Alternative Energy Storage: Why Frequency Regulation Is Important

In August I wrote an introductory article entitled Grid-based Energy Storage: Birth of a Giant. Over the last few months the utility sub-sector been very active and now might be a good time to drill a little deeper into the kinds of utility applications that can benefit from cost-effective energy storage. Once again, I want to caution readers that I’m a little out of my depth when it comes to providing simple explanations for complex technical issues. But I believe an understandable overview can far more valuable to investors than too much detail. Once again, I'm not seeking perfection and will be more than happy with B- grades from power professionals and engineers.

To maintain a level of continuity, I’ll begin with a graph from the Lawrence Berkley National Laboratory that shows statewide electricity use in California on July 31, 2007.

While the 24-hour curve looks pretty smooth, it gets more complex when you start looking at shorter time intervals. I found the following example of what a three-hour morning segment might look like in a paper from the Oak Ridge National Laboratory titled “Frequency Regulation Basics and Trends” (December 2004).

While the ORNL graph may seem complex, it’s really fairly straightforward. The smooth blue line shows how on-line generating resources ramp up from 3,600 MW to 4,000 MW over a three-hour period from 7 to 10 a.m. The jagged green line overlying the blue shows actual demand during the same time period. The very jagged red line at the bottom is simply a scaled up representation of the minute-to-minute differences between the blue supply line and the green demand line. That 60 MW slice of highly variable demand is the domain of frequency regulation.

Historically, utilities were able to avoid frequency regulation issues by maintaining on-line generating capacity at a level that was always higher than expected peak demand. In ORNL graph, that kind of strategy would require a cushion of about 60 MW, or roughly 2% of total system capacity. But with increasing fuel prices the cost of keeping an extra 60 MW of capacity on-line 24 hours a day has become prohibitive. So utilities have turned to frequency regulation strategies as a means of reducing their costs while maintaining reliable service. The following table summarizes the hierarchy of frequency regulation resources that has become a mainstay of today’s utility industry.

		Required Response Speed	Normal Cycle Duration	Period Between Cycles
Voltage Control	Online reactive reserves with automated controls that respond instantly to maintain system voltages within required ranges.	Seconds	Seconds	Seconds
Regulation	Online power sources with automated controls that can respond instantly to minute-to-minute changes in system load and correct for normal fluctuations.	Seconds to <1 minute	<1 minute to 10 minutes	Minutes
Spinning Reserves	Online power sources with automated controls that can respond instantly to major outages or demand spikes.	Seconds to <10 minutes	<1 minute to 10 minutes	Days
Supplemental Reserves	Standby power sources with manual controls that can respond to major outages or demand spikes within 10 minutes.	<10 minutes	10 minutes to 1 hour	Days
Replacement Reserves	Standby power sources with manual controls that can respond to major outages or demand spikes within 30 minutes to free up regulation, spinning reserve and supplemental reserve resources.	<30 minutes	>1 hour	Days

If you consider the first three classes of frequency regulation resources in the table and then revisit the red line in the ORNL graph, it’s easy to see how appropriately sized battery- and flywheel-based storage systems could accumulate surplus power during short intervals of low demand, return that surplus power to the grid during short intervals of high demand and offer substantial response time advantages over mechanical systems.

Frequency regulation has always offered critical cost savings as it improved overall reliability. But it’s important to understand how recent and planned additions of wind and solar power to the grid have increased the importance of frequency regulation due to the inherent variability of their power output. The graph below comes from the ORNL report and shows the typical two-day power output from a wind farm with 138 turbines. While the combined power output is huge, it is also highly variable. Without massive storage resources, the wind farm depicted in the graph will never be able to provide the consistent reliable electric power that customers demand from utilities. That’s why I believe cost-effective storage is a fundamental enabling technology for all emerging alternative energy technologies.

The wind farm depicted in the graph is a great visual example of why storage is essential. But it’s important to understand that battery and mechanical energy storage systems do not have the brute capacity required to smooth the power output for an installation like a large wind farm. At present, the only energy storage technologies that can smooth variations of that magnitude are compressed air and pumped-hydro. Nevertheless, the underlying point remains valid: wind and solar are inherently variable and the more variability you add to the grid, the more critical the need for frequency regulation becomes.

The generation, transmission and distribution of electricity is a regional business, but all electricity consumption is local and customers don’t care about 99.8% system reliability if they happen to be in the 0.2% of the service area that has problems. So the trick for utilities is forecasting what the likely demands of a particular service area will be and deploying frequency regulation and resources to provide the essential stability. Unlike generation capacity, which is frequently centralized in large facilities, frequency regulation resources will usually be located in close proximity to end users and distributed among a larger number of smaller facilities.

In a heavy-demand industrial or business district that is not prone to frequent outages or unusual demand spikes, voltage control and regulation will likely be the primary concerns even though the requirements may be extreme. So the first choice would probably be either a flywheel system or a battery system based on lithium-ion chemistry. As you move away from heavy-demand centers and migrate to the suburbs and beyond, regulation needs often become less important while spinning reserve capacity becomes more important. In less demanding environments, the first choice would probably be a less-expensive storage system based on either lead-acid chemistry or flow batteries. The key point here is that frequency regulation resources will be selected because they provide a cost-effective solution in a specific service area and a variety of solutions will be required to meet a variety of needs.

Over the past few months Altair Nanotechnologies (ALTI), Beacon Power (BCON), Axion Power International (AXPW.OB), ZBB Energy (ZBB) and A123 Systems (IPO pending) have each made significant progress on their respective frequency regulation initiatives. In particular:

• Altair completed preliminary testing of a battery energy storage system (“BESS”) that uses lithium-titanate batteries to provide up to 2 MW of on-demand power for 15 minutes of frequency regulation (500 kWh of total storage capacity). The PJM Regional Transmission Organization approved the Altair BESS last week, which means that the system can now be used to offer frequency regulation services in 13 states and the District of Columbia.
• Beacon completed construction of a 10-unit flywheel array that can provide up to 1 MW of on-demand power for 15 minutes of frequency regulation (250 kWh of total storage capacity). Beacon intends to expand the installation to 2 MW before year-end and 5 MW in 2009. The Beacon system was approved for commercial use in connection with ISO New England’s Alternative Technologies Pilot Program last week; which means that Beacon can now offer paid frequency regulation services.
• Axion completed fabrication of its first modular “Power Cube,” which uses lead-carbon PbC batteries to provide up to 75 kW of on-demand power for up to 3 hours of frequency regulation and spinning reserves (225 kWh of total storage capacity). The Axion Power Cube will begin an 18-month demonstration at a utility substation in New York as soon as the utility completes the permitting process.
• ZBB signed a distribution agreement for the commercial sale of its modular zinc energy storage system (“ZESS”) in Australia. The modular ZESS system uses zinc-bromine flow batteries to provide up to 25 kW of on-demand power for up to 2 hours of frequency regulation and spinning reserves (50 kWh of storage capacity per module).
• A123 completed fabrication and began field-testing of its first Hybrid Ancillary Power Unit (H-APU), which uses lithium-phosphate batteries to provide up to 2 MW of on-demand power for frequency regulation. While A123 has not yet released details on the energy storage capacity or discharge profile of its H-APU, I think it’s probably safe to assume that the H-APU will offer capabilities that are comparable to Altair’s 2 MW/500 kWh BESS.

A couple weeks ago, I explained that I don’t see the future of energy storage as a black or white proposition. I think lithium-ion batteries, lead-acid batteries, flow batteries, flywheels, compressed air and pumped hydro and will each make important contributions to the energy storage solution. I also believe that every company that is actively pursuing the development and commercialization of storage solutions for frequency regulation is likely to experience more growth than we can even begin to imagine.

Since all of the companies that are active in the sub-sector are in the early stages of their technical and commercial development, I believe a balanced portfolio of storage stocks is the only sensible approach for investors who don’t have the time or inclination to do their own detailed research. Articles like this one can provide useful insights, but they should not be relied on as investment advice because every author (including me) has his own agenda, preferences, predilections and prejudices.

Disclosure: Author holds a long position in Axion Power International (AXPW.OB) and is a former director of that company.

Comments

[john s. gordon:]

glad to see you have included pumped hydro in the above. response time to changes in load demand is pretty quick.

a problem with wind is the wind resource is located in sparsely populated areas far from load centers. this places an additional burden on the already stressed transmission network.
> jack

2008 Nov 25 08:58 AM

[John Petersen:]

I've included pumped hydro because it's very cool where the topography is suitable and it's excellent way to store days worth of energy cheaply. It is not terribly useful, however, for frequency regulation which requires huge minute-to-minute changes.

2008 Nov 25 10:24 AM

[creativforce:]

I don't have any mathematical ability, so I don't know what the long term economics of flywheels versus battery systems are, but it seems logical that a flywheel system like Beacon Power's would beat battery systems hands down. Once you build it and get it started, it runs virtually maintenance free forever because the weight is spinning on a friction free magnetic field. Every battery has a certain number of cycles before it needs to be replaced. The fact the BCON has floundered under $2 for the last five years and is now only .62 amazes me. Is the management inept? Or does it cost too much to make a flywheel? Is there someone smarter than me who can explain it? Why aren't energy companies, factories, and technology centers that need consistent uninterupted power buying these things like hotcakes? Disclosure, I have lost lots of money on BCON and no longer own it.

2008 Nov 25 10:32 AM

[John Petersen:]

Mechanical systems that require a high vacuum, rotate at high speeds and rely on exotic bearing and alignment systems are an extraordinarily complex undertaking; and despite the hype, nothing lasts forever. So it doesn't surprise me that BCON has taken years to develop prototypes to a point where they have a 100 Wh / 25 kWh flywheel that they think is ready for grid testing.

Now that BCON has a unit that can be plugged into the grid and tested, it will still take at least a couple years before they have enough data to support widespread sales to skeptical buyers. But the same can be said for any of the emerging storage technologies. There's just not enough data to satisfy a public utilities commission about what the costs and benefits will be. The data is coming, but it's not available yet.

One of the biggest problems in the storage sector is getting a firm grip on comparable numbers and I wish I could do a better job. While ALTI has sold prototype systems for $2 per kWh of capacity, Axion has sold prototype systems for $1 per kWh of capacity and ZZB is talking about price in the $0.60 per kWh range, each of their systems is designed to behave differently and provide different operational benefits.

Frequency regulation is a $20 billion smart grid opportunity because that's what it will ultimately take to replace the fossil fueled regulation assets utilities use today.

But anybody who claims to have the holy grail is blowing sunshine up your skirt and doesn't really understand the complexity of the problem.

2008 Nov 25 11:18 AM

[MILESCFA:]

John. Great piece(s). As soon as you mentioned "... battery and mechanical energy storage systems do not have the brute capacity required to smooth..", I thought about pumped-hydro. Boom, you mentioned it in the very next sentence (I have not head of compressed-air before now).

When people (like Boone) throw out wind "economic viability" numbers, I start watching where I'm stepping! Am I right to think their numbers don't include addressing the problems you site?

Like you, I waiting for the "next great battery" technology, which in my lay terms, has not changed in a thousand years (it's "creeped, not leaped")?

I have a friend who has a solar installation business here in Ga and he essentially won't install battery storage systems (exploding batteries and general cost/reliability).

2008 Nov 25 12:01 PM

[crtoca:]

You mention the need to place regulation resources in selected service areas. However, the various regulated grid operators - Independent System Operators (ISO) - like the California ISO (CAISO) - maintain markets that pay generators to provide frequency regulation, and there is no limitation or advantage on location for this service. Although it should make sense to place an energy storage system at the location of a frequency problem, like a wind farm, the current system allows the farms to push all their energy, and grid instability issues, into the larger grid. The ISO then buys frequency regulation services to manage the entire grid. The ISO accepts FR service from any generator in the larger grid, regardless of location.

Also, I disagree with your assertion that batteries, etc., don't have the "brute capacity" of compressed air or pumped hydro. It seems to me that a 100 MW battery system should have the same capacity as a 100 MW compressed air or pumped hydro system (although some would argue that the fast response from advanced energy storage devices would allow for smaller capacity systems providing superior service). Advanced Energy Systems (AES) can be scaled up to any size or distributed where needed. Compressed Air Energy Systems (CAES), which only exist in a couple of locations, and pumped hydro, sound great on a drawing board, but have such geographically restrictive requirements that they are of only limited potential value.

2008 Nov 25 03:41 PM

[nakedjaybird:]

You covered generation and storage, and to some degree, demand; however, we may have many opportunities to regulate (control) demand, including the regulation of bulk storage (pumped hydro, etc.) with variable continuous (speed, etc.) and flexible incremental loads (number, etc).

And, significantly, future smart metering and control of demand (loads) in residential, commercial and industrial HVAV and lighting (non-critical process demand) to refrigeration and local storage provides opportunity we have not begun to wrap our hands around.

2008 Nov 25 04:57 PM

[nakedjaybird:]

PS - just to keep the EV folks happy, recharging batteries at night at home are included in the demand control equation.

2008 Nov 25 04:59 PM

[John Petersen:]

As legal counsel for BFLS, a company that generates 15 MW of renewable peaking power in north Houston, I can assure you that location makes a huge difference in the efficiency of regulation, as does the cost of natural gas, the primary fuel for regulation. Instability from pushing today's tiny amount of wind and solar into the grid is not a problem, but as the percentage of variable capacity increases, so do the regulation problems.

Battery systems are designed for very short discharge durations (15 minutes for the ALTI and A123 systems and less for BCON). They will never be competitive against a hydro or pumped air system that's designed for hours of operation at a continuous load. Hydro and air are still relatively rare, but they'll be the technology of choice for the mega-projects because nothing else will do the job cheaply enough.

I've been tangentially involved with the Houston Advanced Research Center for years and they made the superconducting magnets for the SSC in the early 90s. When the project shut down, they began looking for alternative uses, including SMES. While the idea is cool (no pun intended) the systems are little more than giant capacitors that can provide mega-watts of storage but only seconds of discharge pulse.

There are a number of technologies that will ultimately be integrated into the smart grid. But since I'm a bit of a one-trick pony the only thing that really matters to me is that storage will be a big part of the solution.

You already know my opinions on PHEVs and EVs so I won't bother repeating them.

2008 Nov 26 02:18 AM

[nakedjaybird:]

Right, one-trick pony; the other ponies, the real demand control technologies will more than likely render current regulation "adequate" as the real manifestation of variable generation is sucked up by control of variable demand at the other end. Humongous surge capacity may indeed not be necessary. Certainly, the goal will be to minimize it as in any good business: excess inventory is a negative.

2008 Nov 26 02:43 PM

[John Petersen:]

The flip side of the coin is that fast storage with instantaneous response is twice as effective as spinning reserves and up to 20 times more effective than spinning steam when it comes to regulation. So the economies from reduced spinning reserves fuel can get pretty substantial pretty quickly.

2008 Nov 27 12:57 AM

[nakedjaybird:]

John - one does not want to spin steam - that is like having your ICE idling while stopped in traffic: come on EV's!

Here's how to do both on the demand side. Real data.


seekingalpha.com/artic...

2008 Nov 27 12:31 PM

[Art Reid:]

The real inherant problem with flywheels is the need to constantly add power because of the losses due to rotational accleration. That any point on the flywheel is always changing direction causing a continous loss of stored energy.


On Nov 25 10:32 AM creativforce wrote:

> I don't have any mathematical ability, so I don't know what the long
> term economics of flywheels versus battery systems are, but it seems
> logical that a flywheel system like Beacon Power's would beat battery
> systems hands down. Once you build it and get it started, it runs
> virtually maintenance free forever because the weight is spinning
> on a friction free magnetic field. Every battery has a certain number
> of cycles before it needs to be replaced. The fact the BCON has floundered
> under $2 for the last five years and is now only .62 amazes me. Is
> the management inept? Or does it cost too much to make a flywheel?
> Is there someone smarter than me who can explain it? Why aren't energy
> companies, factories, and technology centers that need consistent
> uninterupted power buying these things like hotcakes? Disclosure,
> I have lost lots of money on BCON and no longer own it.

2008 Dec 08 11:53 AM

[Ed C.:]

If the price for frequency regulation services roughly follows the price of electricity, how would BCON's business model work?

Wouldn't BCON be buying up the excess electricity at essentially the same price that it is selling it for? Is there a 1-2 cent / kwh premium added on top of the electricity price for the frequency regulation service?

Thanks for the great post.

On Nov 25 10:24 AM John Petersen wrote:

> I've included pumped hydro because it's very cool where the topography
> is suitable and it's excellent way to store days worth of energy
> cheaply. It is not terribly useful, however, for frequency regulation
> which requires huge minute-to-minute changes.

2008 Dec 26 03:48 PM