Electricity is the highest quality energy carrier, increasingly dominant throughout the world's energy infrastructure. Ultimately electricity use can expand to efficiently meet virtually all stationary energy applications, eliminating stationary end-use carbon emissions. This approach is unlikely to work in transportation, however, due to the high cost and low energy density of electricity storage. Chemical energy carriers, such as hydrogen, can more effectively serve transportation fuel and energy storage applications, offering much higher energy density at lower cost. Electrolytic hydrogen, extracted from steam with renewable energy, stored as a high pressure gas or cryogenic liquid, and reconverted to electricity in fuel cells and or used to power hydrogen vehicles, will reduce emissions from both transportation and electric generation. Renewable resources and modular electrolytic technology also permit decentralized hydrogen production, circumventing distribution issues and barriers to market entry. In contrast, sequestration-based fossil-fueled systems must achieve economies of scale by relying on centralized production and hierarchical transmission and distribution of electricity, hydrogen fuel, and carbon (dioxide).
Renewable and fossil approaches may ultimately turn out to be complementary. Use of renewable sources would limit the sequestration burden to modest quantities using the most cost effective methods and reliable disposal sites. Previous analyses have concluded renewable electricity will be cost effective in combination with dispatchable carbonless energy sources (i.e. hydroelectric, fission, and biomass), to minimize energy storage (Union of Concerned Scientists, 1992; Kelly and Weinberg, 1993). Dispatchable carbonless sources only generate a fraction of current electricity, however, and are likely to be limited on the scale of burgeoning demand (Schipper and Meyers, 1992; Fetter, 1999). A future role for biomass, in particular, may be restricted due to competing uses for land, water, and perhaps other agricultural inputs (Smil, 1998). An alternative to expanding fission, hydropower, and biomass is to use modest fossil electric generation and carbon sequestration as a complement to wind and solar energy. An integrated hydrogen transportation sector complements renewable systems both by providing a large, but flexible, use for excess renewable electricity and by enabling dual-use of hydrogen fuel as utility energy storage and transportation.
This chapter surveys energy storage and hydrogen vehicle technologies, analyzing the integration of these technologies into increasingly renewable electricity and transportation sectors. The implications for greenhouse gas reduction strategies are examined using an aggressive efficiency scenario for the United States in 2020, the latest time horizon for which detailed sectoral projections have been made.
Low capital cost, but inefficient, gas-fired peaking plants are used to meet demand fluctuations in present utility systems. Demand fluctuations can also be met using energy storage to shift electric generation to more cost-effective times of day. Utility storage is employed today, in small amounts, using established principles of mechanical energy storage: elevated water or compressed air.
Hydroelectric pumped storage is the most widespread and mature technol ogy, however the theoretical energy density of pumped hydro is quite low, requiring 3.7 tonnes (about 1000 gallons) of water traversing 100m of elevation to deliver 1 kWh. Pumped hydroelectric plants are consequently most viable on a large scale. The largest pumped hydro facility in the world today uses Lake Michigan and an artificial lake averaging 85m of elevation. It has a peak generating capacity of 2000 MW delivering up to 15000000 kWh over a period of about 12 hours, supplying the equivalent electric demand of about one to two million people. Roundtrip efficiencies approach 70%. At present 2% of electric demand is met by pumped hydro systems (Dowling, 1991).
Two disadvantages of hydroelectric energy storage appear in the context of future energy systems. The large scale nature of hydroelectric storage indicates that little if any cost saving will exist for electric distribution systems connected to a pumped hydroelectric facility. Finally, in the context of solar or wind intensive energy systems it seems unlikely that sufficient sites could be found in convenient locations (i.e. where natural formations provide low per kWh storage costs) to contribute more than a minor role in overall energy storage. River-fed reservoir hydroelectric capacity is probably best used to offset seasonal variations in solar or wind electric generation.
Compressed air energy storage (CAES) is also a reasonably mature approach, though only employed in a few sites worldwide. The energy density of CAES is about 50 times greater than hydroelectric storage. Air compressed to 100 atmospheres of pressure in a 20 gallon volume contains 1 kWh of energy. This energy density is still quite low, however: the same volume of compressed natural gas (CNG) contains nearly 100 times more energy. CAES is economic at larger scales (100-200 MW), relying on natural formations for low cost storage capacity, limiting widespread implementation. The compressed air can be run through turbines to generate peak electricity, although the heat of expansion must be supplied by thermal storage or fuel. For economic reasons, interest is greatest in applying CAES if additional fuel is burned with the pre-compressed air, enabling smaller turbines to match peaks in electric demand. In this application, however, the majority of the energy from a CAES system actually comes from fuel rather than compressed air (Gordon and Falcone, 1995). CAES systems are also not the most efficient method of gas-fired electric generation, and their greatest benefit is reducing the cost of generation capacity, rather than energy storage per se. CAES would have little to offer carbonless energy systems which do not rely much on fossil fuels, but do require renewable energy storage. Widespread application of CAES in a greenhouse gas context would require either carbon sequestration, or the use of carbonless fuel (i.e. hydrogen) produced elsewhere. The capacity for compressed air storage in CAES systems would be probably be more valuable as compressed hydrogen storage in the context of carbonless energy systems. Leakage has not been a difficulty when storing town gas (a mixture containing hydrogen) in underground caverns near Paris, France (Ingersoll, 1991).
Advanced energy storage technologies, in contrast to conventional energy storage, are characteristically modular, highly engineered systems without the scale and location constraints of pumped hydroelectric or compressed air storage. Approaches to energy storage include thermochemical (chemical couples), thermal (phase change materials), mechanical (flywheels), and electrochemical (batteries and electrolytic fuel production).
Thermochemical energy storage approaches capitalize on the high energy density of chemical energy storage and the use of low cost and abundant materials. Thermal energy can be stored in reversible chemical reactions (e.g. 2SO3 <—> 2SO2 + O2 or CH4 + H2O <—> CO + 3H2) in which the reactants are transmitted though a "heat pipe" loop between thermal source and end-use over distances up to 100 miles (Vakil and Flock, 1978). For stationary applications, heat can be stored cheaply in the enthalpy of common materials (water, oil, or molten salts).
While thermochemical and thermal storage are expected to be low cost, the thermal energy stored is not as valuable as electric or fuel energy. The chief disadvantage of thermal energy storage per se is thermodynamic. Unlike electricity or fuels, thermal energy "leaks" continuously, and in proportion to the useful work which can be extracted (determined by the Carnot cycle). Today, large amounts of energy are used as heat for low temperature space and water heating, some of which could arguably be saved through judicious use of thermal energy storage. However, thermodynamics again present a disadvantage as future space and water heating needs could be supplied very efficiently using heat pumps. As a method of reducing carbon emissions, thermal energy storage is likely to be most useful at modulating solar power production to more effectively meet late afternoon peaks or nighttime electricity demands, using solar thermal electric plants (De Laquil et al., 1990).
Electricity can be stored reasonably compactly and very efficiently as kinetic energy in flywheels. Flywheel energy storage is in the early stages of commercialization, and is targeted at uninterruptible power supplies (UPS), where the value of energy reliability far exceeds the value of energy. Flywheels can spin at very high velocities (10000-100000 rpm) in vacuum using magnetic bearings. They offer high efficiency (90%+) charging and discharging, low power related costs ($100/kW) and the prospects of very long equipment lifetime
(Post and Post, 1973). A 1 kWh flywheel module may weigh 10 kg and occupy 20 liters. Flywheel feasibility has advanced substantially with the advent of very strong and light carbon fibers and other composite materials. On the other hand, all of this specialized technology and materials (e.g. magnetic bearings to eliminate friction and provide rotor stability) can lead to high costs per unit of energy stored. Cost estimates are currently $100/kW and $600/kWh of storage capacity, although costs may fall to below $200/kWh (Post et al., 1993) in mature mass production. Flywheels store relatively small amounts of electricity (1-300 kWh) and are probably best placed near end-users in the electricity system, easing the burden on distribution, providing peak power and reliability, and making future energy systems uninterruptible.
The chief alternative to flywheels is electrochemical energy storage. Batteries are heavier, and less efficient (70-80% turnaround efficiency), but more compact than flywheels. Batteries have lower capital costs ($100-$200/kWh), but also a much lower cycle life (1000s of cycles) placing in some doubt their role in bulk power storage. The availability of mineral resources for common battery materials (lead, nickel, cadmium etc.) is likely insufficient (Andersson and Rade, 1998) for globally significant amounts of energy storage (e.g. 24 hour storage, roughly 100 billion kWh would require 1-2 billion tonnes of battery materials) in future electricity systems. The most compelling energy application of batteries is efficient electrification of moderate range (100-200 miles) passenger vehicles, assuming battery mass and cycle life can be improved sufficiently.
Less well known than batteries is a closely related alternative: electrolytic fuel production. Electrolysis differs from battery storage in that the electrodes are not chemically changed during electrolysis and do not store energy as in batteries. Energy is instead stored in the chemical fuel produced. A number of electrolytic fuels have been proposed (e.g. lithium, aluminum, and zinc) whose technology is closely related to metal-air (oxygen) batteries. One advantage of electrolytic fuels is the decoupling of power (electrodes and electrolyte) and energy (fuel) functions which are combined in batteries. This reduces the capital cost of achieving high power or large storage capacity. Electrolytic fuels offer potentially rapid refueling and lower weight than conventional batteries, especially when using atmospheric oxygen as a reactant.
Hydrogen has been considered for decades as a universal electrolytic fuel and energy carrier (Cox and Williamson, 1977; Bockris, 1980; Winter and Nitsch, 1988; Ogden and Williams, 1989; Ogden and Nitsch, 1993). Historically, the feasibility of hydrogen has been limited by the fuel economy of passenger vehicles and the corresponding weight of onboard fuel storage systems to achieve good travel range. Recent advances in composite materials, as well as hybrid electric vehicles, have resolved these issues, enabling future hydrogen vehicles 2-3 times more fuel efficient than those envisioned 20 years ago. As a renewable energy carrier electrolytically produced from abundant water, hydrogen is capable of fueling all transportation sectors indefinitely. A spectrum of hydrogen storage methods allow hydrogen systems to be tailored to the economics of individual applications. Electrolytic hydrogen fuel is expected to have low capital costs of production (electrolysis), storage (compressed gas, cryogenic liquid, or chemical storage), and utilization (hybrid electric engines or fuel cells). Estimates in a utility context are $500-1000/kW and less than $5/kWh. The chief disadvantage is that the cumulative process efficiencies of each step in hydrogen systems lead to roundtrip efficiencies of 30-40%, roughly half that of more direct storage technologies. Decentralization may offset this to some extent, potentially making waste heat available for space and water heating.
It is clear from the above discussion that energy storage technologies are best suited to different roles. Flywheels can improve transmission and distribution reliability, storing and delivering electricity perhaps twice daily. At the opposite end of the temporal spectrum, buffered hydroelectric generation may be most useful in adjusting to seasonal electric supply and demand variations. Batteries, most useful contribution would be enabling high efficiency short range transportation. The most important role of thermal energy storage technology is probably allowing solar energy to contribute to nighttime electricity production. Electrolytic hydrogen can serve as a bulk energy storage and universal transportation fuel, even if somewhat energy intensively.
The usefulness of each technology will depend on how well these roles meet the needs of future electricity supply mixes which make increasing use of intermittent electric generation, as well as an evolving transportation sector potentially powered directly by electricity or indirectly through electrolytic fuels.
Transmission technology advances can also play a role in carbonless energy systems, potentially easing local renewable resource constraints by enabling solar and wind energy to be harnessed at greater distances from urban load centers. Transmission is expected to incrementally improve by going to higher voltages, with DC transmission replacing AC transmission lines for long distances. In the longer term high power and perhaps underground cryoresistive and/or superconducting transmission lines could ultimately allow for wholesale long distance electricity transmission, reducing energy storage needed for seasonal and day/night variations around the globe. Another futuristic option may be transmission of energy by relay satellite, similar to proposed satellite or lunar solar power (Hoffert, 1998). Capital costs of transmission are typically moderate relative to both electricity distribution and renewable electricity production. In future energy systems, however, full utilization of electric transmission capacity may become more difficult if large, but intermittent, solar or wind energy facilities are distant from population centers.
An alternative to electric transmission is hydrogen transmission by pipeline or cryogenic tanker, just as natural gas is transmitted today. Pipeline systems offer some buffer capability reducing their sensitivity to short-lived fluctuations in supply or demand. Hydrogen pipelines have been in operation for decades and it is possible for today's natural gas pipelines to transport hydrogen, albeit at reduced pressure and higher cost than today's natural gas (CRC Press, 1977; Bockris, 1980; Winter and Nitsch, 1988; Ogden and Williams, 1989; Ogden and Nitsch, 1993). For equal investment, new hydrogen pipelines also deliver substantially more energy than electric transmission lines, although the conversion losses (electric energy to hydrogen energy and back to electricity) may counter this advantage. The chief factors determining the efficacy of transmitting energy as hydrogen are the scale of energy demand necessary to justify pipelines, and the fraction of demand for transportation fuel vs electricity.
In the near future, energy storage and perhaps transmission improvements can improve electricity distribution and reliability as electricity markets become deregulated, but this will not impact overall energy use or emissions substantially. In the intermediate term, an increasing reliance upon intermittent (solar, wind), and/or less flexible (nuclear) electricity sources will at some point require significant energy storage, as distinct from power storage. This storage will be needed to match non-dispatchable electricity sources with varying electric demands (Iannucci et al., 1998). Finally, energy storage as electrolytic fuels can extend the reach of carbonless energy sources to the transportation sector. This is especially important since the transportation sector is the highest value use of fossil fuels (Berry, 1996), the largest source of carbon emissions, and the least amenable to sequestration approaches.
Was this article helpful?