Renewable Energy and the Need for Storage: Lessons to Be Learned from the Situation in Germany

Posted on October 15, 2015

Intermittent renewables represent the largest share of newly built energy generation capacity into the foreseeable future. The US Clean Power Plan requires all states to source 28 percent of their power from renewables by 2030 (12% from non-hydro renewables). California, usually in the lead on “green” policy initiatives among US states, requires 50% by 2030. 28% is the current annual contribution from renewables to Germany’s power mix, with a recent record peak of 78% on July 25, 2015. IRENA’s 2015 Energy Storage Roadmapnow calls for a tripling of energy storage by 2030 in order to enable a larger share of renewables on the grid. Increased storage, then, appears to be key for integrating renewable energy sources in the coming years.

“Not so!” claim two German studies, both completed in 2014 and both with the participation of Aachen University (RWTH), which has developed an optimization model for European power generation stations and markets for every hour of the year. Examining the situation in Germany, the authors controversially claim that no, or very little, new storage will be required to build a grid that is powered almost entirely by renewable energy sources.

Germany’s Energiewende (probably best translated as “energy paradigm shift”) entails a complete transformation of the electricity system: not only is there a rapidly increasing share of intermittent renewable power sources on the country’s electricity grid but the government’s ambitious plans foresee a nuclear exit by the year 2022. So should Germany not be the number 1 market for storage facilities in order to balance out intermittent power generation with fluctuating demand?

of the two studies was commissioned by Agora Energiewende and found that no additional storage is currently needed in Germany (note that 5.4 GW of pumped storage – 15 facilities – already exist in Germany, p.100), even if the amount of intermittent renewables were to increase to 60% in Germany (by 2030) and to 40% in the rest of Europe. This study determined the economic feasibility of storage solutions and finds that only at very high penetration levels (e.g., 90%), additional storage will definitely be required. Even then, no scenario indicates the need for more than 10 GW of long-term storage capacity, whereas short-term storage is of lesser importance and would only be economical if costs are driven down rapidly. The report indicates that other options, such as curtailing the output of intermittent power sources during periods of overproduction, are technically sufficient to maintain grid stability even with very high penetrations of intermittent resources. The report concludes,

“To achieve the mind-term goals of the Energiewende, does not depend directly on the addition of power storage options… In the longer term, if the share of flexible energy generation options (such as solar thermal or biogas) increases and if demand can be managed flexibly, then even high renewable energy penetrations (about 90% in Germany and over 80% in the rest of Europe) can be balanced by the power system without the addition of new storage options” (p. 45 of the German summary).

In the same vein, the second study, led by the renowned Fraunhofer Institute, concludes that “Up to a renewable electricity share of 60%, the addition of power storage devices is not a condition for the addition of solar PV and wind power plants… Even at high degrees of penetration (90% in Germany), the required balancing can largely be achieved without additional power storage” (p.13).

Both studies point to the expansion of the German and inter-European transmission network, demand-side management and increased flexibility on the generation side as cheaper alternatives that can be harnessed to manage the future renewables-dominated electricity grid without the more expensive storage options. Further support for a high penetration level before batteries are needed comes from Adam Reed of the Renewable and Sustainable Energy Institute in Boulder, Colorado, who also sets the threshold at no lower than 60% in his course material on renewable energy integration (offered by Leonardo Energy).


Or so one could summarize the reaction of dena, the German Energy Agency, a for-profit think tank  created by a consortium including Deutsche Bank and the federal government, to support the Energiewende and inform policy-making. The chairman of dena’s management board, Stephan Kohler, asserts that “energy storage is indispensable for the Energiewende – whoever says otherwise, damages the Energiewende and ultimately jeopardizes the reliability of Germany’s electricity supply.” dena’s main concerns are that a) the expansion of the German transmission system will not proceed as quickly as required to keep up with the growth of intermittent renewables, and b) demand-side management options are at best uncertain means to manage the grid and cannot therefore be relied upon as a substitute for increased investment in storage. Furthermore, whereas fossil fuel-fired power stations can have several weeks of reserves (gas storage) or direct access to fuel (local coal), there is no such strategic reserve in case of a protracted lack of energy supplies with solar or wind energy. Furthermore, there are concerns that the European power market will not operate as one, leading to bottlenecks whenever transborder power balancing is required.

International power flows do already present some problems, as for example Poland is complaining about some unwelcome grid fluctuations coming from its German neighbours (a recent German transmission project is supposed to ease the situation in Poland) – the media even speak of “power wars” breaking out over Germany’s unabated move towards more irregular power generation events. The emphasis, then returns from European integration back to national transmission flexibility.

Some of these limitations are also mentioned by the authors of the two controversial reports: the expansion of the transmission system as planned, for one, is seen as a crucial element of further renewables integration. Both increased demand management capacity and the flexibilization of e.g. cogeneration units, biogas and solar thermal facilities are identified as conditions for avoiding increased storage on the grid (see p. 109 of the second study). Likewise, the addition of large numbers of electric vehicles without flexibility options may lead to increased storage needs. The report even recommends battery storage systems in order to provide some systems services, such as primary balancing (p.110).

One question, which the two German reports answer in the negative, is whether special policy support for energy storage is required. It appears that this may indeed not be the case since this sector is already growing all by itself: “The behind-the-meter market segment of energy storage is widely expected to undergo a similar boom to the solar PV industry, with a tipping point expected within the next ten years as further cost reductions are achieved," states the recent AECOM report, which also predicts cost reductions around 40-60% between now and 2020. This coincides with Bosch’s announcement to produce car batteries at half the cost by 2020, while already offering home-based systems in Germany today. With companies such as British Upside Energy developing tools to integrate thousands of small, distributed storage units (batteries and thermal storage) into virtual power plants, the opportunities for distributed storage to play its part seem to be multiplying.

Indeed, Germany is already expanding both its storage capacity and increasing flexibility in its generation pool. For example, Stadtwerke Kiel is building a 190 MW flexible power plant, composed of twenty Jenbacher gas engines – this capacity can be called upon within minutes to balance the high wind power penetration in Germany’s northern grid. Renewable energy plants will shortly be allowed to provide grid stability services in terms of secondary and tertiary balancing. Batteries are also being added for primary balancing – Europe’s largest energy storage project (10 MW/11 MWh) just opened in Feldheim in Brandenburg state (near Berlin), using Li Ion batteries from Korea. The installed battery capacity for primary balancing in Germany has increased from about 1 MW in 2012 to roughly 27 MW in 2015, and will keep on rising, according to Germany Trade & Invest projections. In part due to a 30% capital cost subsidy, up to 13,000 residential battery systems are predicted to be installed across Germany by the end of 2015. A recent Frost & Sullivan report sees Germany as the fourth largest market for storage systems over the coming decade. And given battery costs are dropping fast, it seems storage will come no matter what.

So what does this all mean for North America? The US, for one, have been much more trigger-happy with trying and installing battery systems in recent years. They are the number one market according the above-mentioned Frost & Sullivan report and the world’s largest battery storage system of 100 MW is, of course, planned in the States, with several systems between 15 and 30 MW already in place. Famously, the State of California mandated the installation of 1.3 GW of storage by 2020. So, are the States getting ahead of themselves by installing storage that isn’t actually needed?

One answer probably is that no man is an island. Or, at least, Germany and North America aren’t. Because the more an electricity grid resembles a small island grid, the more the fluctuations of intermittent renewables will be felt and the more need, therefore, for battery storage. According to Agora’s Country Profile, Germany (with a population of about 80 million) had 21.3 GW of available interconnection capacity in 2012 - a high level compared to an annual peak demand of 83.1 GW. To compare, the (Eastern US) PJM Interconnection grid (serving a population of about 60 million) had a peak demand of 127 GW in 2009 and an interconnection capacity of about 23.5 GW. The California grid (CAISO, about 30 million) had a peak demand of 46 GW in 2009 with an import capacity of 10 GW. As such, the California situation appears to be similar to Germany, with a regional intertie capacity of around 25% of peak demand, whereas PJM shows less flexibility in terms of regional power trading.

Germany already has considerable storage resources, i.e. over 5 GW of pumped storage and some compressed air and battery storage. The growth of storage installations is modest, with Energy Storage Updateconcluding that ”the grid-scale market for energy storage appears to be evolving more slowly in Germany than in Italy or the UK. This is probably because an abundance of grid connections to neighbouring countries has so far enabled Germany to balance increasing renewable energy generation by exchanging it with other European markets.” This stands in strong contrast to California, which currently has almost 4 GW of pumped storage and is seeking 1.3 GW of storage by 2020.

Maybe California’s 10 GW of intertie capacity are of little use if inside the state, electricity cannot flow freely. Burbank Water and Power (BWP) is planning to build a 300 MW CAES facility – a very large number for its service area. But the service area does actually resemble an island, based on what their power resources manager says: “BWP is responsible for its own Area Control Error, has contractual transmission rights and “cannot socially spread” the intermittency of renewables in its territory across the ISO market.”

California’s famous “duck curve” (exacerbated by rapidly increasing solar PV installations) illustrates that not all locations are created equal, and simply adding up MW of intertie capacity may not be enough to determine the need for storage – both the local load curve and the (physical and regulatory) ability to balance power over large geographic areas are significant factors as well. Given the difficulties of realizing effective regional and international cooperation and grid management tools, as well as the long planning timelines and high investment for new transmission capacity, storage options may simply turn out to be a good medium-term solution, even though other options could provide the same services. And with batteries becoming ever cheaper very rapidly (IRENA’s Energy Storage Roadmap mentions a potential cost of only 6 cents (US) per kWh by 2020 for utility-scale storage, p.5), the cost argument is weakening year by year. Whether existing trends to add storage at the distributed level (PV/storage and vehicle batteries) may make centralized storage redundant remains unclear, but can also worsen the situation unless batteries can be harnessed to provide grid services, such as voltage, frequency and ramp control.

Yet, with several time zones across the continent, North America has a great opportunity for not only north-south but especially east-west balancing between regional grids. Utility and regulators’ strategies should, by consequence, include cautious investment into storage but should also tackle the legal and infrastructure bottlenecks that keep electricity markets from providing regional and continental balancing between electricity grids. Absent the success of such longer-term measures, storage, more flexible generation options, and demand management are more immediate means to integrate ever-increasing amounts of intermittent power sources on the grid.

Authored By:
Martin Tampier, P.Eng. is an energy consultant based in the Montreal area working under the name of ENVINT Consulting. Martin is a senior renewable energy expert and provides advice to all levels of government on conventional and alternative energy technologies from algae to wind power. His expertise stretches from technical and economic analysis of commercial and emerging technologies over emissions and GHG management to providing advice on project funding applications

Related Posts



October, 20 2015

David Ciepluch says

Martin, have you run across any load storage plan that is decentralized on the consumer side? Thermal mass storage for heat in electric furnaces, heat pump water heaters, ice making AC equipment, and eventually plug in electric cars, and utility load control timing, offer utilities a rather low technology storage system capability. Equipment cost is about 1.5 times that of more traditional heating and cooling technology but that gap can be closed with economies of scale.

I know utilities would like prefer to maintain their hold on the production side of energy and the prices. The model I suggest could hold capacity of thousands of megawatts of stored energy and even would have a tendency to flatten the load curve.

October, 20 2015

Martin Tampier, P.Eng. says

I mention British Upside Energy in the article, who are attempting to create virtual power plants by controlling customer-side storage devices (heat and electricity). Stem in the US is trying something similar:

October, 23 2015

Jack Ellis says

In the US, NREL has performed some studies that suggest little or no additional storage may be required in the Western Interconnection at penetration levels substantially higher than today's levels. I'm not so sure. I'm also not so sure it's technically or economically feasible to get to California's 50% level by relying primarily on wind and solar, both of which create complications for the grid operator with their variability and uncertainty.

However any attempt to achieve high penetration levels of renewables depends on a number of complementary actions, including flexible demand, coordination across market areas (countries in Europe, Balancing Authority Areas in the US), a lot more storage than we have today, controllable, mostly fossil-fired plants that are more flexible than the existing fleet, and a more geographically and technologically diverse renewable supply. Not just a few of these actions but all of them. Storage will be required not only to store surplus renewable production for later use, it will also be required to help with balancing at a number of time scales, and it is likely that separate sets of storage devices with different and mutually exclusive attributes will be required depending on the balancing time frames. Moreover, the amounts will be measured in many tens of GWH for bulk storage and at least in the high single digits of GW for short-term balancing, and the cost of all this infrastructure will be enormous, even if battery storage is cheap.

I haven't seen the studies done in Germany but I'm highly skeptical about the conclusions.

Jack Ellis, Tahoe City, CA

October, 26 2015

Martin Tampier, P.Eng. says

Jack, I tend to agree with you. The German reports found that "all of the above" measures other than storage are cheaper than storage, and could provide the required flexibility. In reality, though, the European market is not currently as flexible as it would need to be. Certain assumption were made as to how this will change in the future, and if it does not (or not as quickly) then storage will most likely be required (as the authors admit). As I mention, California does not seem to be very well integrated (fairly well in terms of capacities but apparently not at the transaction level, which limits the ability to move electricity around freely), so unless that changes, it will be difficult to get by without more storage. As said, the more your system resembles a small island, the more storage you will likely need.

Add your comments:

Please log in to leave a comment!
back to top

Receive Energy Central eNews & Updates


12th Annual Electricity Flexibility, Ancillary Services and Balancing Forum

Wednesday Feb 19, 2020 - Friday Feb 21, 2020 - Vienna Austria

This marcus evans conference plays a crucial part in a strategy that involves keeping ahead and gaining the decision-making edge for system operators, power distribution companies and ancillary services in the energy industry. more...