Skip to content

Is the Future of Electricity Generation Really Distributed?

Renewable energy technologies have made outstanding progress in the last decade.  The cost of solar panels has plummeted.  Wind turbines have become massively more efficient.  In many places some forms of renewable energy are cost competitive.  And yet…just as these exciting changes are taking place, the renewables movement seems to be shifting its focus to something that has little or no connection to the fundamental environmental goals: distributed generation, particularly at the residential level.  In practice, this means rooftop solar PV.

Instead of seeking the most affordable way to scale up renewables, the loudest voices (though possibly not most of the voices) in the renewables movement are talking about “personal power”, “home energy independence”, “empowering the consumer”, and rejecting “government-created monopolies”.  In the not so distant future, residential PV may be augmented with onsite storage (as suggested by Tesla’s announcement this week of its Powerwall home battery system).


Residential is now a growing share of U.S. PV installations. Source: GTM research

The new emphasis on distributed generation has created a very unusual coalition between some traditional environmentalists and some anti-government crusaders.  Parts of the tea party movement have joined the Sierra Club in advocating for “DG-friendly” residential electricity tariffs, which mean high volumetric electricity charges in order to make rooftop solar economic.

I’m sorry, but count me among the people who get no special thrill from making our own shoes, roasting our own coffee, or generating our own electricity.  I don’t think my house should be energy independent any more than it should be food independent or clothing independent.   Advanced economies around the world have gotten to be advanced economies by taking advantage of economies of scale, not by encouraging every household to be self-sufficient.

That’s not to say that distributed generation couldn’t be the best way for some people at some locations to adopt renewables, but simply that DG should not be the goal in itself.  We desperately need to reduce greenhouse gases from the electricity sector, not just in the U.S., but around the world, including some very poor countries where affordability is a real barrier and electricity access is life-changing.  If DG is the least costly way to get that done, I’m in, but the choice should be driven by real cost-benefit analysis, not slogans about energy freedom.  TopazSolarFarm The 550 MW Topaz Solar Farm in San Luis Obispo County, California

The Pros and Cons

Compared to grid-scale renewables, DG solar has many advantages.  Generating and consuming power onsite means no line losses, which typically dissipate 7%-9% of grid-generated electricity before the power gets to your house. In addition, DG solar occupies your rooftop, a space that doesn’t have a lot of alternative uses, so the real estate cost is essentially zero.[1]  And as an extra bonus those solar panels also shade part of your roof, reducing the heat gain on hot sunny days.

In certain cases, distributed generation delays distribution system upgrades as demand on a circuit grows, because less power has to be shipped into the circuit on sunny days.  It also can reduce the need to build new transmission lines to carry power from distant grid-scale generation.

Having many small DG solar installations also spreads them around – spatial diversification – reducing the overall volatility of generation when clouds roll through.  Plus, spatial diversification and onsite generation can make the system more resilient to natural or man-made disasters, such as storms or sabotage.


The obligatory residential PV photo  (Source:

But distributed generation also has some serious drawbacks.  The first and foremost is that design, installation and maintenance of solar PV small rooftop by small rooftop costs a lot more per kilowatt-hour generated than grid-scale solar, probably about twice as much these days.  The scale economies that are lost with small systems on roofs of different size, shape, and orientation is a big disadvantage compared to grid-scale solar plants that are 10,000 to 100,000 times larger than a typical residential installation.  The size of grid-scale plants also makes tracking devices practical, which allows the panels to move throughout the day to continually face the sun and generate more electricity.

While small scale spatially-diversified generation could in theory reduce distribution upgrades and improve resiliency if the location and types of installations were optimized for those benefits, that’s not how DG solar is actually getting installed.  Systems are put in where homeowners choose to install for their private benefits regardless of the impact on the grid, and they can actually destabilize distribution circuits when they pump too much power back into the grid.  In Hawai’i, where 12% of houses now have rooftop solar, that’s already a serious concern.

Though it’s great that DG solar can contribute energy to the grid when the household doesn’t consume it all onsite, exporting power from the house reduces the DG advantage in line losses and distribution capacity upgrades.  For a typical residential system, at least one-third of the electricity generated is injected into the grid, though that may change with cheaper small-scale storage, one of the many technological factors in flux.

The technology installed with DG solar also is not optimized for the grid, so current systems aren’t contributing to resilience.  Solar PV installed today doesn’t have the smart inverters or the onsite storage that would be necessary for the systems to remain operational when the grid goes down.  Closely related, DG solar systems aren’t communicating with – or controllable by — the grid operator, so the system operator has to just guess when they might start and stop pumping power into the grid.

How do these pros and cons sort out?  Right now, I believe that residential solar loses to grid scale.  But I’m not convinced that will always be true.  And I don’t think that means households should be impeded from adopting DG solar today, just that we shouldn’t be giving it special incentives.   We need to recognize that DG’s role in the electricity future is uncertain and locking in on this (or any other) technology is unwise.  

An economically resilient system for renewables adoption

Well, then, how should we decided whether to go with DG renewables or grid-scale technologies?  We shouldn’t decide.  Instead we should design incentives that reflect the real benefits and costs of each type of system and then let them battle it out.  This has two big advantages.  First, it reduces the political fighting that comes with policymakers choosing one technology over another, or even the share that each technology should get.  Second, it pushes all alternative technologies to keep innovating and lowering their costs.

Designing such science-based incentives isn’t easy.  It requires detailed examination of each of the costs and benefits I’ve listed (and probably others that commenters will suggest).  It will not be possible to nail down each of these factors exactly, but we can’t make good electricity policy if we don’t carefully study what benefits and costs each technology brings to the table.  Tying renewables incentives to the best engineering and economic analyses of their net benefits will involve some heated debates about those analyses, but at least we will then be arguing about the right issues.

Then we should craft incentives that accurately reflect the net benefits each alternative technology offers.  I’m not sure exactly how those incentives should be structured.  But I can tell you that they don’t involve paying households retail rates for power injected into the system, as net metering policies currently do.  And they don’t involve maintaining retail rates that are many times higher than avoided costs — even including pollution costs — in order to create artificially high savings for PV adopters, as the current tiered electricity rates do in many states, especially in California.

They do include much greater use of time-varying pricing and, probably, location-varying pricing to reflect the real value of power on the grid.

Smart incentives based on careful analyses can reflect the dynamic value of distributed solar and distributed storage.  Curtailing net metering would boost the value of battery storage.  A lower cost of storage would smooth out prices over time and location, which would reduce the production timing advantage solar has, but would also reduce the problems of load balancing on individual circuits as DG solar ramps up.  Lowering volumetric residential rates would make end-user storage less valuable by closing the gap between retail and wholesale prices.

If DG solar with incentives that reflect its true benefits wins, that will be great, because we will know we’ve got the least-cost approach to reducing the externalities of electricity generation.  If it sputters, that will be fine too, because it will indicate that there are other less-expensive ways to achieve our environmental goals.  Either way, it’s time for incentives that are truly calibrated to costs and benefits, not to achieving penetration of one low-carbon technology over another.

[1] Though many people don’t have a roof for solar, either because they live in multi-family housing or, in the developing world, because the roof can’t hold the weight of solar panels.

To join the Energy Institute email list and receive notices of new blogs, working papers and events (one or two per week), click here

Severin Borenstein View All

Severin Borenstein is Professor of the Graduate School in the Economic Analysis and Policy Group at the Haas School of Business and Faculty Director of the Energy Institute at Haas. He received his A.B. from U.C. Berkeley and Ph.D. in Economics from M.I.T. His research focuses on the economics of renewable energy, economic policies for reducing greenhouse gases, and alternative models of retail electricity pricing. Borenstein is also a research associate of the National Bureau of Economic Research in Cambridge, MA. He served on the Board of Governors of the California Power Exchange from 1997 to 2003. During 1999-2000, he was a member of the California Attorney General's Gasoline Price Task Force. In 2012-13, he served on the Emissions Market Assessment Committee, which advised the California Air Resources Board on the operation of California’s Cap and Trade market for greenhouse gases. In 2014, he was appointed to the California Energy Commission’s Petroleum Market Advisory Committee, which he chaired from 2015 until the Committee was dissolved in 2017. From 2015-2020, he served on the Advisory Council of the Bay Area Air Quality Management District. Since 2019, he has been a member of the Governing Board of the California Independent System Operator.

71 thoughts on “Is the Future of Electricity Generation Really Distributed? Leave a comment

  1. Severin, thanks for the thoughtful post. I think most in the solar industry are in violent agreement that compensation to DG customers should be based on actual costs and benefits of DG solar. A variety of studies have shown that, depending on what you assume, what you count, and local issues such as current rates, net benefits can either exceed the retail rate (NV, ME), or come in below retail rates (E3/CA).

    I’d love to engage you and your colleagues in an effort to standardize evaluation of costs and benefits, as these issues are live in many states now. RMI has done some useful work here that could be extended,

    It’s also worth noting, as Jack Ellis points out, that utility scale renewables are increasingly difficult to permit and site, which drives some of the interest in expanding DG.

  2. I am in agreement that the rules that specify the rates which an owner of rooftop solar pays for grid power purchases or receives for sale of excess power should be based on solid economic grounds to avoid the creation of inefficient investments or to encumber other electricity users with charges they are not responsible for. In my view, this issue goes beyond roof top solar to include utility scale renewables. It even goes further in my view to include non-renewable DG whether industrial, commercial, as well as utility scale DG. A question that was put to me as an energy consultant in Saudi Arabia was to comment on the rules for the pricing of power purchases and sale of excess power of roof top solar. I share the basic elements we addressed:
    • Net metering is being practiced as a method to charge or pay a roof top solar owner for power purchases and sales. The rate being the same rate whether the home owner has rooftop solar or not.
    • At the same time, any DG facility including roof top solar needs to determine:
     Supplementary power requirements in case roof top capacity is not sufficient to meet peak demand,
     Standby power requirements to meet breakdowns, and
     Required interconnection or changes to interconnection to the grid.
    • A home basically has a contract with the power provider at set rate, for a maximum connected load, and no minimum consumption levels.
    • The power provider integrates in their planning process the household impact on a broad range of key measures: baseload generation, peaking capacity, and reserve margin.
    • A household adopting rooftop solar would: 1) reduce its demand on the grid to residual power only, 2) establish demands for standby power, 3) create a burden for the utility in treating the intermittent nature of power exported to the grid, and 4) may require changes to interconnection to the grid.
    • The introduction of rooftop solar would reduce the demand on grid power reducing revenue and increasing the cost of ancillary services. Both of which are not accounted for by net metering. The net effect is increase to utility costs and grid based rates.
    • The rooftop application at the same time is contributing to reduction in peak demand for power and benefits in addressing externalities whether reduction in line losses and emissions.

    In the event rooftop solar is broadly adopted, the parameters described here can result in major losses to utilities, major increase in rates, and ultimately the imposition of subsidies. This is the situation Germany finds itself in today. Much more can and need to be said on this issue.

  3. Severin, this is a terrific post and I violently agree with most of it. However I’d like to comment on a couple of points.
    First, rooftop solar actually may make good sense in places where there’s not much flat land for PV, like Hawaii, Taiwan or Japan. Second, the grid operator lacks visibility and control over individual distributed load nodes, so I’m not sure why the lack of detailed visibility over DG is more problematic. The fact that grid operators rely heavily on a tool called a state estimator is indicative of the fact that they’re a lot less certain of the state of individual devices on the high voltage network than they lead many people to believe. It may actually make more sense for them to increase the density of sensors that sit on the wires than to insist on invasive monitoring and control at each node in the network.

    The tow factors that could swing the balance toward rooftop solar more than any other is the difficulty of permitting large scale solar, which has already move development for sales to California away from the desert areas, and the cost, time and difficulty of permitting additional transmission.

    • Yes, and most of the profitability of solar panel for individual users in Germany now comes from self-consumption, and the reduction on the cost of electricity it brings.

  4. Love this post. Severin Borenstein, the only honest man in renewables. When you are cursed by all sides, you’re onto the truth! I’m not saying this is exactly what will happen nor that I even support every bit of what you’ve said as there are political realities in all policy, but this is the voice that is so sorely needed in all of our energy policy debates.

  5. I suspect the geographic diversity benefits of distributed renewables is overrated, because often enough, there are big storms that cast a shadow over thousands of square miles for many hours at a time, causing most or all of the local solar to shutdown. The benefit from dispersed exposure to scattered cloudiness is slight compared to this widespread problem when a big storm comes through.

  6. This whole post is fantastic, but I my favorite part was this:

    “I’m sorry, but count me among the people who get no special thrill from making our own shoes, roasting our own coffee, or generating our own electricity. I don’t think my house should be energy independent any more than it should be food independent or clothing independent. Advanced economies around the world have gotten to be advanced economies by taking advantage of economies of scale, not by encouraging every household to be self-sufficient.”

    Hahahaha, this is the perfect response to the “home energy independence” advocates.

    • Agreed with a smile. We know that Dr. B. has condensed a lot of economics into that sound bite.

    • Spoken by a true consumer! There are lots of things households still do for themselves, they cook for themselves, clean for themselves, wipe their own behinds, drive themselves to work or charity organisation. The solar panels on homeowner’s roofs would certainly be mass produced economically and not homemade. In the analogy solar panels would be the shoes and the electricity generated the walking in those shoes. Surely the author is not suggesting the walking should be carried out by professionals at economy of scale or that people should be watching marathons on TV rather than exercising for themselves.

  7. Are home PV systems really protection against storms? My understanding is that after Sandy, home PV came back online much more slowly than other electricity, as the utilities weren’t responsible for the repairs.

    • I think the idea is that if you have home storage that provides whole house UPS functions it will keep the lights on in those houses. It can also lower the load as they bring the grid back up. Attempting to bring an entire fully loaded substation(s) back on line is more difficult due to the large step size of the load.

%d bloggers like this: