Last week I introduced a new study titled “Energy Storage for the Electricity Grid: Benefits and Potential Market Assessment” that was commissioned by the DOE’s Energy Storage Systems Program, identified seventeen discrete storage applications for the electricity grid, discussed the technical requirements of each application and summarized the potential economic benefits.
If the Yahoo! message boards are any indication, investors are already jumping to inaccurate and wildly optimistic conclusions because they don’t understand that many storage applications are synergistic and every storage system purchaser will try to maximize the value of its investment by capturing as many value streams as possible. The process is called “aggregation” and while it will speed the implementation of storage on the smart-grid, it will ultimately destroy the high value niche markets for frequency regulation, short-duration wind integration, electric service reliability and similar ancillary services.
To truly understand the issues, investors need to stop looking at individual trees and focus instead on the forest.
One of the biggest challenges facing developers of grid-based energy storage systems is that electricity is cheap and abundant, and storage can be incredibly expensive. As a result, most of the grid-based applications identified in the new DOE study are not attractive as stand-alone value propositions. In the following table, the applications highlighted in blue make economic sense today as stand-alone value propositions. Conversely, the applications highlighted in yellow won’t generally work unless a particular installation can capture and monetize several value streams. As utilities and other users begin installing significant storage capacity and aggregating value streams to maximize their returns, total system capacity will rapidly outrun demand for niche services, thereby eliminating the value premium. Over the long-term, the economics of grid-based storage will obey the laws of economic gravity. The only companies that will survive, much less thrive, are manufacturers of cheap, durable and dependable energy storage systems that can do the required work at the lowest cost.
A prime example of the prevailing “can’t see the forest for the trees syndrome” is the wildly over-hyped idea that we can use plug-in vehicles to provide ancillary services while they’re connected to a charging station. The silly values floating around for vehicle to grid, or V2G, services are all based on the theory that EV batteries can be used for frequency regulation and other high value ancillary services. While the theory sounds wonderful in the telling, the fundamental premise is fatally flawed and the promised benefits to plug-in vehicle owners will never be realized because they violate the law of supply and demand. The easiest way to demonstrate the point is with an example.
At last year’s EESAT conference in Seattle, a representative of the PJM Interconnect estimated that total national demand for frequency regulation was on the order of 6,000 MW. Storage companies that are actively pursuing opportunities in frequency regulation include Beacon Power (BCON), Altair Nanotechnologies (ALTI) and A123 Systems (AONE). In general the battery companies that are working on fast response products claim their systems can provide two to four MW of frequency regulation service for each MWh of battery capacity. Beacon is claiming a 20-year life for its flywheel systems. Demonstration projects are currently under way to determine whether these performance claims will withstand the tests of time and intensive use. For purposes of this example I will assume that all systems perform up to expectations.
President Obama has established a policy goal of one million plug-in vehicles on the road by 2015. If that goal is reached and the average plug-in vehicle is equipped with 20 kWh of batteries, a figure that’s mid-way between the GM Volt and the Nissan Leaf, then the total battery power available for V2G services will be roughly 20,000 MWh and the aggregate amount of frequency regulation those batteries could theoretically provide would be somewhere between 40,000 MW and 80,000 MW.
It doesn’t take a PhD economist to know that if sellers try to force 40,000 to 80,000 MW of supply into a 6,000 MW national frequency regulation market, prices will collapse. Similar issues exist across the entire spectrum of grid storage applications.
In a 2007 “Guide to Estimating Benefits and Market Potential for Electricity Storage in New York” that was commissioned by the New York State Energy Research and Development Authority, Mr. Eyer and his colleagues identified and evaluated a number of potential synergies between different grid-based storage applications and concluded that users would need to carefully consider the potential value of the following complimentary uses when planning a new grid-based storage installation.
|Electric energy time shift||Transmission and distribution (T&D) upgrade deferral; Transmission congestion relief; Electric service reliability; Electric service power quality; and Ancillary services.|
|Electric supply capacity||T&D upgrade deferral; Transmission support; Electric service reliability; Electric service power quality; and Electric supply reserve capacity.|
|Reduce transmission capacity requirements||Electric energy time shift; T&D upgrade deferral; Electric service reliability; Electric service power quality; Transmission support; and Ancillary services.|
|Transmission congestion relief||Electric energy time shift; T&D upgrade deferral; Electric service reliability; Electric service power quality; Transmission support; and Ancillary services.|
|T&D upgrade deferral||Electric energy time shift; Transmission congestion relief; Electric service reliability; Electric service power quality; and Ancillary services.|
|Operating reserves||Voltage support; Electric service reliability and Electric service power quality.|
|Regulation and frequency response||Limited.|
|Electric service reliability||Electric service power quality and Demand charge management.|
|Electric service power
|Electric service reliability and Demand charge management.|
|Demand charge management||Electric service reliability and Electric service power quality.|
|Time-of-use energy cost management||Limited.|
|Renewables energy time shift||Generation capacity deferral; T&D upgrade deferral; Transmission congestion relief; Electric service reliability; Electric service power quality; and Ancillary services.|
|Renewables capacity firming||Electric service power quality; Electric energy time shift; T&D upgrade deferral; and Transmission congestion relief.|
The point of the foregoing is not to pick winners and losers in the emerging market for grid-based storage solutions. Rather my goal is to highlight the immense differences between demonstration projects that establish whether a particular storage device can meet the technical requirements of a specific application and a detailed cost-benefit analysis that establishes whether a particular storage system will be cost effective for a particular user. As the market unfolds, I expect many demonstration projects to be impressive technical successes. Most of those technical successes, however, will be dismal economic failures because the cost of the storage system will be far too high for widespread implementation by potential users. The utilities all understand they can’t buy a service for dime, sell it for a nickel and make it up on volume.
In a July 2008 report on its Solar Energy Grid Integration Systems–Energy Storage (SEGIS-ES) program, Sandia National Laboratories provided a summary table of current and projected capital costs for grid-quality manufactured energy storage systems. While commenters often criticize this table for conflicting with more the optimistic numbers that appear in corporate presentations and the mainstream media, I tend to believe Government studies are more reliable than public relations.
When I compare the capital cost figures in the SEGIS-ES table with the economic benefit per kWh values that I derived from the new DOE report on grid-based storage applications, the only companies I see that are within reasonable striking distance of a 10-year product life and a capital cost that compares favorably with the economic values are:
- Enersys (ENS), a leading manufacturer of lead-acid batteries for commercial and industrial applications;
- C&D Technologies (CHP), a leading manufacturer of lead-acid batteries for uninterruptible power systems;
- Active Power (ACPW), an established manufacturer of flywheel-based uninterruptible power systems;
- ZBB Energy (ZBB), which is scaling up manufacturing of a zinc-bromine flow battery system; and
- Axion Power International (AXPW.OB), which is preparing to begin commercial production of its PbC line of asymmetric lead-carbon supercapacitors in cooperation with Exide Technologies (XIDE).
All of the other systems that I’m aware of suffer from crushing raw materials or capital cost constraints. I understand that every storage system developer is actively pursuing research and development programs that may significantly reduce costs at some future date. Unfortunately, experience has taught me that it’s unwise to count chickens before they hatch and hope is not an investment strategy.
The grid-based energy storage sector is in its infancy and there is no reasonable way for an average investor to learn enough to pick individual stocks with any level of confidence. While I’m a stock picker when it comes to my personal holdings, I believe that a balanced portfolio of established and emerging energy storage companies is the only rational way for non-professionals to invest in the sector. Disproportionate investments in individual companies should be avoided unless you’re prepared to do a whole lot of investigation and analysis.
Disclosure: Author is a former director of Axion Power International (AXPW.OB) and holds a substantial long position in its stock. He also has small long positions in Enersys (ENS), Exide Technologies (XIDE), C&D Technologies (CHP), ZBB Energy (ZBB) and Active Power (ACPW).
I agree that the niche markets are over-hyped, but I think your assumptions go too far in the other direction.
First of all, the limiting factor for the MW of frequency regulation available for a PHEV will probably be the outlet its plugged into. A 220V 40A outlet is probably only going to be useful for about 10kW of frequency regulation, no matter how big the battery pack plugged in to it, so your numbers for frequency regulation from GEVs are probably a factor of 4 or more too high.
Second, you implicity assume that demand for grid services is constant, as supply increases. That, as you say, not only violates fundamental economics (as price falls, demand will increase), but demand is also increasing with increasing grid penetration by wind and solar.
What a wonderful and insightful comment Tom. While I hadn’t thought it through before, the maximum power flow through a 220 V- 40A circuit is 8.8 kW and the maximum flow through a 110 V – 40A circuit is half that. So while a 20 kWh battery might be able to handle 80 kW of V2G services, a typical connection would be 5% to 10% of that total. The limitation is something that I’m sure V2G advocates have not even thought of, but it pretty much destroys the illusion.
Your point about demand growing is also well taken. The only counterpoint is that every battery installed on the grid has some FR capacity and the owners of those batteries are certain to want to capture every high-value revenue stream they can lay their hands on, which gets us back to supply and demand issues if big chunks of storage are installed.
Aren’t we putting the cart before the horse?
The Nissan Leaf has a home charging station arrangement with Aerovironment (symbol AVAV). This means if you buy a Nissan Leaf, you are installing Aerovironment’s level II charger at your house (240V at 15A or 30A).
AeroVironment has not made the claim that they will enable these Nissan Leaf’s to sell energy back to the grid.
So why even worry about this if it isn’t going to be a factor for some time to come?
So that’s why I’m invested in Aerovironment. They have the Nissan Leaf contract. They’ve been selling/installing home charging stations since the GM EV-1. They’re the leader. They also will produce the 440V fast charging stations which will accelerate charging to 80% of battery capacity in 15-25 minutes.
Don’t sweat the V2G scenario. Figure out who is going to be the leader in charge of charging all these vehicles. 1 million cars X $1500 per home station? = big margins and big $$$$.
I would rather though we change the terms and instead of “cheap” we use “old tech” and instead of “cool” we use “new tech”.
And again I think that we may both agree that there will be a hybrid of technologies that serve the future grid, however to make arguments that all innovations will be snuffed out by cheap technology today, shows no respect for history.
Are you not pointing to individual “trees” instead of noticing that there is a forest with “individual trees”? As I had also pointed out the other day, I am sure the makers of beta tapes (“cheap”) were not looking at the then “DVD cool tree” in that forest?
Trees (beta tapes) in that forest died, others it seems now have flourished
You in that previous post said you would not debate until you wrote more articles on such. So I will keep my comments very brief (this time) on a complicated subject that requires complexity.
If one reads this Sandia paper intensely and taken with many more from the industry, I would argue that your shadings of blue, orange, and yellow to denote “cheap vs “cool” in your previous article do not do your readers justice.
Again, as I had pointed out the other day in a response to your first article in this area. Most researchers and papers on these different subjects refer to a much higher “frequency of events” in the under 15 minute space. Of course every event is different, but when one looks at the “frequency”, one can easily deduce that the “high cycling” rapid “charge/discharge” without degrading will also fit many these “synergies” you refer to.
For lack of space I will point out several “Benefits” that in my opinion Beacon Power Flywheels have the ability to “cycle” and have “deep charge/discharge” without loss of “round trip efficiency” that are more valued in the industry and much more “synergistic” than “old tech” in the design of the new grid.
Thus in my opinion the following Benefits are ignored in your analysis of the flexibility of certain “new tech”.
App#3. Load Following.
220.127.116.11 Load following is characterized by power output that changes as frequently as every several minutes.
18.104.22.168 Storage used for load following will probably need access to automated generation control (AGC). Typically, an ISO requires output from an AGC resource to change very minute.
App#8. Transmission Congestion Relief
22.214.171.124 The discharge duration needed for transmission cannot be generalized easily given all the possible manifestations.
App#9. Transmission and Distribution Upgrade Deferral
126.96.36.199 The key theme is that a small amount of storage can be used to provide enough incremental capacity to defer the need for large “lump’ investment in T&D equipment.
Later on…”a similar rationale applies to equipment life extension.
188.8.131.52 Discharge duration is a critical design criterion that cannot be generalized well.
App#10 Substation On-site Power
A very small market however with small storage.
184.108.40.206 “Competitive options should have a straightforward way to determine the storage system’s remaining useful life and “system health”
Also “Especially important are the ability to serve momentary loads including switchgear operation, motor driven valves, isolating switches, and the field flashing generators.
App#12. Demand Charge Management.
220.127.116.11 In many cases, the demand charge is assessed if load is present during just one 15 minute period.
18.104.22.168 For maximum load that occurs for even a few minutes, storage must be reliable. It must have acceptable or better quality power quality for loads served.
App#16 Renewables Capacity Firming.
22.214.171.124 Output can change rapidly over short periods of time. “ramping”.
Plenty of research in this area.
I would agree that in all of the above “Benefits”, Beacons flywheel technology may not have the duration to handle the maximum events, but that is why there are duration technologies that obviously would sit behind the quick rapid technologies. That is also why in the Sandia paper many times it is said “minutes up to 2 hours”. We would prefer however to focus on the “frequency” of these durations and eat that “huge” piece of the pie rapidly, over and over again with quick charge/discharge cycles.
Beacon Power estimates 120,000 plus cycles over a 20 year life. Can fully deep charge within 15 minutes and discharge within the same over and over. That is a minimum of 6,000 of these cycles per year.
“Cheap” batteries life can range in the 5-7 year frame and only a maximum of 4,000 -7,000 cycles over this period, thus only about 1,000 cycles per year, of course with a longer duration, but also they cannot do deep charge/discharge without efficiency degrading, so it takes them longer to recharge at a slower rate.
Which system would it seem to provide more “synergistic” flexible to the complex grid?
I will take the reps and the DOD.
You have also ignored “Incidental Benefits” #’s 18-26.
There is certainly more to add….Section E.
An avid Yahoo Message Boarder
I appreciate the suggestion, but think I’ll stay with the cheap vs cool distinction. My rationale is simple. Every industrial revolution has come on the heels of an innovation that has proven its ability to do more valuable work with fewer economic inputs. The operational specifics of a technology make for fun discussion, but at the end of the day the only things that matter to owners, and by transitivity to stockholders of owners, are reducing operating costs or increased revenues. It all boils down to money.
I truly like Beacon’s technology and think it fills a valuable niche. But I don’t believe for a minute that other alternatives are incapable of eating into it’s potential markets.
Flywheels are great in spiky markets where demand tends to yo-yo up and down through supply on a minute-to-minute basis. It has very little value where demand tends to exceed supply for an extended period of time because there are no recharge opportunities.
Many storage devices have some ability to provide the equivalent of FR services with rapid shallow cycling. When there are enough storage resources online that are able to satisfy modest FR demand, I expect the demand for special purpose installations to fall off.