Energy Storage and Solar Power

Perhaps the greatest technical problem facing the increased adoption of solar power is to find ways to meet our expectations of electricity - that it be available upon demand, when we flip the switch - with the intermittent and variable nature of solar production. As critics of wide-scale adoption of solar energy technologies are quick to point out, there is not only every night, but a good portion of many days when the sun doesn't shine. While such statements miss the advances that have been made in accommodating the electricity grid to this natural phenomenon, various forms of energy storage are important for wider use of solar power.

This phenomenon is also true of wind energy, and in many ways wind and solar share similar challenges in terms of integrating more wind into our expectations of electricity production. As a result, much of the research into integrating more solar into electricity production looks at integrating both solar and wind.

Variability and intermittency

But what do we mean by variable and intermittent? These two concepts are important to understanding the challenges of renewable energy generation. Solar is an intermittent power source - as every night, solar photovoltaics (PV), solar hot water and concentrating solar power (CSP) systems do not produce power, unless it is the release of a power storage system that accompanies the system. Furthermore, their output is variable – solar systems produce different amounts of power at different times of the day, and on different days. One important aspect of the variability is the predictability of power output.

Types of demand response

However, just as solar output fluctuates, so does demand from power consumers – not only from day to day and from week to week, but within 24-hour cycles and smaller increments. Grid operators currently have multiple ways of meeting this demand – from turning on and off large coal, nuclear and natural gas plants, to increasing and decreasing rates of generation within plants, to releasing power stored in pumped hydroelectric units and using capacitors to meet instantaneous demand spikes. The type of response naturally varies according to the type of demand. The most difficult and expensive changes to meet are the rapid, short-term fluctuations in energy demand, known as “regulation” demand.

Sample graph of regulation versus load following. Courtesy National Renewable Energy Laboratories (NREL)
Sample graph of regulation versus load following. Courtesy National Renewable Energy Laboratories (NREL)

In order to meet consumer expectations of reliable power on demand, grid operators must also keep enough power production on hand to deal with contingencies both predictable and unseen, including downtime for maintenance and repair of generation units, loss of transmission capabilities due to electric storms and other weather events, and other accidents.

Current energy storage status

Until recently, the ramping up and down of natural gas units, some of which are only turned on when needed, has been used in many places to meet variations in demand. Many of these reserve units are kept operational, as “spinning reserves”. As a result, many nations, including the United States, have not invested heavily in energy storage.

Diagram of pumped hydro system:Courtesy Georgia Institute of Technology
Diagram of pumped hydro system:Courtesy Georgia Institute of Technology

In the mid to late 1970's, the increase in the cost of oil and gas spurred the development of pumped hydro reserves, which grew to 20GW of capacity in the United States; however after gas costs dropped again in the 1980's natural gas combined-cycle units became the preferred way to meet many types of electricity demand. Globally, only three technologies other than pumped hydro are at used in a total quantity greater than 100MW: sodium-sulfur batteries, compressed air storage, and thermal storage. In the United States, the only other large energy storage project besides pumped hydroelectricity is a single 110MW compressed air storage system; the nation uses slightly more than 1GW total power storage.

Many existing concentrating solar power (CSP) plants, also use energy storage in the from of heated fluids that are stored in tanks, to be released later. In some cases, the heat that is generated is transferred to another medium, such as molten salt, for storage.

While large-scale power storage is very limited in much of the world, small power storage is another story. Until the last decade the majority of PV installed was not connected to the electricity grid, with most of these “off-grid” systems using batteries to store and release electricity. The most common technology used is lead-acid batteries, similar to the type used in automobiles.

Alternatives to storage: Other means of meeting the challenges of variable generation

Batteries and other storage technologies, both at small and large scales, add significant costs to PV and other renewable electricity generation. As a result a variety of other strategies have been pursued to minimize the use for storage.

The subject of how much variable renewable energy can be incorporated into a grid without difficulties is the subject of debate, with groups like the American Institute of Chemical Engineers stressing the need for massive electrical storage. However, recent research suggests that larger amounts of renewables than previously thought can be integrated into electricity grids with a combination of better weather forecasting, more rapid scheduling of generation, better coordination between grid operators, and energy markets that utilize resources over larger geographical areas. The International Energy Agency has gone so far as to state that: “There is no intrinsic ceiling to variable renewables’ share from a system integration point of view. The integration potential of a region/country depends on the flexibility of its power system.”

More concrete assessments have also been made. In 2006, the Government of Alberta imposed a cap of 10% of the amount of wind that could be integrated into the power mix, and then lifted the cap in 2007. In 2007, 20% of the Danish electricity demand was met with wind energy, and in that year the Chairman of the West Danish system operator ELTRA stated that even more could be integrated if grid operators were allowed to use the right tools to manage the system. Most recently, in a May 2010 report, the Western Wind and Solar Integration Study, the US Department of Energy's National Renewable Energy Laboratories (NREL) stated that with the proper mix of tools, different grid operation practices and limited new infrastructure, as much as 35% variable renewable energy generation – 30% wind and 5% solar – could be integrated into the grid in states in the Western US with significantly affecting system reliability.

Impacts of variable renewable generation on electricity demand, projected for a typical June and August. Courtesy NREL, Western Wind and Solar Intergration Study.
Impacts of variable renewable generation on electricity demand, projected for a typical June and August.
Courtesy NREL, Western Wind and Solar Intergration Study.

Such research on practical grid integration is ongoing, and currently the Hawaiian Electric Companies, the Sacramento Municipal Utility District, and the University of California at San Diego are all conducting research on greater grid integration of solar power. UC San Diego and the Hawaiian Electrical Companies both have the advantage of conducting research on small grids. UC San Diego's Jacobs School of Engineering uses a “micro-grid” which is powered by a variety of sources including 1.2MW of PV. In Hawaii, electrical grids are limited to individual islands in the chain, meaning that each island – Hawaii, Oahu, Maui, etc. has its own grid. 

Forecasting and scheduling

The International Energy Agency has stated that the greatest challenge with variable renewable generation is not variability, but predictability, or lack thereof. A key emphasis of NREL's "Western Wind and Solar Integration Study" was the need for sophisticated weather forecasting. This has also been an emphasis of research at the UC San Diego Jacobs School of Engineering. 

The NREL study found that with better forecasts could reduce the need for power reserves. However, as anyone who has used a weather report to plan a trip knows, there are currently no perfect weather forecasts. Along with developing more accurate forecasting, NREL recommended that grid operators increase their scheduling of generation deliveries more frequently than every hour, which is currently the practice.

Flexibility

Researchers have also found that the overall flexibility of a power delivery system is very important for integrating variable sources. Flexibility can mean a number of things; that more resources are available to be activated, or that power is imported from other sources in times of low generation.

As has been mentioned before, combined cycle natural gas generation is favored for the ease in which it can be turned on and off and ramped up and down, in a less costly manner than other forms of “traditional” generation, including coal plants.

Geographic coordination and grid operation

Both NREL and European studies have found that coordinating generation over a wider geographic area is important for overcoming the limitations of variable generation. In the case of Denmark, electricity imported from Norwegian hydroelectric generation is used to meet consumer needs in moments of low wind intensity. As weather events tend to be highly localized, even more variable generation can be used to meet dips in output if drawn from far enough away. Put simply, when the sun is not shining one place, it is somewhere else. Other variable resources such as wind can be used on a smaller geographic scale, and German data shows that wind resources often produce at an opposite daily cycle to solar resources.

Electricity trade between Denmark and Norway. Courtesy International Energy Agency.
Electricity trade between Denmark and Norway. Courtesy International Energy Agency

Such geographic balancing of resources requires coordination between grid operators and active “spot-markets” for electricity producers in one region to be able to quickly sell excess power to another region. In some regions, this requires additional transmission capacity. An example is the ERCOT grid in the US state of Texas, which is not connected to the grid in the neighboring state of Louisiana, meaning that utilities in the US Deep South cannot currently take advantage of excess electricity from wind and solar generation in Texas.

Power storage

However, to move beyond these ceilings and limitations, as well as to provide backup power from non-fossil fuel sources in the event of system breakdowns, some form of power storage is necessary. The following section will look both at thermal storage for concentrated solar power (CSP) technologies and massive energy storage.

Thermal storage and CSP

Of all the solar technologies, the one with the greatest advantages in terms of supplying electricity on-demand without such intensive coordination is CSP. CSP uses sunlight to produce heat as a step towards producing electricity, and the advantage of CSP is that this heat can be stored more efficiently and at a much lower cost than electricity, and rapidly deployed.

CSP storage technologies vary according to the medium used, the amount of storage, and other factors, meaning that costs for CSP storage vary as well. Parabolic trough CSP systems typically use thermal oil for a heat transfer medium, however many of these systems, such as the Spanish Andasol 1 plant, store this heat as molten salt. Solar tower systems typically use water or air as the transfer fluids, and the PS10 and PS20 tower systems store heat as superheated, pressurized water. In both cases the stored heat is used to power turbines when sunlight is not available, allowing these plants to be used to supply base-load power and power on demand.

Recent studies have indicated that the addition of thermal energy storage systems makes CSP plants more cost-competitive, even in California, where retail electricity prices are relatively low by global standards. Forecasting electricity demand and solar availability also increases the profitability of CSP plants with thermal energy storage.

Battery storage – utility-scale solutions

Due to their cost, batteries traditionally have not widely been used for large-scale energy storage. Significant exceptions are the use of lead-acid batteries to serve as backup power for internet and communications systems in the United States, and the recent increase in the use of sodium-sulfur  batteries by utilities in Japan and other nations.

Power storage cost comparisons: Courtesy of the Electricity Storage Association
Power storage cost comparisons: Courtesy of the Electricity Storage Association

While batteries represent an additional cost to generation, there are situations where strategic usage of batteries can prevent utilities from needing to construct additional transmission, which also represents large capital costs. In the United States, American Electric Power is using a 5MW sodium-sulfur battery in Southern Texas to alleviate a transmission problem. However, batteries are also seeing increasing use for conventional reasons; American Electric Power has stated that it plans to increase its battery storage capacity to 1GW by 2020.


Battery technologies

Lead-acid batteries

Lead-acid batteries have been in commercial use since 1970 and are the most common type of battery in use today. The use of lead-acid batteries for commercial, industrial and automotive use was a USD$2.9 billion industry in 2008, and is growing by 8% annually.

Lead-acid battery diagram. Courtesy Florida State University
Lead-acid battery diagram. Courtesy Florida State University
Sodium-sulfur battery bank. Courtesy of Hawaiian Electric Companies
Sodium-sulfur battery bank. Courtesy of Hawaiian Electric Companies

The technology of lead-acid batteries is uncomplicated and manufacturing costs are low; however such batteries are slow to charge, cannot be fully discharged and have a limited number of charge/discharge cycles. The lead and sulfuric acid used are also highly toxic and create environmental hazards which can be particularly ironic when used to accompany “clean” sources of power such as PV. A number of companies are currently working on improvements in lead-acid battery design, often incorporating super-capacitors.

Sodium-sulfur (NaS) batteries

Sodium-sulfur batteries are a relatively new technology that has emerged as the energy storage option of choice to accompany new wind and PV generation. 300MW of sodium-sulfur batteries are currently in use on the electric grid worldwide. Many of these batteries are being used by utilities in Japan and other nations and US utilities are increasingly investigating the use of these batteries for larger applications.

Sodium-sulfur batteries use molten metal, and operate at temperatures above 250 degrees celsius. They have very high power densities and work well at storing large amounts of power, with a long cycling lifetime of 15 years. They are also relatively inexpensive to manufacture.

Utility-scale sodium-sulfur batteries are manufactured by only one company, NGK Insulators Limited (Nagoya, Japan), which currently has a 90MW per year production capacity. Because of the demand for sodium-sulfur batteries to accompany variable renewable generation, the company is currently building a new 60MW production line, which it expects to be in place by October 2010. In June 2010 NGK told the Wall Street Journal that it is considering building another 60MW production line to be completed by the summer of 2011.

Lithium-ion (Li-ion) batteries

There are three types of lithium-ion batteries in commercial use: cobalt, manganese and phosphate. Lithium-ion batteries are widely used for portable electronic devices such as cell phones and increasingly for portable power tools, but have not shown themselves to be economical for most utility applications. When lithium-ion batteries are used for utility-scale applications, it is to perform regulation and power management services, and are used for minutes of runtime.

Lithium-ion batteries. Curtesy of the Zentrum für Sonnenenergie- und Wasserstoff- Forschung Baden-Württemberg
Lithium-ion batteries. Curtesy of the Zentrum für Sonnenenergie- und Wasserstoff- Forschung Baden-Württemberg

Other types of conventional batteries

There are a number of other battery technologies currently in use in other applications that are either in decline or have not shown themselves to be useful for such applications. Nickel cadmium batteries were widely used for consumer electronics in 1990's and first decade of the twentieth century, but have gradually been replaced by nickel metal hydride and lithium-ion batteries. Nickel metal hydride batteries have found uses in a number of hybrid electric vehicles as well as consumer rechargeable batteries, but have lower energy densities than lithium-ion batteries.

Flow batteries

Flow batteries store energy chemically in liquid electrolytes, which are pumped through cells in the battery when electric current is flowing, with two different electrolytes for the positive and negative charge separated by a thin membrane through which selected ions flow. Flow batteries provide high power and high-capacity batteries, and the amount of power in kWh generated by these batteries increases with the quantity of electrolyte, with virtually no upper limit. As they also have fast response time, flow batteries can be useful for a wide variety of utility-scale applications.

There are two types of flow batteries: hybrid and redox (oxidation-reduction). In hybrid batteries, one of the electrodes is reactive; in redox batteries, all active components are immersed in the electrolyte. In such batteries the electrodes do not change chemically or physically, and thus they can be cycled for a very long period of time. Zinc-bromide batteries, a type of hybrid flow battery, are currently the most widely used flow battery technology used in the United States.

Vanadium-redox batteries

The vanadium-redox battery uses vanadium salts and sulfuric acid in the electrolytes. Vanadium is present in both positive and negative electrolytes in different oxidation states. These batteries have the same advantages of large potential power capacities as other flow batteries, but also suffer no permanent damage from mixture of the electrolytes.

Vanadium-redox battery diagram, courtesy Cellstrom.
Vanadium-redox battery diagram, courtesy Cellstrom.

Researchers at the Fraunhofer Institute are currently working on improvements in the vanadium-redox battery. Vanadium-redox batteries have been explored for potential use in electric cars, with the used electrolyte pumped out and replaced with recharged electrolyte at filling stations.


Other energy storage solutions

Compressed air

Compressed air systems can be seen as either a type of storage or a type of natural gas generation. Compressed air plants store their air in natural geological formations, such as underground caves. When released, the compressed air is mixed with natural gas to run turbines which use 40% less fuel than conventional natural gas turbines.

Supercapacitors and flywheels

Supercapacitors and flywheels are used in modern utility applications to meet regulation demand, the short-term spikes in electricity use that are part of supplying electricity to large numbers of consumers. They are generally not seem as solutions to other utility-scale storage problems.


Conclusion

Even with improved scheduling, forecasting, coordination of generation and power system flexibility, it will at some point be necessary to make greater investments in power storage solutions to accommodate renewable generation if solar and wind generation are to make up more than a third of electricity generation in a given geographical area. Fortunately, several promising technologies for power storage either exist or are being developed.

Sodium-sulfur batteries, courtesy Honda Motor Company Limited
Sodium-sulfur batteries, courtesy Honda Motor Company Limited

Sodium-sulfur batteries are currently the optimal choice for large scale power storage to accompany renewable energy generation. However, given the rapid changes in battery technology in recent decades, it is quite possible that a new technology or a development on existing technology will replace sodium-sulfur batteries in the near future.

2010-07-07  |  Christian Roselund