Inexpensive Grid Stability Solutions

Tom Konrad, Contributor
July 07, 2011 | 16 Comments

With all the discussion of grid based energy storage for renewable energy integration, the two cheapest and most mature solutions are overlooked. They are also the solutions most often overlooked by investors captivated by the story of clean technology.

A few years ago, I put together some graphs to show this as dramatically as possible. I surveyed the available data on energy storage and other grid integration technologies for the costs of existing installations, and calculated average cost per installed kW (power), per installed kWh (energy) and round trip efficiency (the percent of energy lost through round-trip charging and discharging.)

The results are shown in the graph below. Keep in mind that the data is a few years old at this point, and all numbers are approximate, since they are culled from a variety of different sources. The graph shown is on a log-log scale, so a technology at the top of the graph is 10 times cheaper when it comes to delivering power to the grid than a technology on the horizontal line. A technology at the far right is 100 times cheaper for storing energy than a technology on the vertical line.

The most cost-effective technologies are closer to the upper right hand corner of the graph, and have relatively large bubbles (high round trip efficiency.)

Large Scale Energy Storage Technologies

The most talked-about energy storage technologies, Pumped Hydropower [PHES], Compressed Air Energy Storage [CAES], and molten salt thermal storage in conjunction with Concentrating Solar Power [CSP-Tower] can clearly be seen to outperform both batteries and flow batteries for energy storage applications. Note that the numbers are approximate. PHES is shown on my graph as slightly less viable than CAES, but the balance of opinion favors the economics of PHES. The CAES bubble may overstate the viability of that technology: there are only two operational CAES plants, which leads to considerable uncertainly in the construction costs for future plants. Similarly, the Economics of PHES may be understated. Each Pumped Hydro site is unique and has its own economics, and the best sites are likely to be considerably better than shown. Such sites will have existing reservoirs that can be raised and lowered at will to reflect current electricity supply imbalances. I discuss PHES in more detail here.

Hydrogen [H2], flow batteries, and conventional batteries are simply too expensive to be viable as an energy storage medium except in situations such as remote power, where transmission, demand response, and PHES or thermal storage are impractical. While the economics of large scale energy storage for Hydrogen compare well with those of molten salt thermal storage, the high cost of fuel cells makes hydrogen storage nearly useless as a power resource, and the low round trip efficiency means that much energy is lost transforming electricity back and forth in to hydrogen. Large hydrogen tanks are relatively inexpensive to build, but filling and emptying those tanks is too slow a process to be practical as a grid based storage solution.

Yet all these solutions pale in comparison to the virtual energy storage provided by high voltage transmission. When a region has excess electricity, it usually makes much more sense to sell it to a neighboring region that can use it than try to store the electricity locally. Hours or months later, the same transmission line can be used to re-import the power when relative prices in the two regions reverse, making a transmission line to a neighboring region act as if it were a connection to a battery with infinite capacity.

Grid Stability (Power) Technologies

While the energy storage technologies on the right side of the graph are good for smoothing out long term imbalances between electricity supply and demand, short-term variations in supply and demand are best addressed by the cheap power resources towards the top of the graph. The quicker the fluctuations that need to be smoothed, the more important it is that the technology be able to absorb or deliver power quickly, and the less important it is that a large amount of energy be stored for extended periods of time.

Three highly effective technologies for producing quick bursts of high power but without much energy storage capacity are flywheels (currently in their earliest stages of deployment by Beacon Power (BCON), Superconducting Magnetic Energy Storage [SMES] a technology provided by American Superconductor (AMSC) that has been shown to be able to maintain grid stability when events such as lightning strikes would otherwise overload the grid with large, sharp jolts of power, and ultra-capacitors such as those provided by Maxwell Technologies (MXWL) which are generally too expensive for grid based applications, but are beginning to find a niche in vehicles. These technologies are not shown on the graph because I would need to expand the vertical axis multiple orders of magnitude upwards.

Among established technologies, Lithium-Ion [Li-ion], Nickel-Metal Hydride [NiMH] and Lead-Acid [PbA] batteries perform acceptably in remote grid stability applications where few other options exist, but all are eclipsed by the low cost and effectiveness of Demand Response. Demand Response is a suite of technologies which allow the utility to ask energy users to reduce their energy usage when the utility's generation capacity has trouble meeting current demand. Like transmission, but unlike batteries, flow batteries, thermal storage, PHES, or CAES, the electricity storage provided by demand response technologies is virtual: when a customer temporarily turns up the thermostat in response to a signal from the utility, the use of energy to cool the building is delayed until after the event when the customer drops the thermostat back to its usual setting. This avoids the cost of physical electricity storage, and makes Demand Response the most economical way to meet short-term spikes in energy demand (such as on hot summer days when air conditioning demand is high) and short term supply shortfalls, such as when power plants fail to come online at the scheduled time, or when power output suddenly drops.

The Bonneville Power Association's Hydropower Surplus

Recently, a heavy snow pack and a quick melt have caused the Bonneville Power Association (BPA) to shut down wind power generation for several hours each night in the Colombia Gorge. This has wind farm owners (who stand to lose Federal tax credits for energy production) heading to court. BPA claims shutting wind farms is necessary, but wind farm owners claim that two inexpensive solutions exist to deal with the excess power: Unused transmission capacity to Canada and Southern California, and the possibility of paying customers to shift their energy consumption from daytime to nighttime hours. Both these solutions would cost BPA money, while their current approach of refusing to accept wind power at night is free. This is why BPA chose not to honor its contracts with wind farms. While this makes economic sense for BPA, it sets a bad precedent because it was poor planning on BPA's part to sign such contracts in the first place. Should wind farm owners have to bear the financial consequences of BPA's bad planning? If they had known that BPA would not honor those contracts, they might have spent their capital in other regions of the country where the most productive season for wind does not correspond with the most productive season for hydropower.

In my opinion, this ruckus is more about industry players jockeying for position, than about wind being too unstable for the grid or incompatible with salmon. Both Demand-Response and Transmission are existing, cheap ways to deal with the potential power surplus, and no matter what the courts rule, Demand Response and High Voltage Transmission are both key in allowing wind to achieve its full economic and development potential. In fact, Tim Healy, CEO of Demand-Response firm EnerNOC in a recent interview, said his firm has been helping BPA shed some excess power duing nighttime hours by turning some of their customers equipment on when it might otherwise be off.

Investments

I've written extensively about transmission stocks in my "Strong Grid" series. I included the two exchange-traded Demand Response companies, EnerNOC (ENOC), and Comverge (COMV) in my recent list of Ten Clean Energy Stocks I'd Buy Now, because their prices are looking very attractive. I've already written about World Energy Solutions (XWES), and I spoke with EnerNOC CEO Tim Healy about his company last week in preparation for a this article. I plan to follow that with articles about Comverge, and EnergyConnect (ECNG.OB), an OTC-traded demand-response provider.

This article was originally published on AltEnergyStocks.com and was reprinted with permission.

DISCLOSURE: Long COMV, ENOC, BCON.

DISCLAIMER: Past performance is not a guarantee or a reliable indicator of future results. This article contains the current opinions of the author and such opinions are subject to change without notice. This article has been distributed for informational purposes only. Forecasts, estimates, and certain information contained herein should not be considered as investment advice or a recommendation of any particular security, strategy or investment product. Information contained herein has been obtained from sources believed to be reliable, but not guaranteed.

16 Comments

Gerry Wootton
July 13, 2011

Longwatcher - you're on track. Geographic diversity stabilizes variable generation. It is a good observation that storage need not be 100% of capacity: only a reasonable amount of ride-out capacity will do.
For a working example of geographic diversity, consider hydro power along North American west coast: some reservoirs are drawn down after several seasons of drier than normal weather while others are fully charged. Since power is 'shared' up and down the coast, local variation (-20% or more power availability) has little impact: certainly not so it receives any press. This is an easy exercise to do as production stats are available.
Typically, solar capacity leads demand by a bit less than 2 hours. For a totally local solution, storage in the order of 20% of daily production is sufficient. The cheaper solution of transmission would require a west-east transport of ~3300 km - sounds like a lot but there are transmission networks in North America that do that and more. Geographic diversity also means a diversity of sources where the use of one resource (say solar) defers the use of another or, more likely, provides a utility with exports to its neighbors. The advent of modern switch gear and power inverters has made high-voltage DC transmission practical; this is less lossy and can be applied in buried or submarine cable which helps to mitigate ROW issues.
One aspect of solar power that has not been heavilly considered is resiliance to insolation. Even on relatively clear days, a portion of insolation is indirect light which can be utilized by flat-plate modules. So far, the emphasis is on module peak power (that's how they're priced) not productivity; that can change as basic cell efficiencies go up and costs go down. Eventually, module manufacturers can take on diffuse and low light level yield as performance criteria mitigating cloud effects. Currently, DC/DC microinverter technology provides some accomodation at the array level for cloud effect.

Tim Dolan
July 12, 2011

Just thought I would share - it regards predicting clouds moving over larger solar arrays - basically it becomes relatively easy. Still easier if the solar PV is distributed, but they are demonstrating that even single large arrays can be reasonably predicted (or at least that is my take on the article)

http://www.sciencedaily.com/releases/2011/06/110621141851.htm

I am thinking if it was required to install a small load balance device with my solar panels to give the grid time to adjust to a cloud passing over (or worst case a cloud front passing over - the only case where there would be a wide area drop in power generated with a distributed solar PV system) I would be okay with that required installation. I would also be okay with that device talking with the grid to let it know a decrease or increase in power is coming and then level out that increase or decrease.

Just my thoughts,

Tom Konrad
July 12, 2011

mitch3-
The cost advantage of Demand Response over most storage solutions is currently enormous for short-term, infrequent dispatch, and the cost of DR is likely to fall at least as quickly as storage costs since DR is IT while storage is bounded by chemical and physical constraints.

GeraldR- Excellent points, I've nothing to add.

Garth Barker
July 12, 2011

I think demand response will help the situation today and help avoid some congestion issues as well as balancing issues however we must remember the economic arena we in at this time; if thing improve and its more costly to reduce, say, a factory's out put thru demand response than to buy stored power or even peak power those demand response contracts will dissolve. Its all about economics; good times change the way we do business.

Gerry Wootton
July 12, 2011

If one looks at some of the regulatory actions surrounding hydro-electric projects, it's not surprising that this is waning. The burdens imposed on operators are huge and include a lot of indirect liability: they not only have to take care of their direct impact, they have to support parasitic uses; for example, they become obliged to private associations or even government agencies that decide a flood pond would be a good place to raise fish or milk tourists. With the cost of compliance now nearly doubling the operating cost, it may be less risky for a company to just focus on NG power plants. I think there is a fair chance that even dedicated pumped storage operators will find themselves in a regulatory choke hold of a similar kind.
While it seems obvious that the grid could benefit greatly from some of the low cost technologies cited, it also seems obvious that this can't happen without political will of the right sort. And that means less interference from special interests including erstwhile environmentalists and the legion of NIMBYs. I suspect that demand response is the most feasible - at least until utilities discover its great potential to reduce their revenue.

Gerry Wootton
July 12, 2011

With respect to demand response, I know of an individual who can start up the HVAC system at the cottage before they leave home using a smartphone or computer - the technology exists; it's only a question of whether we choose to use it. System operators could even implement a real-time bidding system for power to orchestrate demand (if they can do it for hotel rooms ...).
A serious problem is regulatory Interference. I also think that geographic diversity is a good low-cost means of stabilizing the grid; however, local greed and parochialism gets in the way. Many states have laws and/or politicians that oppose anything that would be of material benefit to their neighbors or of less benefit locally. This is a serious impediment to rolling out an effective broadly distributed grid. Some governors are on record as opposing any development that would create jobs or industry elsewhere even if it means higher rates for their citizens. There are even published studies factoring higher utility rates against local job creation. While a prime rule of quality managment is "avoid local optimization", quality of service is not a political virtue. Local FF power plant == source of campaign contribution. To all of this, the feds have added more levels of bureaucracy (e.g. FERC)to muddy the waters and make regulatory compliance a major component of cost (more expense for lawyers than line crews).
The lowest cost hydro storage, leave out the word pumped. Hydro can be moderately dispatchable and at least mostly available. Simply synchronizing dispatching of hydro to variable generation and/or demand gets the job done. However, some utilities actually run in reverse to this policy. One major issue is that a larger fluctuation in water levels and water flows is a likely consequence - this generally raises the hackles of upstream and downstream land owners. Pumped storage is more expensive but may be desirable as it can be implemented out of run i.e. create a dedicated reservoir.

Gerry Wootton
July 12, 2011

As usual, quite insightful Tom.
The issue is how to balance supply and demand in a system. Systematic solutions may be lower cost than the obvious technology ones. In addition to generators coming not coming on-line when scheduled or experiencing a low capacity factor, there is the case of generators being on-line when not needed.
Although everyone is quick to point out wind and solar as variable sources, many other sources are variable. Nuclear is a case in point: power availability is historically around 80%. The other 20% has to be made up from long term storage or, more likely, redundancy (more generators in the system). Poor dispatchability means that nukes have to operate below baseline. In a nuke rich region like Ontario, curtailment is the solution although underutilization is not a cost effective way to operate. Hydro has a substantial number of dispatching constraints, particularly since most conventional hydro is multi-use. Water flows must be managed to control water quality, prevent erosion, maintain water levels for recreational use and land use, manage a fishery including possible migratatory diversion, supply water for irrigation or utility water supply, etc. At some times water is spilled unproductively when demand is low; at others, power generation is curtailed in spite of demand.
Regulation of demand in the form of demand response (I believe in Europe they call it energy management) is a very economical solution. Level 1 is to shift or curtail loads based on TOU; level 2 is to do it in response to system operator scheduling. The communication technology already exists and is pervasive (Internet/WiFi) so cost is low. The primary investment is for point of load switch gear which can be deployed at any scale including residential. This can be implemented at the distribution panel and/or at individual loads/appliances. In Germany, there is a special rate program for small users who have energy management systems or appliances.

Tom Konrad
July 12, 2011

Nigel-
Molten salt thermal storage probably does not make sense for most direct use thermal storage except perhaps in some high-temperature industrial operations because it requires a minimum 350C operating temperature. Usually processes that need this much heat are fossil fueled, so there is no need for storage.

The Ice storage systems you're referring to are the central thermal bubble on the chart, labeled "HVAC"

Garth Barker
July 12, 2011

Very good article however there is one point I'd like to make concerning PHS; most studies done in the last few years (Sandia, NREL, California) have not considered the variable pump/turbines that will be installed in near term Closed Loop Pumped storage projects in the licensing process at this time.
These turbines combined with advanced stators can respond at rates conducive to 15 min markets; 5mw/sec and because of some foresight in siting have removed most if not all of the environmental and recreational nexus that have plagued PHS since the formation of the EPA. The biggest hurtle at this time with the first of 11 preliminary permitted projects is the transmission interconnect; these storage projects do replace the need for natural gas integration of variable energy and offer a clean stable product for ramping and load following. The fossil fuel based utilities see their asset base becoming obsolete when these 600 to 1300Mw projects come on line in late 2015, early 2016 and they are fighting these projects as much as they can.

Nigel Morris
July 10, 2011

We've seen using ice generation overnight as an energy_storage/load_shift/peak_reduction application in southern states.
We've also seen molten salt storage recently becoming more developed with concentrated solar thermal power stations and in at least one proposed energy storage park.
Is molten salt (or some other heat storage), used purely as a thermal storage i.e. using excess wind generated electricity at night to heat salt and then use the stored energy purely as heat, during winter time in the BPA area a viable option to investigate? Using the energy purely as heat, to heat buildings during the daytime or even using it as process heat for industry, rather than return it to electricity, would remove one transformation loss.

Tom Konrad
July 8, 2011

etcgreen,
You're right, I already knew the answer when I asked, but I get into arguments about this with people fairly often, and it helps if I can back up my theoretical understanding with someone your real world experience.

Also, you discussion of solar cloud transients confirmed what I already knew, but the specifics (regarding maximum array sizes, etc are interesting.

Steve Frazer
July 8, 2011

Tom (Comment 3), you do not really need to ask as it is intuitive for someone with your experience.

Wind farm developers focus on economies by scale in grid connect, land cost, ... This concentrates the turbines which also results in more erratic energy levels. Geo-diverse farms access geo-diverse winds and provide more consistent energy levels.

We collected wind data along one ridge line that presented exactly this model (we wanted to distribute across 2Km), but we could not make the numbers work based on the additional capitalization in wire runs and land costs.

Modern NG turbines can respond quickly to demand, but the minimum safe margin continues to grow based on wind and clouds (solar). So utilities have to embrace increased cost for renewable energy and actually increase their safe margins.

Our local utility will not discuss a geo-point solar array larger than 30Mw as they do not have the capacity to load balance against a cloud.

Tom Konrad
July 8, 2011

Derek,
The viability of windfuels will all depend on conversion efficiency. The article you cite talks about a 58% conversion efficiency, which will give them a very small bubble size, a little smaller than the H2 bubble on my chart. The good part is that liquid fuels are very easy to store, which will put them on the right of the chart, and we have an existing infrastructure to use them well, which will put them towards the top.

So, while the position of the bubble will be very good, the size of the bubble will be small, meaning that Windfuels will only be practical with very low cost electricity. The question then is, who is going to build a wind farm to sell electricity at very low prices? No one. And that means that wind fuels will only be viable for the short time each year when there is an excess of electricity. That's fine, so long as the equipment to make wind fuels has very low capital cost. High cost capital equipment needs to be operated much more often in order to produce an economic return... which is exactly the problem with most EV technologies: the batteries need to be cycled several times each day to me competitive with conventional cars that have lower capital cost.

Tom Konrad
July 8, 2011

etcgreen,
Thanks for your comment. Given your experience managing a diverse set of wind generators, would you agree with the point I've made in many other articles, that while wind does require backup (build a gas plant first), a diverse portfolio of wind will require less backup than and equal amount of wind all operating in a single farm?

Derek Boyle
July 8, 2011

Another option is additional revenue streams for excess renewable energy:

Kicking Oil Addiction with Windfuels: "off-peak excess wind energy to recycle CO2 into standard fuels that work seamlessly in the one billion cars and trucks"
http://www.greentechmedia.com/articles/read/guest-post-kicking-oil-addiction-permanently-with-windfuels/

"Windfuels provide the potential for the U.S. to transition from the world's largest importer of oil to the world's largest exporter of carbon-neutral transportation fuels."…"synthetic fuels could be competitive when oil is as low as $50/bbl -- and always when oil is above $95/bbl."

Steve Frazer
July 8, 2011

I regularly visit a dozen renewable energy websites in an effort to maintain "perspective" on the deltas between our reality and the articles and bloggers' perceptions. I have worked in the renewable energy industry for the past 6 years and I feel most at home on this site and the NYT.

Mr. Konrad, excellent article and analysis. A good balance of "Just the facts, Ma'am" data and connecting the dots.

I was the PM for a load balancing system for geo-diverse wind turbines years ago and it is actually rather difficult. As a mathematician, you might enjoy the challenge of all the feed-back loops, historic demand data, short-term projected wind energy calculated from surrounding weather stations, weather variables, social event adjustments, ...

The Chief Engineer for the utility we were working with at the time would tell his favorite joke - "How do you replace a coal power plant with a wind farm? Answer: Build a natural gas power plant first".

I like your common sense approach - eliminate the demand load level.

We are taking the same common sense approach to what we believe has the potential to help pull the U.S. from its current economic status. It is already gaining good momentum - new sales records every month...

etcgreen.com Article: U.S. Migration

Add Your Comments

To add your comments you must sign-in or create a free account.

Tom Konrad

Tom Konrad is a financial analyst, freelance writer, and policy wonk specializing in renewable energy and energy efficiency. He manages green stock market portfolios. He writes articles about investing in clean energy for Forbes.com AltEnergyStocks.com....

To register for our free e-Newsletters, create your free account here:

Inexpensive Grid Stability Solutions

16 Comments

Add Your Comments

Tom Konrad

Editors' Picks

Most Commented

Total Access Partners