Author: Ryan McEnrush, Partner at a16z; Translation: 0xjs@Golden Finance
The electric grid is a vast, complex system of wires and power plants that is critical to our economy and the foundation of our industrial strength. The United States faces a daunting challenge: U.S. electricity demand is expected to nearly double by 2040 due to factors such as artificial intelligence computing, reflow, and “electrification,” but our grid infrastructure and operations are struggling to keep pace.
To seize an energy-rich future, we must simplify the production, transmission and consumption of electricity; this requires a decentralized grid. The construction of large power plants and long-distance transmission lines is cumbersome, but technologies such as solar energy, batteries and advanced nuclear reactors bring new possibilities. It is these and other more "local" technologies that can avoid costly long distance cabling and be deployed directly on site that will help support significant load growth over the coming decades.
While historical industrial expansion relied on large, centralized power plants, the 21st century marks a shift toward decentralized and intermittent energy, moving from a “hub and spoke” model to distributed networks. Of course, this evolution creates new challenges, and we need innovation to bridge the gap.
Growing Pains
The U.S. electric grid consists of three major interconnections: Eastern, Western, and Texas, managed by 17 NERC coordinators, with ISOs (Independent System Operators) and RTOs (Regional Transmission Operators) overseeing regional economies and infrastructure. Actual generation and delivery, however, is handled by local utilities or cooperatives. This structure worked in an era of low load growth, but it is becoming increasingly challenging and expensive to expand the grid infrastructure to meet today’s demands.
Access issues
Grid operators use queues to manage new asset connections, assess whether the grid can support the additional power at that location without unbalancing it, and determine the cost of necessary upgrades. Today, more than 2,000 GW is waiting to be connected to the grid, with more than 700 GW of projects entering the queue in 2022 alone. That’s a lot: The entire U.S. grid has only 1,200 GW of installed generating capacity.
But in reality, many projects have withdrawn due to grid connection costs. Historically, only 10-20% of queued projects have been realized, and it usually takes more than 5 years after application to finally connect - and these times are only getting longer. Power generators often submit multiple speculative proposals to determine the cheapest access point, and then withdraw unfavorable proposals once the costs are known, complicating feasibility studies. Due to the surge in applications, California grid operator CAISO was forced to stop accepting any new requests in 2022, and plans to do so again in 2024.
This is a key rate limiter and cost driver in our energy transition. A recent report from the U.S. Department of Energy found that to meet high load growth through 2035, intraregional transmission integrating new assets must increase by 128% and interregional transmission must increase by 412%. More optimistic forecasts predict growth rates of 64% and 34%, respectively.
Proposed reforms could help ease this development backlog. The Federal Energy Regulatory Commission (FERC) is pursuing a “first-in, first-out” policy that adds fees to filter proposals and speeds up review. The Electric Reliability Council of Texas (ERCOT) uses an “access and manage” approach that enables faster access but disconnects projects if they threaten grid reliability—a huge success in rapidly adding new grid assets. While these policies signal progress, streamlining other regulations, such as NEPA, will also be critical to speeding up construction.
But even if approved, grid construction faces supply chain hurdles, including lead times of more than 12 months, a 400% surge in prices for large power transformers, and shortages of specialty steel. Meeting federal goals to grow transformer manufacturing also depends on support for the electrical steel industry, especially for upcoming 2027 efficiency standards. All of this comes as grid outages, mostly weather-related, are at a 20-year high, requiring hardware replacement.
This does not include transportation
Ultimately, the cost of building grid infrastructure reforms manifests itself as higher prices for consumers. The “retail price” that consumers pay is a combination of the wholesale price (the cost of generating electricity) and the delivery charge (the cost of the infrastructure needed to get that electricity to you). Crucially, while the price of cheap renewable and natural gas generation has fallen, the price of delivering that electricity has risen dramatically.
There are many reasons for this. Utilities use distribution charges to offset customer generation losses, designed to ensure fixed-return revenue for infrastructure investments (similar to cost-plus defense contracting). The development of renewable energy requires extending power lines to remote areas, which are less used due to intermittency. In addition, as electrification and self-generation increase, loads become more variable, and infrastructure designed for peak demand becomes inefficient and costly.
Policy and market adjustments are responding to these rising transportation costs, with California’s massive adoption of distributed power systems such as rooftop solar being a notable example.
California's Net Energy Metering (NEM) program originally allowed homeowners to sell excess solar power back to the grid at retail prices, ignoring the utility's distribution costs. Recent changes now essentially buy back power at a variable wholesale price, reducing the revenue that solar panel owners receive during peak generation times, which typically coincide with the lowest electricity prices. The adjustment extends the payback period for solar installations and encourages homeowners and businesses to invest in electricity storage so they can sell energy when it's more profitable.
California utilities have also proposed a billing model in which a fixed fee is based on income levels and a usage fee is based on consumption. The move is intended to have wealthier customers shoulder more of the grid infrastructure costs and protect lower-income individuals from rising retail electricity prices. Although this specific policy was recently shelved in favor of a similar but less extreme version, such an idea could lead wealthy users to leave the grid entirely. Defection could result in higher costs for remaining users and trigger a "death spiral." Some believe this is already happening in the Hawaiian electricity market, where some areas quickly switched to electric heat pumps.
Keep the lights on
Electricity isn’t magic; the grid is complicated. At all times, the amount of electricity generated must be matched to the demand for electricity, or “load”; that’s what people mean when they say “balancing the grid.” At a high level, grid stability relies on maintaining a constant frequency—60 Hz in the United States.
Congestion caused by exceeding power line capacity (dumping power onto the grid) can lead to brownouts and local price differences. Any frequency deviation can also cause damage to generator and motor equipment. Wind, solar and batteries – reverse resources that lack inertia – also complicate frequency stabilization as they surge. In extreme cases, deviations can cause blackouts and even damage grid-connected equipment.
Because of the inherent fragility of the grid, the assets connected to it must be carefully considered to align reliable supply with forecasted demand. The growth of intermittent power sources (unreliable supply) combined with the rise of “electrification” (surging demand) is creating serious challenges.
When is enough, enough is enough?
About two-thirds of the load is balanced by wholesale markets through (mostly) day-ahead auctions, where the price is determined by the cost of the last unit of electricity required. Renewables have no marginal cost and typically outbid other energy sources when active, resulting in price volatility - very low prices when renewables meet demand, and skyrocketing prices when more expensive energy sources are needed (note: bids are different from the levelized cost of energy LCOE.)
The unpredictable nature of solar and wind power, along with the closure of aging fossil fuel power plants, puts pressure on grid stability. This can lead to blackouts (underproduction) and brownouts (overproduction), such as the 2,400 GWh of waste that will occur in California by 2022. Addressing this will require investments in energy storage and transmission improvements (described below).
Furthermore, as electricity supply becomes more unpredictable, natural gas plays an increasingly important role due to its cost-effectiveness and flexibility. Natural gas is often used to support renewable energy through "peaker plants," which are powered only when needed. In general, the intermittent nature of solar and wind power makes natural gas plants and other types of power plants intermittently profitable, and sometimes even continuously operating at a loss for technical reasons. As a result, when "peaker plants" set wholesale prices when renewable energy sources are out of power, this leads to higher costs, which in turn creates volatility for consumers.
The demand for electricity is also changing. Technologies such as heat pumps, while energy efficient, can cause winter load peaks when renewable generation is low. This requires grid operators to maintain a buffer of power assets, and renewables are often overlooked in resource adequacy planning. Grid operators often follow a “one in ten” rule, accepting a power shortage once every decade, but the actual calculation is more complicated. In ERCOT, we have seen “emergency reserves” grow as renewables come on the grid, in the absence of a traditional capacity market to replace price increase incentives.
Regions with high solar penetration, like California, also face a “duck curve,” requiring grid operators to quickly add more than 20 gigawatts of power as sunlight wanes and demand rises. That’s technically and economically challenging for plants designed to deliver continuous output.
The intermittent nature of renewable energy creates hidden costs that force grid operators to take risks or invest in new assets. While the levelized cost of energy assesses the economic viability of a project, it oversimplifies the asset’s true value to the wider grid. However, LCOE does highlight the economic challenges of building new assets like nuclear plants. Although more expensive than today’s natural gas, nuclear energy offers a compelling path to reliable decarbonization of electricity. We just need to scale up reactor construction.
But we can’t rely on nuclear power alone. Relying on a single energy source is risky, as shown by France’s nuclear challenges during Russian energy sanctions and natural gas issues during cold weather in the southern United States, not to mention commodity price volatility. Regions with large renewable energy sources, such as California, also face uncertainty due to their daily reliance on imports. Even places that run on nearly 100% clean energy, such as Iceland or Scandinavia, maintain reliable backup or import options during crises.
Become Smart
As demand for electricity grows, the grid struggles to cope with the increasing complexity of decentralized and intermittent renewable energy. We can’t force this transition by brute force; if we’re going to do it, we really need to get smart.
The current grid is aging and dumb, relying on power plants to adjust production based on forecasted demand while making small real-time adjustments to ensure stability. Originally designed for one-way flows from large power plants, the grid is challenged by the concept of multiple smaller sources providing power in all directions, such as your rooftop solar charging your neighbor’s electric car. Additionally, the lack of true visibility into real-time power flows poses looming problems, especially at the distribution level.
Residential solar, batteries, advanced nuclear and (possibly) geothermal energy provide decentralized power, reducing the need for infrastructure. However, integrating an ever-changing and unstable grid still requires innovative solutions. Additionally, the efficient use of utility-scale power systems can even be significantly improved through local storage and demand-side response (such as turning off the thermostat when the grid is stressed), thereby reducing the need to build underutilized assets that are only online briefly during peak periods of demand.
The “smart grid” aims to achieve all of these goals and more, and can be divided into three main technology groups:
Electric meter front end
Dynamic line rating, solid-state transformers, voltage management and power flow systems, better conductors, infrastructure monitoring, grid-scale generation, grid-scale storage, and more.
Instrument backend
Heat pumps, appliances, residential solar, home energy storage, electric vehicle chargers, smart thermostats, smart meters, microreactors and small modular reactors, microgrids, etc.
Power Grid Software
Virtual power plants, better forecasting, equipment management, energy data infrastructure, cybersecurity, ADMS, interconnection planning, power financial instruments, automation of bilateral agreements, etc.
Specifically, two trends are critical to the future of the "smart grid."
First, we need to build a lot of energy storage to smooth local peak loads and stabilize intermittent power across the grid. Batteries are already essential for small bursts of power, and as prices continue to fall, they can cover even longer periods of time. But scaling up batteries by hundreds of gigawatt hours will also require scaling up the supply chain. With luck, strong economic conditions are likely to continue to accelerate deployment; entrepreneurs should seek to plug in batteries wherever they can.
The second is to accelerate the deployment and integration of distributed energy asset networks. Everything that can generate electricity will be electrified. Allowing these systems to interact with home and grid-scale energy systems will require a variety of new solutions. Collections of "smart" devices such as electric vehicles or thermostats can even form virtual power plants that mimic the behavior of larger energy assets.
What is the future?
A core challenge of grid expansion is to carefully balance the transition between centralized and decentralized systems, taking into account both economic and reliability issues. Centralized grids, while simple and (usually) reliable, also face the problem of complex demand fluctuations and high fixed costs—most of the world’s large nuclear power plants are government-funded, for example. Decentralized grids, while still in the early stages of deployment, are cheap but do not automatically ensure reliable power, as the preferences of some rural communities in India show.
To be clear, the centralized grid we have today will certainly not go away—in fact, it will need to scale up—but it will be consumed by a growing network of decentralized assets around it. Ratepayers will increasingly adopt their own generation and storage, challenging traditional electricity monopolies and driving regulatory and market reforms. This trend toward self-generation will reach its extreme in energy-intensive industries that particularly value reliability—Amazon and Microsoft are already pursuing nuclear-powered data centers, and we should do everything we can to accelerate the development and deployment of new reactors.
More broadly, ratepayers need electricity that is reliable, affordable and clean, often in that order. ERCOT, with its unique location, easy-to-innovate “pure energy” market, and relaxed interconnection policies, will be key to watch to understand if, when, and how this can be achieved with a decentralized grid. There is no doubt that successfully navigating this transition will result in significant economic growth.
Crucially, building this decentralized grid will require our most talented entrepreneurs and engineers: we need a “smart grid” that seriously innovates in pre-user, post-user, and grid software technologies. Policy and economic trends will accelerate this electricity evolution, but the responsibility for ensuring this decentralized grid works better than the old one will fall to the private sector.
The future of America's electric grid lies in harnessing new technologies and embracing free markets to overcome our nation's challenges and pave the way for a more efficient and dynamic energy landscape. This is one of the great endeavors of the 21st century, but we must rise to the challenge.
The world is changing rapidly, and the electric grid must change with it.
