an office building. It is industrial infrastructure.
Cooling systems can operate continuously. Transformers operate continuously. Fans and pumps operate. Backup generators require testing. Electrical systems create sound. Construction and expansion can continue for years.
Virginia’s Joint Legislative Audit and Review Commission found that constant low-frequency data-center noise has created problems when facilities are located near residential areas and noted that conventional noise rules do not always address this type of sound well.
https://rga.lis.virginia.gov/Published/2025/RD206
That is exactly why siting matters.
Large data centers should be subject to appropriate industrial zoning, meaningful setbacks from homes, buffers, independent acoustic modeling, low-frequency noise analysis, nighttime standards, post-construction testing, generator-noise analysis, environmental review, and meaningful enforcement.
Residential zoning should mean something. A family should not buy or rent a home in an established residential community only to discover that their neighborhood has effectively become an industrial district because a developer found enough acreage nearby.
CONTINUOUS AND LOW-FREQUENCY NOISE DESERVES SPECIAL TREATMENT
Traditional noise ordinances often address parties, vehicles, temporary construction, lawn equipment, dogs, and other intermittent sound.
A data center can produce something different: continuous, relatively stable, low-frequency or tonal sound twenty-four hours a day.
Virginia’s JLARC found that residents can experience quality-of-life problems even where data-center sound does not necessarily violate conventional noise ordinances.
https://rga.lis.virginia.gov/Published/2025/RD206
Policymakers therefore need measurement standards appropriate for the source. Rules should consider continuous sound, low-frequency components, tonal characteristics, nighttime background levels, cumulative noise from multiple buildings, transformers, cooling systems, generators, and continued campus expansion.
STOP GIVING BLANK CHECKS TO SOME OF THE WEALTHIEST COMPANIES ON EARTH
According to the National Conference of State Legislatures, 38 states offer dedicated tax incentives for data centers, including sales-tax exemptions, use-tax exemptions, property-tax abatements, and other programs.
https://www.ncsl.org/fiscal/subsidizing-servers-how-states-are-competing-to-attract-data-centers
Economic-development incentives are not automatically bad. There may be circumstances where incentives produce genuine public benefit.
But they should not become an automatic bidding war where communities compete to offer increasingly favorable treatment to companies already capable of financing enormous projects themselves.
Any major subsidy should include a public cost-benefit analysis, clear employment and investment commitments, transparent infrastructure costs, expiration dates, periodic review, measurable performance requirements, and enforceable clawbacks if promised benefits do not materialize.
A data center can represent billions of dollars of capital investment while employing significantly fewer permanent workers than many other forms of industrial development. That does not mean data centers provide no economic benefit. It means the benefits need to be evaluated honestly.
If a project receives major tax breaks while ratepayers finance grid expansion and taxpayers finance supporting infrastructure, public officials should demonstrate why the arrangement actually benefits the public.
Private investment should not mean private profit and public infrastructure bills.
ON-SITE GENERATION CANNOT BECOME A LOOPHOLE
Some developers are responding to limited grid capacity by considering or constructing dedicated natural-gas turbines, reciprocating engines, generator installations, pipelines, and islanded power systems.
The issue is not simply that natural gas exists. The issue is whether moving generation behind the fence becomes a way to bypass responsibilities that would otherwise apply.
EPA maintains specific Clean Air Act resources for data centers and relevant stationary sources.
https://www.epa.gov/stationary-sources-air-pollution/clean-air-act-resources-data-centers
If a data center needs dedicated generation, the full system still needs to be accounted for: pipelines, fuel, emissions controls, noise, equipment, maintenance, safety, environmental impacts, and eventual decommissioning.
Running out of grid capacity should not mean creating an entirely separate power system and pretending its impacts do not exist.
DATA-CENTER DEVELOPERS SHOULD HAVE DECOMMISSIONING RESPONSIBILITIES
Computing infrastructure changes quickly. Facilities become obsolete. Companies consolidate. Buildings become economically undesirable. Equipment reaches the end of its useful life.
Communities should not be left with abandoned industrial sites and expensive infrastructure built around facilities that no longer operate.
Large projects should therefore have credible decommissioning and remediation plans, potentially including financial assurance appropriate to the size and environmental risks of the project.
If specialized pipelines, electrical infrastructure, cooling systems, fuel systems, batteries, industrial equipment, or contaminated materials need to be removed or remediated later, taxpayers should not automatically inherit those costs.
The company benefiting from the facility during its useful life should bear responsibility for responsibly closing it too.
PUTTING DATA CENTERS IN SPACE DOES NOT MAKE PHYSICS DISAPPEAR
There is growing interest in orbital computing and the idea of moving substantial computing infrastructure into space.
Space computing may eventually make sense for specialized applications. I am not arguing that research should stop.
But “put the data centers in space” is not a magic escape hatch from terrestrial infrastructure constraints.
As an amateur radio operator, the communications problem is one of the first things that jumps out at me.
A computer in orbit still needs to communicate. You still need spectrum, antennas, ground infrastructure, uplinks, downlinks, potentially inter-satellite links, link budgets, redundancy, and enormous amounts of network capacity. You still have path loss, interference, finite bandwidth, latency, orbital geometry, and handoffs.
Moving enormous amounts of computation into orbit means figuring out how enormous amounts of data move into, out of, and between those systems.
Physics does not disappear because the server rack is above the atmosphere.
SPACE IS COLD, BUT COOLING IN SPACE IS NOT SIMPLE
One of the most misleading intuitive ideas surrounding orbital data centers is “Space is cold, so cooling will be easy.”
Vacuum does not work like cold air.
On Earth, data centers can move heat through air and liquid systems and reject that heat into the environment. A spacecraft cannot use the surrounding vacuum as an enormous convection cooling system.
Waste heat has to be transported through the spacecraft and ultimately rejected primarily through radiation. NASA treats thermal control and radiative heat rejection as fundamental spacecraft engineering requirements.
https://www.nasa.gov/smallsat-institute/sst-soa/thermal-control/
That does not mean orbital computing is physically impossible. It means moving computation into space trades one thermal problem for another.
Engineering still has to close the thermal budget.
MAINTENANCE AND HARDWARE REPLACEMENT BECOME RADICALLY HARDER
On Earth, a failed server can be pulled from a rack and repaired or replaced. GPUs can be upgraded. SSDs and RAM can be replaced. Networking equipment can be upgraded. Cooling systems can be modified. Obsolete equipment can be removed and recycled.
Space changes that equation dramatically.
Every piece of hardware must remain reliable for its intended lifetime, have redundancy, be remotely or robotically serviceable, or eventually be replaced by launching new equipment.
Computing hardware also becomes obsolete quickly. A GPU that makes economic sense today may be considerably less attractive several years from now.
That creates a fundamental lifecycle question: How frequently will orbital data centers need to be replaced to remain economically competitive, and what happens to the old hardware?
WE RISK TURNING E-WASTE INTO AEROSPACE WASTE
Terrestrial computing already creates an enormous electronic-waste problem. Moving large amounts of rapidly depreciating computing hardware into orbit adds another layer.
If the operating model becomes launch, use, deorbit, replace, repeat, then environmental accounting needs to include the entire lifecycle: manufacturing, spacecraft materials, launch vehicles, propellants, launch effects, replacement hardware, reentry, and disposal.
That is not an argument to ban research. It is an argument to count all of the costs before presenting space as an easy environmental solution.
ORBITAL DEBRIS IS ALREADY A REAL REGULATORY ISSUE
Orbit is not an infinite garbage dump.
The FCC already maintains orbital-debris mitigation requirements and end-of-life rules for satellite operators.
https://www.fcc.gov/document/fcc-adopts-new-5-year-rule-deorbiting-satellites-0
This matters tremendously if orbital computing eventually involves very large constellations. More objects mean more collision-avoidance requirements, tracking requirements, failure scenarios, debris risks, replacement missions, and end-of-life operations.
The lifecycle has to be part of the economics from the beginning.
THE NIGHT SKY AND RADIO ASTRONOMY MATTER TOO
There is also a public resource that rarely appears on a corporate balance sheet: the night sky and radio environment.
Large satellite constellations already create concerns for optical astronomy and radio astronomy. The International Astronomical Union’s Centre for the Protection of the Dark and Quiet Sky works specifically on the impacts of large constellations.
Researchers have also detected unintended electromagnetic radiation from large satellite constellations affecting radio-astronomy observations.
As a ham radio operator, this aspect particularly concerns me. Spectrum and the radio environment are shared resources.
We should not wait until enormous orbital computing infrastructure exists before asking how it affects communications, astronomy, spectrum management, interference, orbital congestion, and the night sky.
Those questions should be addressed before deployment at enormous scale.
SPACE COMPUTING MAY HAVE A FUTURE, BUT IT REQUIRES FULL LIFECYCLE ACCOUNTING
Specialized orbital computing may eventually make sense. Processing space-generated sensor data closer to where it originates may make sense. Some workloads may make sense in orbit. Research should continue.
But orbital data centers should be evaluated under the same principle as terrestrial data centers: account for the complete system, not merely the server.
For orbital computing, that means power generation, storage, thermal rejection, launch requirements, spacecraft mass, spectrum, ground stations, inter-satellite communication, space-to-ground bandwidth, latency, radiation, hardware lifetime, servicing, redundancy, replacement cycles, debris, collision avoidance, reentry, astronomy, radio interference, manufacturing impacts, and end-of-life disposal.
You cannot simply move a server into space and declare that its terrestrial environmental footprint disappeared.
You cannot launch bad infrastructure policy into orbit and call the problem solved.
THE LARGER OPPORTUNITY
The frustrating thing about all of this is that the data-center boom could actually become an enormous opportunity.
America already needs major electrical infrastructure investment. We need more transmission, generation, transformers, substations, storage, regional interconnection, resilient distribution, sophisticated control systems, distributed resources, cybersecurity, physical resilience, and emergency capability.
There is also an unprecedented amount of private money flowing into computing infrastructure.
That investment could help accelerate the modernization we already needed.
Instead of asking only, “How do we find enough electricity for data centers?” we should also ask, “How do we use this investment to strengthen the electrical system for everybody?”
A transmission project required by a data-center campus might also increase regional capacity. A new substation might account for future community growth. A giant facility could finance storage. A data center could operate as a microgrid. Flexible computing workloads could participate in demand-response programs. Projects could finance additional generation. Communities could receive resilience improvements.
Data-center investment could also help expand domestic manufacturing of transformers, switchgear, conductors, and other critical electrical equipment.
There is an opportunity to turn an infrastructure crisis into an infrastructure modernization effort, but that only happens if policy requires it.
WHAT WE ARE ASKING POLICYMAKERS TO DO
DEMAND 01: PASS THE RATEPAYER PROTECTION ACT AND BUILD UPON IT. Congress should pass H.R. 9340, the Ratepayer Protection Act, and establish strong protections ensuring that extraordinarily large electrical customers pay the incremental generation, transmission, distribution, and other infrastructure costs created by their demand.
DEMAND 02: PROTECT RESIDENTIAL AND SMALL-BUSINESS RATEPAYERS FROM COST SHIFTING. Utilities and regulators should use appropriate large-load tariffs, minimum-demand commitments, long-term contracts, and financial assurances so existing customers are not forced to subsidize infrastructure built primarily for giant private loads.
DEMAND 03: PROTECT CUSTOMERS FROM STRANDED INFRASTRUCTURE COSTS. If a data-center developer causes a utility to invest in major infrastructure and later cancels, relocates, delays, or substantially downsizes the project, the developer should remain financially responsible for the appropriate costs.
DEMAND 04: REQUIRE PROJECT READINESS BEFORE MASSIVE CAPACITY RESERVATIONS. Developers seeking extraordinary electrical capacity should demonstrate site control, financing, realistic demand estimates, credible construction schedules, utility plans, and meaningful financial commitment before tying up scarce grid resources.
DEMAND 05: STOP SPECULATIVE CAPACITY HOARDING. Companies should not be allowed to reserve enormous quantities of electrical capacity indefinitely or across multiple speculative sites without meaningful financial consequences.
DEMAND 06: CREATE A NATIONAL GRID-MODERNIZATION STRATEGY. America should accelerate investment in generation, transmission, substations, transformers, distribution systems, regional interconnection, storage, grid-enhancing technologies, cybersecurity, physical security, monitoring, controls, and resilience.
DEMAND 07: MAKE DATA-CENTER INVESTMENT PART OF GRID MODERNIZATION. The larger the new electrical load, the greater the project’s responsibility should be for financing infrastructure required to accommodate it.
DEMAND 08: PURSUE A DIVERSE, TECHNOLOGY-NEUTRAL ENERGY STRATEGY. Solar, wind, geothermal, hydroelectricity, nuclear energy, storage, emerging reactor technologies, alternative fuel cycles, and other resources should be evaluated based on reliability, safety, environmental effects, cost, geography, scalability, construction time, and engineering reality.
DEMAND 09: EXPAND DISTRIBUTED GENERATION AND MICROGRIDS. Communities, hospitals, campuses, businesses, critical facilities, and neighborhoods should have greater ability to generate, store, and manage electricity locally while remaining connected to a stronger regional grid.
DEMAND 10: BUILD DISTRIBUTED CELLS OF ENERGY STORAGE. Home, community, commercial, industrial, microgrid, utility-scale, thermal, pumped-hydro, vehicle-to-grid, long-duration, and emerging storage technologies should be encouraged where technically appropriate.
DEMAND 11: TREAT ENORMOUS DATA CENTERS AS ACTIVE GRID PARTICIPANTS. Regulators should evaluate storage, demand flexibility, microgrid capability, generation investment, grid-support services, and infrastructure contributions proportional to the scale of a project’s demand.
DEMAND 12: CREATE A SAFE PATHWAY FOR INEXPENSIVE SMALL-SCALE PLUG-IN SOLAR AND STORAGE. Federal regulators, standards organizations, utilities, manufacturers, states, and electrical-code bodies should establish safe standards allowing appropriately certified small systems to be deployed with significantly less cost and bureaucracy than full rooftop installations.
DEMAND 13: PLAN DATA CENTERS AND OTHER LARGE NEW LOADS TOGETHER. Data centers, transportation electrification, industrial electrification, housing growth, manufacturing, and other major electrical demands compete for the same generation, transmission, transformers, equipment, and labor and should not be planned as unrelated problems.
DEMAND 14: REQUIRE WATER-USE TRANSPARENCY AND SUSTAINABLE WATER PLANNING. Developers should disclose expected and actual consumption, identify water sources, demonstrate long-term availability, prioritize efficient cooling and reclaimed water where practical, and pay for infrastructure their demand requires.
DEMAND 15: PROTECT DRINKING WATER AND GROUNDWATER. Projects in drought-prone regions, areas with stressed aquifers, or communities with limited municipal capacity should receive additional scrutiny.
DEMAND 16: KEEP NEW HYPERSCALE DATA CENTERS OUT OF ESTABLISHED RESIDENTIAL NEIGHBORHOODS. Local governments should require appropriate industrial zoning, setbacks, buffers, environmental review, acoustic analysis, and meaningful public participation.
DEMAND 17: ESTABLISH ENFORCEABLE DATA-CENTER NOISE STANDARDS. Rules should account for continuous sound, low-frequency and tonal noise, cooling equipment, transformers, backup generation, nighttime conditions, cumulative impacts, and post-construction verification.
DEMAND 18: END BLANK-CHECK SUBSIDIES. Data-center tax incentives should require transparent evidence of genuine net public benefit, measurable commitments, expiration dates, periodic review, and enforceable clawbacks.
DEMAND 19: REQUIRE PUBLIC DISCLOSURE OF MAJOR RESOURCE COMMITMENTS AND INCENTIVES. Communities deserve to know anticipated electricity demand, water consumption, generation plans, emissions, infrastructure requirements, tax incentives, and other public commitments before projects are approved.
DEMAND 20: DO NOT ALLOW ON-SITE GENERATION TO BECOME A REGULATORY LOOPHOLE. Dedicated natural-gas or other power systems should remain subject to appropriate environmental, air-quality, safety, noise, public-review, infrastructure-cost, and decommissioning requirements.
DEMAND 21: REQUIRE DEVELOPERS TO FUND DECOMMISSIONING AND REMEDIATION. Communities should not inherit abandoned facilities, obsolete infrastructure, contaminated property, or cleanup expenses when data centers close.
DEMAND 22: REQUIRE FULL LIFECYCLE ANALYSIS FOR ORBITAL DATA CENTERS. Any proposal to move computing infrastructure into orbit at large scale should account for communications, spectrum, ground infrastructure, thermal management, power, radiation, servicing, replacement, launches, orbital debris, astronomy, radio interference, reentry, and end-of-life disposal.
THIS IS NOT ANTI-TECHNOLOGY. IT IS RESPONSIBLE TECHNOLOGY.
I believe in cloud computing, cybersecurity, artificial intelligence research, scientific computing, modern communications infrastructure, electrification, renewable energy, nuclear research, energy storage, microgrids, distributed generation, and exploring technologies we have not perfected yet.
I believe in innovation.
But innovation is not a magic word that makes engineering constraints disappear.
My cybersecurity background tells me why we need this infrastructure. My electrical background tells me that infrastructure has physical requirements. My amateur-radio experience reminds me that even our most advanced communications networks are still governed by physics.
Power has to come from somewhere. Heat has to go somewhere. Water has to come from somewhere. Signals have to travel somehow. Equipment has to be manufactured. Infrastructure has to be maintained. Somebody ultimately has to pay for all of it.
The question is whether those costs are honestly included in the economics of a project or quietly transferred onto everybody else.
If your project requires a new substation, help pay for the substation. If it requires new transmission, help pay for the transmission. If it requires additional generation, help pay for the generation. If it requires storage, invest in storage. If it requires new transformers, account for them before construction. If it requires new water infrastructure, pay for the infrastructure your project requires. If it needs millions of gallons of water, demonstrate that the community can sustainably provide it. If it creates industrial noise, do not put it next to people’s homes. If it requires new pipelines or on-site generation, account for those costs and impacts. If it asks for hundreds of megawatts of electricity, prove the capacity exists or finance the infrastructure necessary to create it.
If utilities build major infrastructure because of your project and you later abandon that project, do not leave ordinary utility customers with the bill.
And if a project’s economics only work because taxpayers subsidize it while ordinary ratepayers absorb its infrastructure costs, then the project is not actually paying for itself. The public is simply paying part of the bill.
WE CAN BUILD BOTH FUTURES
America needs data centers. America also needs a dramatically better electrical grid. Those goals do not have to compete.
We can modernize transmission, expand generation, develop a diverse energy portfolio, build storage throughout the grid, establish resilient microgrids, expand distributed generation, make small-scale solar more accessible, strengthen local distribution, improve regional interconnection, invest in grid cybersecurity, and build infrastructure capable of supporting electric vehicles, manufacturing, housing, businesses, critical infrastructure, and data centers together.
And we can require companies creating extraordinary new demand to become part of the investment required to make that possible.
The answer is not “No data centers.”
The answer is responsible data centers.
Plan them before building them. Power them before filling them with hardware. Account for their water. Account for their noise. Account for their environmental impacts. Put industrial facilities in appropriate locations. Stop speculative projects from monopolizing scarce grid resources. Stop giving blank checks to some of the world’s wealthiest corporations. Protect residential and small-business ratepayers. Pass the Ratepayer Protection Act and build upon it. Modernize the grid. Build storage. Build local resilience. Strengthen national transmission. Support diverse generation. Continue energy research. Make distributed energy easier for ordinary Americans. And make private companies pay the fair cost of the extraordinary private demand they create.
Build the data centers we need. Modernize the American electrical grid. Build a diverse, interconnected, decentralized, resilient energy system capable of supporting the future. Protect our water. Protect our communities. Protect ordinary taxpayers and ratepayers. And make the companies creating extraordinary new demand pay their fair share of what it takes to support it.
DETAILED CASE — END
The Decision Makers
Supporter Voices
Petition Updates
Share this petition
Petition created on August 14, 2026