This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.

10,282 Members


Energy Storage and the Cost of Going Off-Grid

off grid utility 

  • New energy storage offerings from Tesla and other manufacturers are widely expected to enhance the attractiveness of rooftop solar power and other renewables.
  • However, recent analysis from the Brattle Group shows that even with rapid cost reductions, grid-independence will remain beyond the reach of most consumers.

Last month’s Annual Energy Conference of the US Energy Information Administration included speakers and panels on topics such as crude-by-rail, potential US oil exports, and the role of the Strategic Petroleum Reserve, all of which should be familiar to my readers here. However, the topic that really caught my interest this year was energy storage.

Storage has been in the news lately, particularly since the launch of Tesla’s new home and commercial energy storage products. In fact, Tesla’s Chief Technology Officer spoke on the first morning of the conference. Much of his talk (very large file) focused on Tesla’s expectations for the cost of storage to decline sharply as electric vehicles (EVs) and non-vehicle battery applications grow. Whether battery costs can drop as quickly as those for solar photovoltaic (PV) cells or not, storage is likely to become a more important factor in energy markets in the years ahead.

One of the most interesting presentations I saw examined a provocative aspect of this question. Michael Kline of The Brattle Group, which consults extensively on electricity, took a detailed look at whether rooftop PV and home energy storage might become sufficiently attractive that a large number of consumers would employ the combination to enable them to disconnect from the power grid entirely.  That would be an extremely appealing idea for a lot of people. The author of a book I received from the publisher a few years ago referred to it as a movement.

Most people by now appear to understand that solar panels alone can’t make a household independent of the grid. The daily and seasonal incidence of sunlight aligns imperfectly with the peaks and troughs of typical home electricity demand. This is why “net metering“, under which PV owners sell excess power to their local utility–effectively using the grid as a free battery–has become contentious in some electricity markets.

In a true off-grid scenario, net metering would be unavailable. Onsite storage would thus be necessary to shift in time the kilowatt-hours of energy produced from a home PV array. However, a standalone PV + storage system must be sized to deliver enough instantaneous peak power to handle periodic high-load events like the startup of air conditioners and other devices. Another presenter on the same panel had a nifty chart demonstrating how wide those variations can be, with multiple spikes each day averaging above 12 kilowatts (kW)–several times the output of a typical rooftop PV array.

Brattle’s off-grid model included PV and storage optimized to “meet load in every hour given a battery with 3 days of storage (at average load levels.)” Although that is still probably less than the peak load such a system would encounter, it is the equivalent of multiple Tesla “Powerwall” units and would only be practical with the kind of drastic cost reductions Mr. Kline assumed by 2025: PV at $1.50/W and storage at $100/kWh, installed. That equates to around a third of last year’s average US residential PV installation and 1/7th the estimated installed cost of Tesla’s offering on a retail basis.  

Mr. Kline framed this exercise as a “stress test”, not just of the off-grid proposition but of the future of the electric power grid. If many millions of customers were to “cut the cord” for electricity as others have for wireline telephone service, even a “smart” power grid would become much less important and might shrink over time. That same logic should extend to the power generators supplying the grid. If most consumers went off-grid, the value of even the most flexible generation on the grid, which today is often provided by natural gas turbines, would fall, as would demand for the fuel on which they run.

In Brattle’s assessment, despite the assumption of very cheap PV and storage, that prospect seems remote. For the three markets analyzed (California, Texas and Westchester County, NY) the levelized cost of energy (LCOE) for the off-grid configuration modeled was significantly more expensive than the EIA’s projected cost of electricity in those markets in 2025. In fact, for consumers in California and Texas, as well as in all cases of the parallel commercial customer analysis Brattle performed, PV + storage would  be expected to cost a multiple of retail electricity prices.

As Mr. Kline explained, under more realistic assumptions the comparison was likely to be even worse for off-grid options. However, his conclusion that , “going off-grid…is unlikely to be the least expensive option for most consumers” does not mean that some consumers would not choose to do so, anyway. To them, a premium of 10-20 cents per kWh might seem like a small price to pay for personal energy independence. Yet at that price, it is hard to envision it would become a mass-market choice. 

Mr. Kline made a point of reminding his audience that Brattle’s analysis did not mean that distributed energy  would  not be competitive in the future, or that it could not provide valuable services to customers and to the grid. Importantly, the figures he presented underlined the continued value of the power grid to customers, even in a future in which large quantities of PV and storage are deployed.  As he put it, “Distributed energy is a complement to the grid, not a substitute for it.”

By extension, flexible generating assets like fast-reacting gas turbines should also continue to provide significant value, especially during those seasons when daily solar input is low, and in locations where average sun exposure is generally much weaker than in the US Southwest and other prime solar resource regions.  As appealing as the idea might be to some, storage seems unlikely to make either the grid or any class of generating technologies obsolete for the foreseeable future. As Bill Gates recently observed, that has implications for the cost of a wholesale shift to current renewables and away from fossil fuels.

A different version of this posting was previously published on the website of Pacific Energy Development Corporation.

Photo Credit: Energy Storage and the Grid/shutterstock

Geoffrey Styles's picture

Thank Geoffrey for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


Joe Schiewe's picture
Joe Schiewe on Jul 8, 2015 5:47 pm GMT

Darn, I was hoping that all these renewables with their variable production impacts and ugliness would get off the grid.  I sure don’t want to subsidize or endure the impacts of hundreds of new power and gas line corridors, ‘smart’ invasions on power use and the high minded attitudes of the green promoters.  This makes me like the thought of inexpensive small modular reactors located near the demand even more.

Geoffrey Styles's picture
Geoffrey Styles on Jul 8, 2015 7:12 pm GMT

The average US residential roof area is apparently around 1,500 ft2. For at least several months of the year the most populous parts of the US receive less than 3 kWh/m2/day of sunlight, suggesting that at 20% efficiency a fully covered solar roof in the Northeast, Midwest and Northwest would generate no more than 2,400 kWh in November, December, and January. That sounds like a lot until you factor in keeping a couple of EVs charged up, which seems likely to be equally appealing to those attracted to the off-grid lifestyle. That would consume around 600 kWh/month, leaving an amount roughly twice today’s average household electricity usage. Of course storage capacity would have to be proportional to the size of the PV installation, to capture all the energy not used when generated.

It’s not impossible, but it strikes me as wasteful, at least economically, considering the low cost (to the individual and society) of using the grid as your backup.

Rick Engebretson's picture
Rick Engebretson on Jul 9, 2015 3:54 pm GMT

In the spirit of giving fair consideration for everybody, I entirely agree with this electricity overview.

Business leaders like Elon Musk should be given respect for what they have done. But also qualified for what they did not do. Google did not create internet capacity, they used internet capacity; they did not create data storage capacity, they used data storage capacity; they did not create CPU capacity, they used CPU capacity. So when experts like Musk take on creating new battery designs or new space rocket designs, they find out it is not as easy as they thought. Fairness spreads a lot of credit around.

For our part, in Minnesota, we apparently have a new “bioenergy” law on the books. From what I can understand, the law includes advocacy for heating use. If we can figure out improved ways to use biofuels for simple heating applications, we free up very high quality natural gas fuel for production of even higher quality reliable electricity production as you describe.

Joe Deely's picture
Joe Deely on Jul 9, 2015 8:22 pm GMT

To see some actual examples of solar generation go here

To look at energy for charging(Tesla) go here.

Geoffrey Styles's picture
Geoffrey Styles on Jul 9, 2015 10:18 pm GMT

Thanks for that. Looks like a 9kW system (not small) in Mass. averaged 800 kWh/mo. last year.

Mark Heslep's picture
Mark Heslep on Jul 10, 2015 2:28 am GMT

The average US residential roof area is apparently around 1,500 ft2.”

Much of which is distributed over the average residential “A” roof.  Only the southern side of an “A” roof is suitable for PV, and even that is often at less than optimal angles.  Then there’s the roof area shaded from surrounding trees or other structures, etc. 

Bruce McFarling's picture
Bruce McFarling on Jul 10, 2015 4:22 am GMT

It’s also inefficient because the solar generation at one site is more variable than the solar generation across a town or city, and because moving surplus power from one part of the country to another that has net-load with HVDC cross-haul transmission is less expensive than storing the energy, and because windpower and run of river hydro are available when the sun has set or when a rainstorm is passing through.

And its likely economically inefficient to build up the capacity to run residences in Washington or Illinois or Ohio or Connecticut on solar off the grid in December-February, given the generation that this would provide in May-August … and the extremely high cost of batteries for seasonal storage.

It would be as economically inefficient as it would be to require each new nuclear power plant to provide the generation and storage capacity required for seasonal and daily peaks … bearing in mind that while conventional nuclear power operated as in France offer scheduled variable generation, but that is in the aggregate, since the technique for varying output is most effective early in the fueling cycle, and in the last half of the fueling cycle they are operated as relatively constant output plants.

Willem Post's picture
Willem Post on Jul 14, 2015 4:16 pm GMT


This article explains in detail the PV solar, thermal storage and battery storage and propane generator required to be off the grid in a 2000 ft2, energy-efficient house. 

A house south-facing roof, 44 ft x 18 ft = 792 ft2.

About 10% of free-standing houses have such large, flat (no dormers), south-facing roofs. Other roofs are facing in unsuitable directions or are otherwise unsuitable, based on Vermont survey data.

Area of a 250 W PV panel, 64.5 in x 39 in = 17.55 ft2

Maximum installation 250 x 792/17.55 = 11.25 kW, 45 panels in 3 rows of 15; cost about $45,000

Northeast production = 11.25 x 8760 x 0.145 = 14290 kWh/y, or an annual average of 39 kWh/d; this is for a solar-south-facing roof, at the proper fixed angle for maximim annual production, with NEW, CLEAN, unshaded panels.

Less production for deviations from these ideal conditions. In fact, on a national basis, these deviations cause about 15% less PV production in Germany.

Northeast ratio of best summer month/worst winter month generation is about 3.8, but this ratio could be greater, say 6 – 8, if much snow and ice and overcast conditions occur.

Two EVs use 2 x 0.30 kWh/mi x 12000 mi/y = 7200 kWh/y

A standard house uses about 6,000 kWh/y; an energy-efficient house about 3500 – 4000 kWh/y

If no heat pumps, a 2000 ft2, energy efficient house uses about 600 gal fuel oil/y for heating and hot water.

If cold climate, minisplit heat pumps, about 75% (450 gal) of the heating and hot water is with heat pumps (increasing the 3500 – 4000 kWh/y) and the rest with fuel oil (150 gal).


Bob Meinetz's picture
Bob Meinetz on Jul 13, 2015 4:53 pm GMT

Geoffrey, your article implies that going off-grid means going on-solar. Yet for about $6,000 at today’s prices, homeowners can purchase a 8kW natural gas generator and a Tesla PowerWall battery system, permitting them to take their electricity completely off-grid  for a marginal price of 16¢/kWh. In some areas of the country, that’s already less than utility electricity.

Options – freedom – are a wonderful thing, except when they limit the options of someone else. What we lose with the blind assumption that distributed generation will be cleaner than utility electricity is the ability to take responsibility for our electricity emissions – should that not be the case. We’d be making generation less efficient, and making the job of lowering emissions exponentially more complex and expensive by transferring solitary point sources of dirty generation to millions of independent sources.

Geoffrey Styles's picture
Geoffrey Styles on Jul 13, 2015 7:09 pm GMT


You’re right about my assumption that off-grid = solar + storage. Not only was that the focus of the analysis on which I was reporting, but it is the combination I most often see in reference to going off-grid. It’s also perceived as being universally available in the sense that the sun shines everwhere, aside from major differences in annual insolation, while not every home is connected to the gas “grid”.

Your gas-generation option has other advantages, besides price. Because interruptions in gas supply are much rarer than power outages, let alone the cycles of PV cells, storage would only be necessary if peak instantaneous load were greater than the maximum output of the generator. That would allow storage to be sized differently: not just smaller but perhaps with a different chemistry or even based on ultracapacitors that excel at high discharge rates.

I also agree that choice entails trade-offs. Even the natural gas fuel cells marketed by high-profile companies with sterling green images may emit more CO2 per kWh than today’s most efficient combined-cycle plants.


Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »