STORAGE AND BETTER THAN STORAGE FOR NEW ENERGY

The Role of Energy Storage with Renewable Electricity Generation
Paul Denholm, Erik Ela, Brendan Kirby, and Michael Milligan, January 2010 (National Renewable Energy Laboratory)

THE POINT
It is not accurate to say that wind and solar energies are intermittent and the scientists and engineers who are studying, developing and managing them have been working hard to change the description to the more accurate term of “variable.”

Variable is not the same as intermittent. Many people find their loved ones to be variable but the lucky ones also find them reliably constant and not intermittent. The challenge is to handle and depend on the variability. Without their variability, they would not be the ones we love.

Fossil fuels are not variable. They can be reliably be counted on to constantly generate energy and pollution in large quantities. Because of that, the New Energies’ scientists, engineers and developers are aggressively pursuing affordable, utility-scale electricity storage, the solution that would most obviously eliminate the variability of wind and solar and do away with concerns about their dependability. It turns out, though, that it is also possible to accept variability as an essential part of the nature of wind and sun and develop ways to successfully live with it.

The Role of Energy Storage with Renewable Electricity Generation, from Paul Denholm, Erik Ela, Brendan Kirby, and Michael Milligan of the National Renewable Energy Laboratory, says the issue of electricity storage is not a need but a want, something that is a matter of cost. Variable New Energy generation (VG) has technical and economic impacts but managing it successfully can make it dependable. This can be done with enabling technologies and grid operation practices, including most prominently (1) demand response, (2) increased transmission capacity, (3) flexible generation, and (4) improved operations.

The first and most important thing to understand about variability is that grid operators already cope with it on a daily basis. Demand varies constantly, by the hour of the day, by the day of the week and by the season of the year. Grid operators regularly manage a set of generation sources and ramp supplies up and down to match demand. Integrating New Energy sources is, in some ways, more of the same.

click to enlarge

Though some grid operators may nostalgically, if unrealistically, long for a simple yesteryear when coal was burned and electricity sent, variable New Energy generation can be equally serviceable, if not 100% dispatchable, by integrating it into a transmission system that can smartly process and dispatch electricity from a wide range of sources.

The are 2 real questions about energy storage. First, how much will it cost and is it worth the cost in comparison to the value from (1) building a smarter grid with more demand response capacity, (2) building more high capacity transmission to expand the range of available sources and load centers serviced, or (3) developing operational practices that allow for more effective variable New Energy integration into the existing grid? And second, what is the full potential of VG and how much can be used by a high capacity smart “future grid” before it becomes necessary to invest in energy storage or limit New Energy generation?

It is also worth remembering that the Old Energies have reliability issues as well. Nuclear energy is considered by some to be a preferable emissions-free source of electricity generation because it does not vary as often as solar and wind. Coal is considered preferable because it is cheap and certain.

But the variations in sun and wind are entirely predictable and can be compensated for with a flexible grid and distributed sources. When a nuclear plant crashes there is no warning. Coal plants, dependent on an antiquated transmission system, can be completely out of business in an unpredictable instant from the loss of a single line. When such events have happened, cities, parts of states, even interstate grid systems have been left in the dark for hours or even days.

Predictable, variable New Energy generation can be managed by a flexible transmission system and grid operator and, perhaps eventually, energy storage. That has costs. The unpredictable variation of the Old Energies cannot be managed and must be endured. That has enormous costs.

click to enlarge

THE DETAILS
The present day grid already faces variable demand. Grid operators cope with predictable but variable daily, weekly and seasonal demands with ramping operating reserves. Grid operators have both regulation reserves and contingency reserves that serve as backup to base load sources.

The need for reserves increases the costs and decreases the efficiency of any electricity supply system: (1) Fast response reserves are uneconomic; (2) partial loading is inefficient; and (3) reserve generation, though not fully utilized, nevertheless demands the same O&M expenses as base load capacity.

For these reasons, the pursuit of economic, utility-scale electricity storage is not a new phenomenon.

The most advanced and successful type of electricity storage yet developed is pumped hydro storage (PHS). Off-peak electricity is used to pump water against gravity and hold it behind a dam, making it available at times of peak demand to be released to flow with gravity and generate hydroelectric power.

The U.S. presently has ~20 gigawatts of PHS capacity. There is expense associated with PHS and its most active development has been when the overall cost of natural gas-generated electricity has been highest, such as during the oil embargo periods in the 1970s.

click to enlarge

A single prototype 110-megawatt compressed air energy storage (CAES) facility was also built in that era but has so far proved uneconomic.

Five current factors have created new interest in energy storage: (1) Advances in the technologies of energy storage, (2) rising fossil fuel prices, (3) deregulated energy markets producing higher returns on ancillary services like storage, (4) obstacles to new transmission such as siting wires and substations, and (5) the potential returns for discovering an affordable, scalable form of storage for the variable New Energies.

This historically adds up to increased demand but inadequate data to determine how much wide-scale deployment of energy storage would cost and which forms are the best. The assumption is that it would at least require sustained high costs for fossil fuels to be economic.

Another way to look at energy storage in the context of the past and present grid system is that it would be an alternative to building more Old Energy infrastructure for generation and transmission. Seen this way, storage offers several advantages: (1) It only has to be big enough to meet peak demand needs while additional infrastructure would be indiscriminately system-sized, (2) storage theoretically avoids many siting and permitting issues and puts the extra supply as adjacent to the demand as present plants are, and (3) reduces line loss inefficiencies.

click to enlarge

Energy storage can also (1) be a way to provide electricity after a system crash, (2) buffer fluctuations in supply that improve electricity supply stability and quality, and (3) serve remote and off-grid needs.

The impact of variable generation (VG) is best understood by thinking of the New Energies (especially wind power and solar PV) as sources of demand reduction that have predictable supply fluctuations over time. In other words, they are a way to reduce the grid's load. Old Energies meet grid operators’ “residual load” of normal demand but that is reduced by what New Energy generation can provide.

Impact challenges: (1) Increased need for regulation reserves; (2) increased ramping rate; (3) more uncertainty; and (4) increased range between daily minimum generation and daily maximum generation. Taken together, this necessitates greater operating flexibility and creates integration costs.

Benefits: (1) Reduced fuel use; (2) reduced emissions; and (3) reduced need for other generating capacity. Taken together, this creates reductions in costs.

click to enlarge

Multiple studies show the integration costs for wind are likely to add less than $5 per megawatt-hour ($0.05 per kilowatt-hour) or less than 10% to the price of electricity. If the benefits offset this and/or technology improves, integration costs could drop.

Less work has been done on the integration costs of solar but an Xcel study showed them to be in the $3.51-to-$7.14 per megawatt-hour range for 800 megawatts of solar PV in a 6,922 megawatt-peak system.

The studies on wind integration have been done on a wide range of percentages of wind integrated. The larger the percentage of wind being fed into the system, the higher the costs. Only one good study has been done with integrating more than 20% wind but research suggests 30% or 40% wind keeps costs consistent. Relatively little is substantively known about integrating more than 40% of VG into a transmission system.

click to enlarge

Research has produced a “flexibility supply curve” principle that suggests the cost of integrating VG as the percent of New Energy increases. It includes 6 sources of flexibility:
(1) Supply and reserve sharing moves the available VG around to where it is needed and requires increased transmission and forecasting capabilities.
(2) Generation flexibility requires the deployment of greater VG capacity.
(3) Demand flexibility requires expanding market-driving pricing schedules and more smart grid tools to increase demand response (DR) awareness and capabilities.
(4) VG curtailment gives utilities more freedom to ramp down the percent of New Energy in their mix if it is necessary to cut costs.
(5) Counter-intuitively, adding new loads could give utilities the options to higher levels of efficiency from space and process heating and vehicle-to-grid (V2G) technologies.
(6) Electricity storage includes a range of potential technologies.

The flexible enabling technologies that would serve New Energy are already being used in the present grid to facilitate the function of the Old Energies. They are generally less expensive and more efficient when applied across a wider transmission system. Exceptions to this are localized energy storage systems like solar power plant CAES systems and remote wind project battery storage systems.

click to enlarge

Energy storage technologies are divided into 3 categories that are based on the total power that can be stored and supplied:
(1) Power quality lasts seconds to minutes: (a) flywheels, (b) capacitors, (c) superconducting magnetic energy storage;
(2) Bridging power lasts mintes to ~1 hour: (a) lead-acid batteries, (b) nickel-cadmium batteries, (c) nickel-metal hydride batteries and (d) lithium-ion batteries.
and
(3) Energy management lasts hours: (a) high temperature sodium-sulfur batteries in use and hypothesized sodium-nickel-chloride (ZEBRA) batteries, (b) vanadium redox and zinc-bromine flow batteries in limited use and hypothesized polysulfide-bromine batteries, (c) pumped hydro storage (PHS), (d) compressed air energy storage (CAES) and (e) thermal (solar power plant heat) storage.

Only 4 energy storage technologies have so far demonstrated 100-megawatt or greater capacity: Arrays of sodium-sulfur (Na-S) batteries, PHS, CAES and thermal storage. With such limited real-world data, limited conclusive judgments about the costs or relative eficiencies of the competing technologies are possible.

PHS
The U.S. has 20 gigawatts of PHS capacity across 39 sites ranging from 50 megawatts to 2,100 megawatts. Some sites store as much as 10 hours of capacity. Some achieve 75%+ efficiencies. Availability of environmentally acceptable sites may be limited but are being sought. Underground reservoir technologies, though not yet proven, are in development.

click to enlarge

CAES
Th idea is to compress air in airtight underground storage caverns, then draw it out, heat it and use it, first to drive a turbine and then, in combustion with natural gas, to drive a 2nd turbine. Because this requires fuel, it is a hybrid generation/storage technology and its efficiency cannot be precisely calculated. This and finding adequate secure, large scale storage caverns make the concept’s potential questionable.

Thermal energy storage
The most promising example of storing “temperature” is that in which solar power plants heat water or molten salts, hold them under pressure and then (when there is no sun) release the heat to boil water and create steam to drive a turbine and generate dispatchable electricity. Efficiency may be quite high. Limitations include finding adequate storage capacity and the fact that it is a technology only available at solar power plants.

A second form of storing “temperature” is using off-peak electricity to cool water and release it to reduce the cooling load during peak power demand periods. This is a highly efficient process. A similar process can be used to capture the heat of the day for hot water or the heat thrown off during the electric power generation process for space heating. These processes are efficient but limited by the time of day and the output of the systems on which they are built.

click to enlarge

V2G
Charging battery electric vehicles (BEVs), both all electric EVs and plug-in hybrid electric (PHEV) vehicles, can be controlled and timed to periods of high VG output and done at the rate of generation. This makes the charged vehicle batteries a form of distributed storage. BEVs can then partially discharge stored energy to the grid during periods of peak demand, serving as distributed generation.

Use of BEVs is entirely speculative because there is almost no penetration in any vehicle fleet and the technology needed to use V2G storage and generation is not in use. Analysis has demonstrated high potential benefits and clear capability of the technology.

Bottom line on energy storage: There are substantial enabling technologies that are right now more economic choices. As New Energy comes to make up more than 20-to-40% of the power supply, storage may be necessary.

How much New Energy can be used before storage is the economic choice is simply not yet clear. Probably, there is no simple answer because the availability and cost of grid flexibility vary in many ways.

At current and predicted levels of New Energy capacity for the coming 2-to-3 decades, all grid flexibility options, including storage where it is economic, are needed.

click to enlarge

QUOTES
- From the NREL report on energy storage: “Because the wind doesn’t always blow and the sun doesn’t always shine at any given location, there has been an increased call for the deployment of energy storage as an essential component of future energy systems that use large amounts of variable renewable resources. However, this often-characterized “need” for energy storage to enable renewable integration is actually an economic question. The answer requires comparing the options to maintain the required system reliability, which include a number of technologies and changes in operational practices. The amount of storage or any other “enabling” technology used will depend on the costs and benefits of each technology relative to the other available options.”

click to enlarge

- From the NREL report: "The introduction of variable renewables is now one of the primary drivers behind renewed interest in energy storage. A common claim is that renewables such as wind and solar are intermittent and unreliable, and require backup and firming to be useful in a utility system – energy produced by wind and solar should be “smoothed” or shifted to times when the wind is not blowing or the sun is not shining using energy storage. These statements are generally qualitative in nature and provide little insight into the actual role of renewables in the grid, (including their costs and benefits) or the potential use of energy storage or other enabling technologies...It is easiest to understand the impact of VG technologies on the grid by considering them as a source of demand reduction with unique temporal characteristics. Instead of considering wind or PV as a source of generation, they can be considered a reduction in load with conventional generators meeting the “residual load” of normal demand minus the electricity produced by renewable generators..."

click to enlarge

- From the NREL report: "There are technical and economic limits to how much of a system’s energy can be provided by VG without enabling technologies based on at least two factors: coincidence of VG supply and demand and the ability to reduce output from conventional generators. At extremely high penetration of VG, these factors may cause excessive (and costly) curtailment, which will require methods to increase the useful contribution of VG However, the concern regarding how much VG can be used before storage is the most economic option for further integration currently has no simple answer, primarily because the availability and cost of grid flexibility options are not well understood and vary by region."