Solar companies are quickly solving all the technological problems except for one. The sun doesn't shine at night. Costs of panels and installation are plummeting but the technology still faces a major scaling issue. Solar cannot cost effectively store energy generated during the day for consumption at night. Luckily, solar has odd an odd ally who has it in their best interest to should solve this problem: fossil fuel utilities.
Critics of solar often complain about solar power’s intermittency problem. They claim the technology does not work because the panels only provide intermittent power. These critics prefer traditional power sources because those sources can produce electricity on demand.
Despite these boasts, traditional sources of energy suffer from an intermittency problem of their own. These sources may generate power on demand, but that power comes with the caveat of immediate use by consumers. While on-demand generation coupled with immediate use does not sound like an intermittency problem, the illusion only lasts so long as you ignore the capital costs behind the energy and cycle of demand over time.
Utilities must build systems that can satisfy the maximum potential demand. This means that utilities must build enough capacity to satisfy demand during customers’ peak hours of consumption.
Peak consumption usually occurs in the middle of the day. The lowest valleys in the consumption cycle come in the middle of the night.
Power plants are working at their most efficient levels during hours of peak consumption. The plants are built to scale and meant to generate large amounts of electricity. During these hours the plants are generating and as much energy as they can. During the consumption valleys the plants are operating at sub-optimal efficiency, wasting electricity, or in the case of newer plants, shutting down.
The newer plants, like natural gas, are the lucky ones. These plants can fire up and shut down rapidly and efficiently. Meanwhile old coal fired power plants take a long time to fire up and shut down and do so inefficiently. Some sources, like nuclear do not even have the options of turning on or off.
While newer plants save costs by shutting down, they are still not maximizing efficiency. Plants do not pay off any of the capital costs of the plant during downtime. Thus, at best these plants only save on the variable costs. This means that traditional power sources have intermittent ability to operate efficiently.
The cost to prepare for this intermittent efficiency even caused some states and utilities to turn to real time energy pricing. Real time energy pricing allows utilities to change prices as demand varies throughout the day. The utilities offer low prices during the night and high prices during the day. These policies hope to encourage consumers to shift certain activities from peak hours to the valleys.
Utilities may have a more creative and beneficial way to solve this problem. If utilities could generate and store excess electricity at night, then they could resell that electricity at a higher price during the day. This could allow utilities to build fewer plants and have them operate efficiently around the clock.
Thus, current power plants suffer from the inverse of solar power’s current storage dilemma. Traditional power benefits from storing energy generated at night for daytime use. Meanwhile solar benefits from storing energy generated during the day for nighttime use.
Currently solar power has an uphill battle to make this battery technology profitable. For batteries to make sense the solar panels have to generate excess energy during the day (at the same time demand is peaking) and store it for night time use (when utilities already offer their lowest rate). This means that solar companies need to develop battery and storage systems efficient enough to compete with utilities’ lowest daily rates.
Utilities face a much different situation. For battery power to make sense for a traditional utility, they only have to develop systems that can store energy cheaper than the difference between the daily rate variations.
Most new power plants in the U.S. are natural gas plants. Even more so for plants built just to handle peak demand (thanks to gas’ on/off efficiency). If most of the difference in price between day and night rates comes from the additional cost of capital for plants used to satisfy peak demand, then most of the future variation in real time pricing comes from the cost of natural gas plants. Thus, it makes sense for a utility to opt for energy storage over energy generation to satisfy peak demand if the price of storage plus generation from existing infrastructure during off hours falls below the capital and operational cost of a new gas powered plant.
Regardless of the current prices and economics behind energy storage, traditional generation sources are much closer to making them cost worthy or even already there just based on the nature of the market. However, it still may take time and investment to scale storage to the economic tipping point.
However, utilities also have another stake in the storage game that may justify this investment. Some day in the future utilities must handle a coming “Alternative Flip.” The “Alternative Flip” references the point in time when renewable energy becomes so dominant that traditional sources become the alternate. At this point in time natural gas plants do not fire up during the day to supplement a continuously burning coal plant. Rather, natural gas plants fire up at night to compliment their solar rivals who have completely displaced the need for the coal plants during the day.
Once solar has displaced the majority of traditional daytime generation it becomes the solar company’s problem to figure out a way to store energy cheaper than a natural gas plant can generate energy. After all, at this point natural gas just acts like nature’s storage mechanism for energy needed at night. Thus, the structure of the market solar must compete with changes and the choice for solar becomes relying on someone else’s natural gas or building their own storage. Thus, when the utility had the choice of building their own natural gas or their own storage, now the solar companies have the choice between their own storage or using some else’s natural gas. The flip changes the factors in solar power’s energy storage equation.
Also, not only does the equation change, the numbers change. Right now traditional utilities sell most of their power during the day and need to recoup more money per electron than they do at night. However, after the flip, most of a traditional utility’s generation and sales will happen at night. This means the flip causes the peak demand for traditional utilities to switch to night and the valleys to switch to day. If this happens, the utilities may have to flip their prices and charge higher real time rates at night. Thus, the numbers in solar power’s storage cost equation changes because they get to compete against traditional energy’s highest prices rather than the lowest prices.
This future leaves current utilities with two options.
The first option involves building new gas power plants to satisfy peak demand and replace antiquated facilities. However, these plants are ripe for solar to displace with their own energy storage systems in the future.
The second option calls for these utilities to build storage capacity capable of transferring excess electricity from off hours to peak hours. This option allows utilities to meet their current demands while also building infrastructure that retains value after solar power becomes the dominant source of energy.
Solar companies are slowly building advantages in energy production through technology and learning by doing. Eventually these advantages will grow large enough for them to displace traditional energy sources. The question a traditional energy provider face is whether they should build infrastructure that will be as useful as a typewriter in a couple decades, or if they should invest in the intellect and infrastructure that allows them to play a role as a linchpin in the energy provision and distribution system regardless of means of production.
 Before hearing comments attacking the specifics in this post I want to acknowledge that this article oversimplified some issues such as rate making (The higher/lower costs of storage would likely just get factored into the variation in energy prices, thus creating or destroying profit margins), the structure of utilities (horizontal vs. vertical), and the market utilities compete in (regulated monopoly only competing to some extent with distributed energy, energy efficiency, and sometimes other utilities participating in regional exchanges).
The largest misconception this oversimplification may create is a dichotomy between the utility and some “other” solar company. In most cases, the choices between solar, storage, gas, etc. will happen within utilities. However, the theme of this post aims to help demonstrate the case about the long-term trends that utilities should consider when building their infrastructure (and what investors should consider when selecting which projects to put capital behind and what risk is acceptable for those projects).
The article also ignores getting into the details of capital financing for utilities and looking at how the accounting for power plants with expected 40 year lifetimes compare to the costs of storage and what the timetable for declining storage efficiency etc.
Some of these flaws are addressed by the fact that investing in storage may give first adopters the ability to counter overcome some of these issues. For example, utilities may not directly compete with each other, but if a California utility begins investing in storage now why a Nevada utility invests in gas plants, then the California utility may have IP/knowhow advantages they can sell to Nevada after the “Alternative Flip.” This creates some form of competition in wider energy marketplace that extends beyond a utility’s service area. The same IP/knowhow advantages also apply to some of the accounting errors overlooked, as the depreciation of the physical assets may not take into account the depreciation of the intellectual assets. This is probably less true for hard IP like patents that someone can put a price tag on, and less true for IP like efficiency from “learning by doing.”