Mind the “storage” gap: how much flexibility do we need in a high renewables future?

Source: By Brendan Pierpont, America's Power Plan • Posted: Tuesday, June 20, 2017

Imagine for a moment that we have built enough wind and solar power plants to supply 100 percent of the electricity a region like California or Germany consumes in a year. Sure, the wind and sun aren’t always available, so this system would need flexible resources that can fill in the gaps. But with continuing rapid cost declines of wind, solar, and batteries, it’s possible that very ambitious renewable energy targets can be met at a cost that is competitive with fossil fuels.

Every region has a different climate and demand profile. Taking California or Germany as an example, and assuming no interconnections with neighboring regions, up to 80 percent of the variable renewable power produced could be used in the hour it is generated with the right mix of wind and solar – in other words, 80 percent of supply could be coincident with demand. Still, a reliable grid needs fast-responding resources to satisfy the remaining 20 percent of demand; filling this gap is one of the principal flexibility challenges of a low-carbon grid. But what will that flexibility cost?

The answer is surprising – by 2030 an 80 percent renewable energy system including needed flexibility could cost roughly the same as one relying solely on natural gas. As Climate Policy Initiative demonstrated in our recent report Flexibility: the path to low-carbon, low cost electricity grids, if prices for renewable generation and battery storage continue to fall in line with forecasts, meeting demand in each hour of a year with 80 percent of electricity coming from wind and solar could cost as little as $70 per megawatt-hour (MWh) – even when accounting for required short-term reserves, flexibility, and backup generation. Of course, this analysis makes some simplifying assumptions; it represents the new-build cost of generation and flexibility to meet demand in every hour of a year using historical wind, solar and demand profiles from Germany, and it doesn’t factor in transmission connectivity or model the constraints of existing baseload power plants in detail. But it also leaves out the significant potential for cheaper flexibility from regional interconnections, existing hydroelectricity, and the demand side.

Still, this analysis helps us understand what kinds of flexibility we will need and what it will cost. The promise of a low-cost grid based on wind and solar is so compelling, it’s worth digging into what we’d need to do to realize this vision.

What is flexibility, anyway?

A power system has a wide variety of flexibility needs – with time scales ranging from seconds to seasons – and a range of different technology options can be used to meet those needs, depending on the time scale.

On very short time frames from seconds to minutes, fast-responding resources are needed to keep the grid in balance and compensate for uncertain renewables and demand forecasts. These needs should grow only modestly as shares of renewables climb to high levels, and they could be accommodated cheaply using existing hydro generation (where it exists), or even smart solar and wind power plants. Fast-responding demand response or energy storage would also be good choices, particularly after storage costs decline further as projected.

Solar and wind output can change rapidly on a predictable, hourly basis as well, requiring flexible resources that can quickly pick up the slack. One feature of California’s now-infamous “duck curve” is the need for fast-ramping resources to meet the evening decline in solar production. California has devised innovative market mechanisms to ensure flexible gas and hydro generators are available to meet these ramping needs.

On a daily basis, the profile of renewables production doesn’t neatly match demand, requiring resources that can store or shift energy, or otherwise fill in the gaps across the day. Today, daily imbalances are met primarily by dispatching fossil fuel fired power plants. But a number of solutions are gaining momentum, such as automatically shifting when consumers use energy and building large batteries.

At even longer time frames, there can be multi-day and seasonal mismatches between when renewable energy is produced and consumed. The need for long-term, multi-day energy shifting – exemplified by several windless, cloudy winter days with high electric heating demand – is perhaps the biggest challenge to complete decarbonization of the power grid, because batteries are ill-suited to seasonal shifting needs. In fact, using lithium ion batteries for seasonal storage, cycling once per year, would cost tens of thousands of dollars for each MWh shifted.

Graphic: Technology fit with flexibility needs

The challenge of power grid decarbonization hinges on this ability to store or shift energy. But how much energy would the power grid really need to shift, and over how long?

Solar drives daily storage needs

A power system that relies primarily on solar would have abundant power in the middle of each day, and scarcity during the night. Trying to exclusively power the grid with solar, with no ability to store or shift energy, would mean more than half of demand would go unmet.

Many technologies are well-suited to shifting energy within a day. Today solar generation relies on dispatching hydro and thermal power plants to meet changing demand, but in the future, lithium ion and flow batteries promise multiple hours of storage and shifting capability. Thermal energy can be stored in buildings, shifting when electricity is used for heating or cooling. And as electric vehicles become more widespread, ubiquitous charging infrastructure, electricity pricing and automated charging could shift when drivers charge their vehicles.

But are the daily energy storage and demand-shifting solutions emerging today going to be enough? Well, it depends.

In California, demand is highest during the summer, when solar production is at its peak. If California could store and shift solar energy to any time in each day, solar could meet nearly 90 percent of California’s electricity demand. Only 10 percent of energy demand would go unmet by solar because of multi-day and seasonal storage gaps.

In Germany, however, demand is highest during the winter months, driven in part by electric heating demand. So storing and shifting solar energy within each day would still leave 30 percent of energy demand unmet. In other words, the long-term storage gap for solar in Germany is three times larger than in California.

Wind drives storage needs of up to a week

Wind, on the other hand, is a better match with demand hour-by-hour, with 70-80 percent of wind production coincident with demand in California or Germany. And compared with solar, daily storage has little value for wind. Shifting energy within the day could only improve wind’s match to demand by a few percentage points. For wind, the biggest gains come from shifting energy by up to a week. In both California and Germany, the ability to shift energy by up to a week could allow nearly 90 percent of energy demand to be met with wind.

Beyond a week, seasonal storage needs depend on regional demand and renewable resource profiles, and, critically, what mix of renewable resources the region has installed. A system incorporating both wind and solar can have lower storage needs than a system based predominantly on one resource or the other. In Germany, a mix of 70 percent wind and 30 percent solar could meet 90 percent of demand on a daily basis, reducing the need for longer-term storage. In California, solar is already a pretty good fit for seasonal energy needs, but the addition of around 10 percent of electricity from wind could slightly lower both daily seasonal storage needs in California.

Graphic: Storage gap for 100 percent wind or 100 percent solar in California and Germany

Graphic: Storage gap for a wind and solar mix that minimizes long-term storage needs in California and Germany

But far fewer technology options allow for long-term energy shifting. Consumers can’t go for a week without heat, cooling, or charging vehicles, and many long-term storage technologies like hydrogen are still too costly and inefficient for widespread use. The default option for long-term storage is a familiar one – rely on fuel-burning power plants that provide flexibility to today’s power systems. Finding cheap, reliable and carbon-free ways to shift energy for periods longer than a week may be the key decarbonization challenge.

So how should we approach the seasonal storage gap?

Policymakers and planners have several strategies they can use to bridge the storage gap:

  1. Target a mix of renewable resources that minimizes long-term storage needs. Procuring the right mix of resources can be the easiest way to reduce the seasonal storage gap.
  2. Connect with neighboring regions to trade surpluses and shortfalls of energy. Northern Europe and the Western U.S. are taking steps to better integrate regional grids, although getting neighboring states and countries to cooperate can be challenging.
  3. Make use of existing hydropower. Regions with abundant hydroelectricity may already have enough existing flexibility to completely satisfy seasonal storage needs. But electricity and ecological needs don’t always align, and drought years could spell trouble for grid reliability.
  4. Make industrial demand seasonal. Paying the fixed capital and labor costs of an electric arc furnace for several months of the year while a steel foundry lays idle may in fact be cheaper than building the storage or generation needed to meet that demand carbon-free year-round. But this solution would require a careful balancing act between maintaining industrial competitiveness, complying with trade agreements, and ensuring job stability for workers.
  5. Develop long-term storage technologies to shift energy across weeks and months. Turning renewable electricity into hydrogen or synthetic natural gas can enable longer-term and larger-scale storage, and can be used directly for transportation, heating and industry. But so far, these conversion technologies are inefficient and expensive.
  6. Develop flexible, dispatchable carbon-free power plants to cover shortfall periods. A recent survey of decarbonized grid models suggested that nuclear and carbon capture and storage may be needed to completely decarbonize the grid. But market models and technologies will need to evolve for these resources to operate flexibly and profitably.

Transitioning to a low-carbon grid

A low carbon grid is the lynchpin of any serious plan to avoid the dangerous impacts of climate change. And with solar, wind, and energy storage costs dropping year over year, the vision of a low-cost, flexible grid driven by renewable energy seems tantalizingly within reach. But if we are going to fully decarbonize the grid, the long-term storage gap is one of the biggest challenges that lies ahead. We already have many of the technologies and tools we need for this shift, but our electricity policies and markets need to evolve for a new generation of technologies with different cost and risk profiles. If we start laying the groundwork today, we’ll be ready to keep pace with the rapid transition ahead.


Brendan Pierpont is a Consultant with the Energy Finance team at Climate Policy Initiative