PHS is the most widely used and technically matured bulk electrical energy storage technology. It currently accounts for approximately 98% of all global energy storage installations. PHS is based on the hydroelectric principle and hence does not have issues associated with SOx, NOx, greenhouse gases, and particulate matter emissions. PHS stores excessive electricity as the potential energy of raised water against gravity. Due to the use of efficient water wheels or hydraulic turbines to extract the kinetic energy of moving water and convert the energy into electricity through generators, it can provide ancillary services with better part-load efficiency, better controllability, lower maintenance costs, and minimal start-up costs (Succar and Williams, 2008). These advantages enable PHS (as well as hydroelectric generation) to play a central role in the operation of large-scale electricity grids, in particular when there is a large share of renewable electricity such as solar power. As a stabilizer for electricity grids, PHS systems could ensure that solar electricity supply is both regular and steady.

The major drawback of PHS technology lies in the fact that suitable and available sites for water reservoirs, essential for a PHS plant, are scarce, and in most cases are far from solar power plants. As a result, some new ideas have been developed in the past decade to address the issues. The following are two examples:

• Underground PHS uses underground caverns or water structures as lower reservoirs (Gonzalez et al., 2004). Conceptually, this seems a logical and sound solution for energy storage. However, the practical design and actual construction of large underground PHS systems are highly challenging. Consequently, this is still in the stage of feasibility study, and no project has ever been built (Crotogino et al., 2001; Schulte, 2001).

• Ocean PHS uses the high-elevation coastal region as the upper reservoir to retain ocean water, and the ocean itself as the lower reservoir. Challenges associated with this technology include reservoir sealing and corrosion of the pump turbine (DIT, 2004). Despite the challenges, there is already an ocean PHS in Okinawa, Japan. This plant, the first in the world, has been in operation for more than 10 years (EPRI, 2012).

Other recent developments in the area of PHS technology are the introduction of adjustable-speed motor/generators. Conventional PHS plants with a single-speed machine can only provide frequency regulation in the generation mode. With adjustable speed, it is possible to operate in pump mode as well in the range of 60-100% of rated capacity (Pimm and Garvey, 2009). This means an adjustable speed unit can change pumping power and provide load following and frequency regulation in both the generation mode and the pumping mode, making the technology more attractive in absorbing fluctuating solar power. Furthermore, with conventional single-speed machines, the pump turbines can only achieve their peak efficiency in one of the two modes, not both, whereas the adjustable technology can achieve both. Two Japanese companies, Toshiba and Hitachi, are leading the design and manufacture of high-power adjustable-speed motor/generators. In fact, eleven adjustable-speed pumped storage units have entered commercial service, with nine in Japan and two in Germany (Havel, 2011; Lima et al., 2004).

2.4.2 Compressed Air Energy Storage

CAES is a commercially available technology capable of providing very large energy storage capacity (above 100 MW in single units). Currently, CAES accounts for approximately 20% of global energy storage installations, except for pumped hydro. CAES is based on conventional gas turbine technology, with the compression and expansion processes decoupled. In a traditional gas turbine power station, around two-thirds of the turbine output is needed for compressing the combustion air (Ibrahim et al., 2008). In a CAES power station, however, no compression is needed during turbine operation because the required enthalpy is already included in the compressed air. As a result, the net power output for a given sized turbine expander can be 2.5 times that for the compressor load (Septimus, 2006).

Large CAES systems are designed as central storage facilities to cycle on a daily basis and to operate efficiently during partial-load conditions. This design approach allows CAES units to swing quickly from generation to compression modes, which can efficiently absorb or compensate solar power. CAES plants can also respond to load changes to provide load following, because they are designed to sustain frequent start-up/shut-down cycles.

The technical feasibility of CAES has been proven by two large-scale installations. These installations have been successfully running for decades, and as a result, CAES can be regarded as a mature technology. However, the fact that there have been no installations in recent years raises some concerns about the technology. (In fact, some planned large-scale CAES projects have been either terminated/canceled or postponed.) Technically, this may be associated with low round-trip efficiency and economics. The low round-trip efficiency is due to the following aspects:

• The air compression process is highly irreversible, particularly for high-pressure ratios.

• Storage of compressed air is commonly under constant volume conditions. This leads to a great loss of energy due to the throttling process of higher pressure charging and lower pressure discharging (Jovan et al., 2011).

• The expansion process is irreversible, though it is less serious compared to the compression process.

The combination of the previous points gives relatively low round-trip efficiency of around 40%. Even for the new planned facilities in the United States, the round-trip efficiency levels are expected to be between 42% and 54%, depending on the way the waste heat is used (Safaei et al., 2013).

To resolve the issues, recent developments of CAES technology have been in the storage vessel and energy conversion processes, which could ultimately contribute to round-trip efficiency improvement and cost reduction. These include constant-pressure storage vessel (e.g., water-compensated container (Kim et al., 2011), underwater container (Pimm and Garvey, 2009), adsorption-enhanced CAES) and novel energy conversion processes (e.g., near-isothermal CAES (Li, 2011; Li et al., 2011), advanced adiabatic CAES (Bullough et al., 2004)). However, most of these are still in their early research stage, and no reports have been seen on demonstrative scale systems.

2.4.3 Thermal Energy Storage

Thermal energy storage (TES) refers to technologies that store a thermal form of energy, which can be heat or cold. TES technologies can be broadly divided into two categories of thermophysical and thermochemical, with the principle of the former based on the temperature difference and/or latent heat due to phase change, and that of the latter on reaction or sorption enthalpy. Thermophysical-based TES is much more technically developed than thermochemical-based TES. Globally, TES technologies account for approximately 54% of global energy storage installations, except for pumped hydro. TES technologies can be used in both supply side and demand side management. In the demand side, TES is largely distributed and mostly at an individual building scale for domestic heating or cooling. For example, almost 14 million households in the United Kingdom have a hot water cylinder, giving a maximum combined storage capacity of around 80 GWh (Taylor et al., 2013). However, only recently has it been recognized that TES can also be an effective means of electricity management at the supply side.

Potential applications of TES technologies are ample. Examples include solar electrical energy management, particularly solar thermal power generation such as concentrating solar power (CSP), cryogenic energy storage (CES), Pumped Thermal Electricity Storage (PTES), and advanced adiabatic CAES (Bullough et al., 2004).

CSP is unique compared with solar photovoltaics and other renewable energy generation methods because it can be effectively integrated with TES. In addition, TES can decouple the thermal energy production process from the thermal energy utilization (power generation) process in CSP plants. This makes TES highly attractive due to fewer energy conversion steps and hence lower energy losses. At the heart of TES technologies is the storage medium (material). Currently, solar thermal power generation uses synthetic oil (e.g., VP1) and molten salt (e.g., solar salt) as the storage materials. Because these materials also act as heat transfer fluid, they are in the liquid form. As a result, they only store sensible heat with limited energy density. The use of phase change materials and the thermochemical method (with high energy density) is still in the development stage (Gil et al., 2010).

CES is a newly developed electrical energy storage technology that uses cryogen (e.g., liquid air/nitrogen) as the storage medium (Li, 2011; Li et al., 2011). The charging process is through air liquefaction, which consumes excessive electricity, whereas the discharging process is liquid air/nitrogen fuelled power generation. CES is potentially attractive for utility-scale electrical energy storage, as cryogen can be stored at low pressures (close to ambient) and has a much higher volumetric storage density (> 10 times that of CAES and > 100 times that of PHS) (Li et al., 2010). Additionally, cryogen is not only the energy carrier, but also the working fluid in the energy release process. Because cryogen has a relatively low critical point compared to steam, waste heat (e.g., from the exhaust of a new or existing simple-cycle gas turbine) can be recovered efficiently in the CES system (around four times more efficiently than in the Organic Rankine Cycle system) to achieve high energy-storage efficiencies (Strahan, 2013). However, without an external heat source, an independent CES system has relatively low efficiency (below ~60%) due to the high energy consumption of the air liquefaction process (Chen et al., 2009). Another potential issue is that current liquid nitrogen/oxygen production facilities are designed for continuous operation. This is in contradiction to the main function of the storage system to absorb excessive electricity. CES is a pre-commercial technology, with a 350 kW/2.5 MWh scale demonstration in operation since 2011 and a 5 MW/15 MWh demonstration plant to be completed in early 2015 in the United Kingdom (HPS, 2014). Current research and development efforts are mainly on round-trip efficiency improvement through process integration, novel liquefaction process for better flexibility, and system scale-up.

PTES is a recently proposed concept based on reversible heat pump cycles (Howes, 2012). In the charging process, the system works as a heat pump converting electricity into thermal energy (both heat and cold); in the discharging process, the system works as a heat engine to convert cold and heat into electricity. The greatest challenge of the PTES technology is associated with the mechanical components (compressor-expander coupled machinery), which need to have high efficiencies in both forward (charging process) and reverse (discharging process) operations (Desrues et al., 2010). In particular, a PTES system has to use an inert gas such as helium or argon instead of commonly used refrigerants to produce high-grade thermal energy (below − 150 °C at the cold end and above 600 °C at the hot end). The high-grade heat and cold are stored as “sensible heat” in insulated cylindrical vessels containing an appropriate thermal storage medium (e.g., a packed bed of pebbles or gravel, or a matrix of ceramic material). Heat exchange is through direct contact between the working fluid and storage medium, enabling a quick switch between charging/discharging modes and efficient integration with the thermodynamic cycle, and avoiding the “pinch-point” difficulties associated with phase-change storage methods. However, because a complete charging and discharging cycle involves twice as many compression, expansion, and heat transfer processes, the round-trip efficiency of PTES is limited due to the irreversibility of these processes. PTES technology is still in the early stage of development, and according to a UK company's plan, a 1.5 MW/6 MWh demonstrative storage unit will be deployed on a UK grid-connected primary substation in the near future (ETI, 2013).

2.4.4 Flow Battery

Modern redox flow batteries (RFBs) were invented in 1976 by Lawrence Thaller at the National Aeronautics and Space Administration (NASA) (Thaller, 1976). RFBs convert electrical energy into chemical potential energy by means of a reversible electrochemical reaction between two liquid electrolyte solutions contained in external electrolyte tanks. The conversion between electrical energy and chemical energy occurs as the liquid electrolytes from storage tanks flow through electrodes in a cell stack. In contrast to conventional batteries, they store energy in electrolyte solutions and the energy is proportional to the amount of electrolytes in the tanks (Ferreira et al., 2013). The cell has two compartments (positive half-cell and negative half-cell) separated by a membrane to prevent mixing of the electrolytes (Wang et al., 2013; Tan et al., 2013). Compared with other decoupled electrical energy storage technologies, RFBs have relatively high round-trip efficiencies, short response times, a symmetrical charge and discharge process, and quick cycle inversion.

RFB energy storage systems are being developed for use in small-scale applications, such as stand-alone power systems with solar photovoltaic arrays or wind turbines, and distributed energy installations for electric utility services. The first true RFB used a ferric/ferrous (Fe^2 +/Fe^3 +) halide solution electrolyte in the positive half-cell and a chromic/chromos (Cr^2 +/Cr^3 +) halide solution electrolyte in the negative half-cell. The battery soon encountered severe cross-contamination that resulted in dramatic capacity decay. The effort made to mitigate the cross-contamination led to the invention of the all-vanadium redox flow battery (VRB), which concerns only one active element in both positive and negative electrolytes, utilizing four different oxidation states of vanadium ions to form the redox couples. VRBs could have a round-trip efficiency of > 70% (Skyllas-Kazacos and Robins, 1988). They have an expected lifespan of 15 years with a low environmental impact (Rydh, 1999), and as a result, a considerable amount of fundamental and applied research has been carried out (Zhao et al., 2006).

In addition to VRBs, there has been work on other flow batteries (Chakrabarti et al., 2011; Liu et al., 2013; Yan et al., 2013; Ponce et al., 2006). Some of them have been studied for solar energy storage (Chakrabarti et al., 2011; Liu et al., 2013; Yan et al., 2013). Examples include ruthenium-based RFBs (Chakrabarti et al., 2011) and a combination of RFBs with dye-sensitized solar cells (Liu et al., 2013; Yan et al., 2013).

2.4.5 Solar Fuels

Solar fuels store solar energy in a chemical form for use when sunlight is not available. The most widely investigated solar fuels are hydrogen and hydrocarbon (Licht et al., 2001; Yamada et al., 2003). There are a number of methods for solar fuel production. For example, hydrogen production can be done by water electrolysis, thermolysis, and photolysis. These production methods are either indirect or direct. Indirect hydrogen production occurs through converting solar radiation into another form of energy, such as heat and electricity, followed by fuel production. Direct methods harness solar energy to produce fuels without any intermediary conversion steps. In principle, the indirect methods have the disadvantage of low efficiency due to losses in the intermediary conversion steps, but practically, they are easier to implement. Today, many advanced large-scale water electrolysis units are in operation with an electricity-to-hydrogen efficiency of over 75%, while most other solar hydrogen production processes are still at the stage of laboratory research (Licht et al., 2001; Yamada et al., 2003). Production of solar fuels is not the only challenge, particularly for solar hydrogen; challenges are also found in hydrogen storage and electricity regeneration steps. Although hydrogen can be stored as compressed gas, liquefied gas, or in a solid hydrogen carrier, none of these are currently energy-efficient or cost-effective. Fuel cells are currently regarded as a promising energy extraction technology for hydrogen. They are about four times more expensive than combustion engines/turbines, and two to three times shorter in terms of lifespan (Li et al., 2010). Significant efforts are therefore needed to address these issues.

Instead of hydrogen production, combining the water-splitting process with CO₂ can generate liquid fuels (Rakowski and Dubois, 2009; Benson et al., 2009; Centi et al., 2007). These fuels have higher volumetric energy densities compared to hydrogen, and could potentially alleviate challenges associated with hydrogen storage. The liquid solar fuel routes have been shown to be achievable in research labs, but scaling up to commercial level faces a number of challenges, including system efficiency and lifetime and capital costs (Cook et al., 2010).

2.5.1 Lead-Acid Battery

The lead-acid battery, invented in 1859 by Gaston Plante, is the oldest and most widely used rechargeable electrochemical device in automobile, uninterrupted power supply (UPS), and backup systems for telecom and many other applications. Such a device operates through chemical reactions involving lead dioxide (cathode electrode), lead (anode electrode), and sulfuric acid (electrolyte) (Parker and Garche, 2004). Lead-acid batteries have a high round-trip efficiency, and are cheap and easy to install. It is the affordability and availability that make this type of battery dominant in the renewable energy sector. It is also well known that lead-acid batteries have low energy density and short cycle life, and are toxic due to the use of sulfuric acid and are potentially environmentally hazardous. These disadvantages imply some limitations to this type of battery. Indeed, a recent study on economic and environmental impact suggests that lead-acid batteries are unsuitable for domestic grid-connected photovoltaic systems (McKenna et al., 2013).

2.5.2 Lithium-Ion Battery (Li-Ion)

The first lithium battery was built in 1979 (Mizushima et al., 1980). Lithium batteries in the early days suffered from poor cycle life and safety problems. It was the development of lithium-cobalt batteries with carbon as negative electrodes that led to the successful commercialization of lithium-ion batteries by Sony in 1990. The term “lithium batteries” actually means a family of dozens of different battery technologies based on moving lithium ions between a positive electrode consisting of a lithium and transition metal compound and a negative electrode material. Due to the nature of the reactions involved and the structure of electrodes, lithium-ion batteries have a much longer cycle life than lead acid batteries do (McDowall, 2008). Lithium-based batteries have high round-trip efficiency, high energy and power density, and a low self-discharge rate, and have been widely used in portable electronics. High power density also makes lithium-ion batteries an option for powering electrical vehicles. Efforts have also been made to extend the application range to solar energy storage applications (Guo et al., 2012).

2.5.3 Nickel-Based Battery

Nickel-based batteries mainly refer to nickel-cadmium (Ni-Cd), nickel-metal hydride (Ni-MH), and nickel-zinc (Ni-Zn) batteries. Ni-Cd batteries consist of a positive electrode with nickel oxyhydroxide as active material, and a metallic cadmium-based negative electrode with aqueous potassium hydroxide as electrolyte (Shukla et al., 2001). The advantages of such batteries include robustness to deep discharges, long cycle life, temperature tolerance, and high energy density (compared with lead-acid batteries). However, the cadmium used in Ni-Cd batteries is highly toxic. This leads to the development of Ni-MH batteries, which are much more environmentally friendly, though they have a high self-discharge rate. Despite various developments, only Ni-Cd batteries have found commercial application in UPS systems. Recently, Ni-Cd batteries became one of the popular storage technologies for solar energy generation because they can withstand high temperatures, though the high initial investment of Ni-Cd battery systems may hinder widespread application in the sector (Nair and Garimella, 2010).

2.5.4 Sodium-Sulfur Battery

The development of sodium-sulfur (NaS) battery technology has been ongoing for more than 50 years. NGK of Japan was the first to commercialize the technology successfully. NaS batteries consist of liquid (molten) sulfur at the positive electrode and liquid (molten) sodium at the negative electrode as active material, separated by a solid beta alumina ceramic electrolyte (Hadjipaschalis et al., 2009). They have high energy density, long cycle life, and high round-trip energy efficiency. These features make the technology suitable for load leveling, power quality, and peak shaving in distributed energy systems, including solar power (Baker, 2008). In fact, current global installations of NaS batteries exceed 300 MW, despite recent concerns about the safety aspects.

2.5.5 Other Battery Technologies

Intensive research efforts have been made in recent years, seeking new storage technologies that can economically provide the power, cycle life, and energy efficiency needed to respond to the short-term transients arising from renewables and other aspects of grid operation. The following are some examples of recent developments.

A new type of aqueous electrolyte battery has been proposed and demonstrated (Pasta et al., 2012). Such a technology relies on the insertion of potassium ions into a copper hexacyanoferrate cathode and a novel activated carbon/polypyrrole hybrid anode. The cathode reacts rapidly, with very little hysteresis. The hybrid anode uses an electrochemically active additive to tune its potential. It has been demonstrated that a 95% round-trip energy efficiency can be achieved when cycled at a 5C rate, and the efficiency reduces to 79% when cycled at 50C. The results also show a zero-capacity loss after 1000 deep-discharge cycles.

An aqueous rechargeable sodium-ion battery has recently been proposed. This technology uses NaCuHCF as cathode, NaTi₂(PO₄)₃ as anode, and aqueous Na₂SO₄ solution as electrolyte. Lab experiments have shown a voltage of 1.4 V and a specific energy of 48 Wh/kg based on the total weight of the active electrode materials (Wu et al., 2014). In addition, the aqueous Na-ion battery has also shown an excellent high-rate discharge capability and cycle capability, with 88% capacity retention over 1000 cycles at a 10C rate.

A novel symmetric open-framework electrode battery is proposed and demonstrated experimentally (Pasta et al., 2014). The battery is shown to provide a maximum specific energy of 27 Wh/kg at a 1C rate based on the mass of the active material (Pasta et al., 2014). At a 50C rate, the battery has a specific energy of 15 Wh/kg, a specific power of 693 W/kg, and an energy efficiency of 84.2%. These performance data suggest that this type of battery is suitable for grid-related applications, including smoothing of intermittent variations in power production associated with the integration of renewable energy.