Rooftops, Energy Storage, and IOU Business Model

Posted on September 30, 2013
Posted By: Roger Arnold

In January, a report from the Edison Electric Institute painted a bleak picture for the future of investor-owned utilities.  It essentially said that the IOU business model, under the current regulatory environment, is not sustainable in the face of expected growth in rooftop solar[1].

 Dangerous feedback loop?

The crux of the problem is a positive feedback loop that raises utility rates as more customers produce more of their own power.  They use less power from the grid, and so the fixed costs of transmission, distribution, and generating assets must be recovered over a shrinking base  That drives up rates and gives customers that much more incentive to produce their own power.  Moreover, the fixed costs themselves increase, as the shrinking demand for kilowatt-hours torpedoes credit ratings and raises the IOU's cost of capital.  The end result could ultimately be wholesale bankruptcy.

Fans of distributed generation may cheer at the prospect of grid power becoming so expensive that anybody who possibly can installs panels, invests in energy efficiency, and generally does everything they can to minimize what they draw from the grid.  That sounds like the green Nirvana!  But wholesale bankruptcy among utilities would definitely not be good for the country.  The economic damage that would be done would not be limited to utility investors and customers dependent on them.

Nearly all rooftop solar PV systems that are installed today are "grid connected".  They depend on the existence of a working grid to meet power demand when the sun is not shining.  Most also depend on net metering  for economic payback.  Customers are charged only for the difference between what they draw from the grid and what they supply to it, which means that the utility is obliged to buy any surplus power that the PV system provides at the same rate at which power is sold to the customer.  The grid effectively serves as a free battery and backup system for PV customers.  It costs the utility to provide that service, but net metering shields PV system owners from any exposure to those costs.   Instead, the costs are passed on to the general pool of ratepayers.

If they did not have the grid to use in that manner, owners of home PV systems would see their capital costs more than double.  The levelized cost of power used would more than triple.  Batteries of the size needed to serve a home for many hours when the panels aren't producing are expensive to buy and expensive to maintain.  With the consequent higher cost of energy, disposable income and the overall economy would take big hits.

Addressing the Problem

Although penetration for rooftop solar is not yet high enough to cause a serious problem, the trends in panel and installation costs make it only a matter of time.  In the near to mid term, utilities can try to avert disaster by pushing for regulatory changes.  They would like to charge users monthly fees for grid connection, in addition to power consumed.  Also, they want to be able to buy surplus power at rates that realistically reflect the cost of transmission and storage and the impact of intermittency on generating equipment.

In the long term, utilities will need to evolve from a business model of selling kilowatt-hours to one of selling services.  At the top of  the list of services will be energy storage and backup power.  Those services are essential for a renewable energy economy, and utilities are in a position to provide them far more economically than PV system owners can provide them for themselves.  Utilities can capitalize on energy storage technologies that are simply not accessible at the scale of individual consumers.

Conventional wisdom is that there are not yet any generally available solutions for energy storage on the scale needed to deal with high penetration of renewables.  In most regions, variability is accommodated almost entirely by starting up and shutting down dispatchable fossil-fueled units - mostly gas combustion turbines.  That forces a heavy reliance on relatively inefficient peaking units.  That, along with the inefficiencies of the start up and shut down processes themselves, results in higher emissions.  The low average duty cycle for peaking units also results in high capital costs.  Nonetheless, in most locations, that approach is cheaper than installing the amount of battery storage that would otherwise be needed.

In locations that are geographically favored, conventional wisdom does recognize a better alternative: pumped hydroelectric.  Compressed air energy storage (CAES) may also considered viable for large-scale storage, but only where salt domes or abandoned mines make it easy to create high-volume compressed air storage reservoirs. For not-so-favored locations, R&D continues to focus on new battery technologies that might be cheap enough and durable enough to support grid-scale storage.

Storage options revisited

I believe the conventional wisdom needs updating.  There are straightforward technology options that offer the prospect for large-scale storage at a fraction of the cost of the best current battery solutions.  They have yet to be proven in commercial practice, but given the straightforward nature of the technologies, the risks seem minimal.  One such option is championed by Gravity Power LLC.  It's a close kin to pumped hydroelectric, but is not dependent on nearby mountains and preexisting reservoirs.

With Gravity Power's approach, the hydraulic head against which turbine / pump operates does not result from the difference in elevation between two nearby reservoirs.  Rather, it is created by the weight of a massive rock piston bearing down on water in the lower part of the shaft in which the piston travels.  That's illustrated in the two figures below, and explained in more detail in a video on Gravity Power's web site[2].

Figure 1A - GPM Discharging

Figure 1B - GPM Charging

Because the above-ground footprint for a Gravity Power system is so small, siting and permitting should be relatively easy. It is unlikely to rouse opposition from environmental groups.  Indeed, the first commercial Gravity Power system could end up being built in the Nekar-ALB region of southern Germany, where public opposition and zoning issues have stalled the construction of several conventional pumped storage facilities that had been called for in the district's Regional Plan.  In April, the plan was amended to designate Gravity Power Modules (GPMs) as an "acceptable alternative" to conventional pumped storage.

It may seem counter-intuitive that a storage approach involving excavation of a large, deep shaft 500 - 1000 meters down into bedrock could be economical.  But the figures that Gravity Power's CEO, Tom Mason, presented at a recent industry forum were not pulled from thin air.  They were developed in a design study by the German firm Babendererde Engineers, which has considerable experience in underground power and mining projects.  The figures were then vetted by the German engineering, procurement, and construction group (EPC) Hochtief.  Since Babendererde's project plan used only proven equipment and mining techniques, their cost figures can be taken as fairly solid - what a firm like Hochtief would expect to bid on a fixed-price EPC contract. 

For a demonstration peaking plant - 40 MW for 4 hours - the capital cost would be $4390 per kilowatt.  For commercial peaking plants in the range of 200 - 1600 MW capacity for 4 hours, the cost figures ranged from a high of $2050 / kw for the 200 MW plant down to $950 / kw for the 1600 MW plant.  The larger the plant, the lower the specific cost.  When the cost of "fuel" is factored in (surplus electricity for charging the GPM units, or natural gas imported to Europe at high cost), the larger GPM units are very competitive with conventional gas-fired peaking units. 

Isothermal CAES

Another approach to large-scale energy storage is isothermal compressed air energy storage (ICAES).  It is being pursued in competing implementations by companies including LightSail Energy of Berkeley, CA[3], SustainX of Seabrook, NH[4], and General Compression of Newton, MA[5]

Isothermal compression and expansion are processes in which compression and expansion occur with no change in temperature of the gas.  They respectively minimize the work needed for compression and maximize the work obtained from expansion of a given amount of gas.   True isothermal compression and expansion are thermodynamic abstractions that can't be realized in practice.  However, quasi-isothermal variants can be implemented and are almost as efficient. 

In quasi-isothermal compression, the heat of compression is absorbed by heat transfer between the gas, as it is compressed, and a heat storage material.  The heat transfer limits the temperature rise within the gas.  During expansion, the process is reversed; the heat storage material supplies heat to the expanding gas, limiting the temperature drop that would otherwise lower the gas pressure and reduce work output.. The actual temperature change during compression or expansion is determined by the heat capacity of the heat storage material and its amount in relation to the amount of gas. 

The best choice for the heat storage material is usually water.  It has a high specific heat capacity, good thermal conductivity, and of course is liquid and easily handled at the temperatures employed for ICAES.  All three of the companies mentioned above use water for the heat storage material.  They differ in the design of equipment and in particulars of how the interface between the air and water for heat transfer is arranged.  The name of the game is cost-efficieny, so the equipment must have a high specific throughput, without sacrificing too much in round trip energy storage efficiency. 

Achieving high specific throughput and high thermal efficiency requires the largest feasible contact area between the gas and the heat storage material, relative to their volume.  LightSail, for example, achieves that by turbulent mixing of a  spray of fine water droplets into the air that is being compressed or expanded.  The very high surface-to-volume ratio of the water droplets and the very short average distance between any molecule of air and the nearest droplet means that compression or expansion can occur quickly, without developing a significant temperature difference between the gas and the water droplets. 

The major attraction of ICAES is high energy storage capacity relative to both the size and mass of the system.

Both LightSail and SustainX are initially developing above-ground, "site anywhere" modular systems that can be built into standard cargo container modules.  They're competing with battery banks for off-grid military, disaster relief, or construction bases.  The modules have volumetric energy densities comparable to battery banks, but should be substantially lighter and more durable.  In modest arrays, they should be suitable for firming output from wind farms.  Both companies are interested in future grid-scale implementations using underground storage caverns, but that's not their immediate focus.  It is the focus of General Compression, which intends to use natural caverns in the West Texas area for power firming and load shifting for the many wind farms there. 


All of the ICAES approaches being developed suffer a significant complication, in that the pressure in the fixed-size compressed gas reservoir varies with the charge state.  That means that the air compressors and expanders connected to the reservoirs must operate over a wide range of compression / expansion ratios.

 To handle variability, LightSail, for example, uses piston motor-pumps in multiple stages, with electronic valve control and other features that enable the system to operate over the range of tank pressures.  While the approach works, it makes for a complex system with a low specific power rating compared to the Francis turbines and synchronous motor-generators employed in pumped hydro and GPMs.

In addition, gas already in a fixed-size storage tank is further compressed as more gas is added, or expanded as gas is withdrawn.  That means that heat must be removed from the entire tank as the tank is filled, or added to it as air is withdrawn.  Again, that can be done, but it's a complication that adds cost.

Hybrid Gravity Power-ICAES?

The complications of changing pressure in the compressed air reservoir could be neatly avoided by marrying a Gravity Power Module with ICAES.  The piston would be supported by compressed air rather than water, and the pressure in the compressed air reservoir below the piston would remain constant.  It would be the effective volume of the reservoir that would change as the piston was lifted or lowered, and not the pressure.

The switch from water to isothermally compressed air would increase the energy storage capacity of a GPM roughly eight-fold.  The exact amount depends on the actual pressure used to support the piston and on the air-water ratio used to limit the temperature change.  The radial motor-turbines that would be used for expansion and compression of mist-laden air would be larger than the Francis motor-turbines of pumped hydro installations, but not greatly different in design.  They would share the key feature of a ring of moveable guide vanes that vary the net flow rate over a wide range.  That allows the power to be throttled over a correspondingly wide range in both turbine and motor modes.  It makes the systems ideal for load following.  For mist-laden air as opposed to water, multiple stages would be needed.  However, the stages would remain relatively simple, delivering high specific power ratings.  The overall cost of the system should change very little, leading to at least a five-fold drop in cost per megawatt-hour of storage.

Such a large drop in the specific cost of energy storage would enable hybrid GPM-ICAES installations to address a much larger storage market.  They would not simply compete in the market for peaking power and spinning reserve; they could become so dominant as to effectively wipe out those markets.  Units with tens to hundreds of gigawatt-hours of storage are feasible, and could support whole-day or even multi-day load balancing.  They would draw energy indifferently from baseload power plants and intermittent renewables, and deliver it as needed.  The ultimate goal of 100% non-fossil power resources would become quite achievable.

Storage-Centric Utility Model

Whether large-scale / low-cost energy storage capability is achieved via the hybrid GPM-ICAES approach described above, or via a breakthrough in storage battery technology and lifetime, or by some other means, its arrival will transform the industry.  The business model for utilities will shift from one centered on generation to one centered on storage. 

Under a model centered on storage, rooftop solar and other forms of distributed generation are easy to accommodate.  The storage facility is analogous to a bank, and the T&D system to toll roads leading to and from the bank.  Owners of distributed generation would deposit surplus energy into the storage facility, paying a small transport fee to get it there.  They could later withdraw it for themselves or sell it to other parties needing the energy.

With sufficient low-cost storage capacity, the distinction between peak and off-peak power largely disappears.  There will certainly still be variations in demand over the course of a day, and peak demand could still strain the ability of the grid to deliver.  There will remain economic value in responsive loads that can reduce the swing between peak and off-peak demand.  But difference in power auction rates would become much smaller.  Wind farms especially could expect to see much more stable pricing for their output, even in de-regulated power markets. The need for negative pricing and curtailment would be eliminated.

I believe that price stability at realistic levels will ultimately be very good for renewables - however much their more vocal proponents might howl at losing the free ride they now enjoy at the expense of utility rate payers.

End Notes and References






Authored By:
Roger Arnold is a systems architect and engineer, recently focusing on energy systems and controls. His consulting company, Silverthorn Engineering, is developing architectures and software for motion control using custom switched-reluctance actuators and motor-generators. Mr. Arnold graduated from the Honors College at Michigan State University with a triple major in physics, math, and chemistry. He then worked for IBM’s Components Division in Poughkeepsie, N.Y. His budding career at IBM, however,

Other Posts by: Roger Arnold

Related Posts



September, 30 2013

Ferdinand E. Banks says

Only a matter of time, you say. Interesting. I was somewhere, in one of the University Buildings at my University - the University of Uppsala (Sweden) - when some crank who is now an important executive in an important bank explained to me the Beauty of Electric deregulation, and in the course of this hanangue mentioned households producing their own electricity, and utilities going broke.

I certainly am in favor of households producing their own electricity, but even though it has been about thirty years since that fool told me that I was wrong and he was right about the future of our Electric supply, I congratulate myself every time I remember that incident, because that was Another of those Beautiful occasions when I found out how smart and am, and how.---...well, I Think you know what I am going to say next.

September, 30 2013

Harry Valentine says

Hi Roger, I do commend you on your research into energy storage. In the physical world of energy, economy-of-scale has long determined the economics of power generation. With regard to roof-top solar PV power, the solar panels lose effectiveness over time . . . some delivering less than 50% of original output after 5-years. Electro-Chemical battery energy storage systems are typically rated for about 1,000-deep-cycle recharges and discharges. Both solar PV and EC battery technologies incur a high replacement cost.

While some power generation technologies can be feasibly downsized, most of these cost-competitive distributed power generation technologies will serve the power requirements of multiple buildings, including those in an industrial area. If an industrial park is located next to a river with a ramp, low-head hydro could deliver efficient and cost-competitive power over the long-term.

With regard to energy storage, there is definitely application in many nations for both short-term and long-term (seasonal) energy storage technologies. The concept you describe may be built into an emptied salt dome that occurs near ground surface . . . it could carry an electrically powered municipal and regional transportation system during the peak periods. Many cities operate electric subway trains as well as electric commuter trains . . . sometimes the mix includes streetcars and trolleybuses. In terms of economics, it make zero economic sense to equip several thousand transit vehicles with expensive electro-chemical batteries that will require replacement after a few years. A central energy storage system will last longer and be more cost effective over the long term.

In my view, low-grade solar thermal power that uses salt-pond solar thermal collectors . . . with organic rankine-cycle engines to generate electric power, may be cost-competitive over the long-term (20-years or more), despite low thermal efficiency.

September, 30 2013

Harry Valentine says

With regard to deregulation, many regulations that restrict competition are in my view, unnecessary and unproductive. I would support the repeal of regulations that forbid neighbouring private property owners from connecting private power lines across shared property lines. Most power utilities already own power transmission lines and local distribution networks. There is NO NEED for government to compel them to share their transmission and distribution systems with competitors.

I am all for privately funded (not subsidized) local power generation and distribution via privately owned power lines that connect across private property lines within neighbourhoods . . . whether a residential area, commercial area or industrial area. If PV-windows can generate cost-competitive power electric in a commercial high-rise building, so be it.

Local, regulatory-free small-scale power (EG: up to 10MW) is not going to put the big power producers out of business overnight . . . . and should therefore be allowed to operate, unfettered by political or bureaucratic restraints.

September, 30 2013

Roger Arnold says

@Fred, As an economics teacher, you have the background to understand exponential growth curves. Solar PV installations, both worldwide and in this country, have been averaging about 30% growth per year. That's a doubling time of about two years and four months, and it's been maintaining that rate for well over a decade.

Of course, as they always say, "past performance is no guarantee ..". But unless you can point to something that will clamp a lid on that growth very soon, I'd say yes, it is indeed only "a matter of time" before distributed solar PV causes trouble for the current IOU and regulatory models. That's the Edison Institute's conclusion, not my own. But I happen to agree.

I'll be the first to admit that the situation is volatile. Rooftop solar has reached the point that, in California at least, it is a nicely profitable investment for most home and business owners. But that's under the current regime of tax credits, subsidies, and net metering policies. Take those away and it would not be such a good investment. That would certainly crimp the growth rate.

OTOH, PV module costs have dropped well below $1.00 a watt wholesale, and are now only a fraction of installed PV costs. There's no good reason for installation and BOS costs to be as high as they are, and we're almost certainly due for multiple rounds of cost cutting and consolidation in those areas. So rapid PV growth can likely survive some reductions in subsidies -- as long as the reductions aren't too drastic or imposed too suddenly.

All this has nothing to say about what the best and most rational direction for future energy supply would be. I'm only speaking of what's actually happening, where the trends are likely to lead us, and what we might want to do about it. Develop storage, young man, develop storage!

@Harry, I didn't write about it here, but there's a different approach to CAES that's being pursued in Europe. You might like it better. It uses a combination of salt domes and thermal storage. Air is compressed adiabatically, and the resulting hot air is cooled quasi-isentropically, transferring heat to a thermal store. When compressed air is withdrawn, it's re-heated by passing through the thermal store in the other direction. That gives it this A-CAES the same theoretical energy density as I-CAES.

In the proposed A-CAES system, the cooled compressed air is stored in a salt cavern at constant pressure. The pressure is maintained by changing the effective volume of the storage cavern as air is added or removed. That's done by removing or adding a corresponding volume of saturated brine. For technical reasons related to high temperature thermal stores, I don't think A-CAES is as practical or promising as I-CAES. I could be wrong though. It will be interesting to see how it plays out.

Regarding loss of performance in solar panels over time, I believe that the incidents you're referring to are atypical. They're supposed to be related to poor quality control and production shortcuts in some financially stressed Chinese producers. Properly made crystalline and polycrystalline silicon PV cells and panels should live up to their advertised 30-year lifetimes with little loss of output.

Of course, there's the not-so-minor problem of how to determine if the cheap Chinese PV panels you're buying have been "properly made". You need to know your supplier, and ideally have a big enough operation to run your own sample testing and analysis. Or buy through an organization that can do it for you.

I don't want to talk about deregulation and access to T&D lines. The topic is too big and messy. But I probably agree with you about private lines between neighboring properties.

September, 30 2013

Harry Valentine says

Hi Roger, there are certainly some interesting variations and developments occurring in grid-scale energy storage technology . . . . a possible result of environmental opposition to pumped storage involving valleys in mountains. Now the pumped storage needs to go underground . . . and Gravity Power's technology certainly has future potential. The power storage level and storage duration happens to match the rush hour requirements of electric transit systems that serve several major cities internationally (New York City, San Francisco, Mumbai, Chennai, Sao Paulo, Rio de Janiero, London, Rome, Paris, Toronto, Montreal, Berlin, Moscow etc).

One novel technology from Hopper Energy Systems (I have no connection to them) is based on changing the density of a fluid or liquid ( ) that flows through turbines. Be interesting to see as to whether Hopper's technology could carry a big city electric transit system (subway trains, suburban commuter trains, streetcars and trolleybuses) through the twice daily rush hour periods.

October, 01 2013

Ferdinand E. Banks says

The more that we can get from solar and wind the better, and I Think that we can get more. But you see, while trying to live up to my belief that I am the leading academic energy Economist in the World - based on publications and professorships - I am ultra sensitive to the lies and misunderstandings being circulated about various energy sources. For example, the decision in Germany to dump nuclear and invest in solar and wind is strictly fruit cake, and behind that decision was the desire of Ms Merkel for Another term at the head of the Government government.

But I am not surprised. If you read German read the speech that Adolph Hitler made when he declared war on the US. Merkel and her foot Soldiers are pikers as compared to Mr Hitler's favings..

October, 09 2013

Finley Shapiro says

Mr. Arnold writes "Although penetration for rooftop solar is not yet high enough to cause a serious problem, the trends in panel and installation costs make it only a matter of time." It will probably be a long time.

While residential solar is growing, most of the growth in PV systems is in the larger commercial and utility scale systems. Utility scale systems are not a challenge to utilities. They either own the systems or buy the output as they would buy the output from other power plants, and they transmit and sell it to their customers as they do with electricity from other sources. Commercial scale systems, such as those found on factories and warehouses, do not seem to be the utilities' primary worry either. Tiered pricing can make sure that the value of this electricity stays low enough, and peak power trimming can be a benefit to the utility.

The primary concern to utilities seems to be residential rooftop solar, which is less than 1/4 of the installed PV generation and is projected to remain so. (See, for example, "U.S. Solar Market Insight Report, Q2 2013" by GTM Research and the Solar Energy Industries Association.) But the residential rooftop part of the solar market is limited by the number of houses with good rooftops for solar, the number of people who can afford the investment in solar, and the number of companies that are willing to invest in loans or PPAs for residential solar.

In spite of the fears expressed in the Edison Electric Institute's report, these factors are likely to keep residential rooftop solar well below a level that threatens electric utilities with "wholesale bankruptcy" for quite a long time.

November, 07 2013

peter snell says

Without the regulatory and statutory subsidies for PV and wind, they would be static or DECREASING in capacity as systems age and wear. Those subsidies will not long survive during stringent state audits and the prolonged period of stagnant economy caused by the policies of the current administration.

Add your comments:

Please log in to leave a comment!
back to top

Receive Energy Central eNews & Updates


Getting Physical with Cybersecurity

Wednesday Sep 7, 2016 - 1:00 PM Eastern - Your office

IT-OT-Physical Security Convergence for Critical Infrastructure Protection In spite of huge investments in security, attacks in critical sectors are on the rise. Most owners and operators of critical infrastructure assets deal with security in departmental silos which is not only more...

Demand Response Management Systems: The Evolution from Point Solution to Enterprise Platform

Wednesday Sep 28, 2016 - 2:00 PM Eastern - Virtual Event

Demand response management systems (DRMS) play a critical role in the deployment and management of demand response programs. more...

PLMA Demand Response Training Series hosted by Ameren

Monday Sep 12, 2016 - Friday Sep 16, 2016 - St. Louis, MO - United States

The PLMA Demand Response Training Series consists of three courses presented by industry experts and hosted at member offices. All three PLMA Demand Response Training Courses will be offered. Attend one, two, or all three. more...

28th Annual Independent Energy HR Association Conference

Wednesday Sep 21, 2016 - Friday Sep 23, 2016 - Charleston, South Carolina - USA


Mobile Utility Summit 2016

Tuesday Sep 27, 2016 - Wednesday Sep 28, 2016 - Denver, Colorado - United States

Real-World Solutions, Practical Advice, Field-Tested Strategies Mobile technology is delivering big dividends to utilities, helping them capture new customers, boosting loyalty and connecting field service forces with back-office operations. Now in its third year, Energy Central’s Mobile Utility Summit has more...

Smart Grid Customer Education Symposium

Thursday Sep 29, 2016 - Philadelphia

The Smart Grid Customer Education Symposium is the longest running, most influential gathering of utility industry professionals that meet annually to discuss best practices in smart grid customer education. Whether you are a utility looking to roll out Advanced Metering more...


Get your event listing in front of more than 100,000 industry professionals by posting on EnergyCentral's Event Center.

Sponsored Content