Blog Post

Pros & Cons: Distributed Rooftop Solar vs. Desert Solar Thermal

Utilities plan to use solar power from both the massive solar plants that are being built in the California Mojave desert, as well as large scale distributed rooftop solar projects, like the one Southern California Edison is planning. So which technology is better? Centralized solar systems that use the sun’s heat to generate electricity, or hundreds of rooftops covered in solar panels strung together to generate power?

Roy Kuga, the vice president of the Energy Supply Division at California utility PG&E, had some interesting ideas about the pros and cons of each technology at the Berkeley, Stanford CleanTech Conference Series on Wednesday. Basically, while solar thermal plants provide lower solar prices, higher efficiencies and better energy storage, distributed solar rooftop programs are quick to deploy, and less costly when it comes to transmission lines and water needs. Check out the detailed list below:

Distributed Photovoltaic Solar Rooftop Projects:


  • These projects can get up and running fast. Around 8 months, Kuga says, noting that the solar industry is also trying to bring down this time dramatically.
  • Distributed projects are not dependent on building long transmission lines to remote locations (such as the desert).
  • Distributed projects are also not dependent on the high water needs that solar thermal plants require for cooling.


  • Distributed systems have high costs of deployment. Because each system is a separate project, each rooftop installment requires a lot of labor, transaction and implementation costs.
  • They scale more slowly because it takes time to get all the rooftops up and running.

Solar Thermal Plants:


  • Solar thermal plants benefit from the economies of scale that can deliver lower solar power prices.
  • Solar thermal plant efficiencies are commonly higher than rooftop systems.
  • Solar thermal systems have pretty good energy storage technologies, so they are compatible with the intermittencies of solar. Solar thermal plants can store energy for when the sun goes down better than rooftop systems.

Cons: (Also check out our 8 Offbeat Hurdles for Solar Power Plants)

  • Solar thermal plants in the middle of the desert are transmission line dependent. Transmission lines are costly and difficult to get built.
  • Solar thermal plants need a lot of water, which is costly and delays permitting.
  • Solar thermal plants need a lot of land and require extensive permitting processes to get approved.

19 Responses to “Pros & Cons: Distributed Rooftop Solar vs. Desert Solar Thermal”

  1. timothy helfrich

    with all the flat roofs available on schools in everyones neighborhoods why not co-op our rebates and incentives and build arrays on property we ,technically,already own and share the rewards

  2. The NREL website has maps that show the areas suitable for CSP. The states include California, Arizona, Nevada, Utah, Colorado, New Mexico and West Texas.

    Why dispatchable power from CSP is better than coal.

    The above is one of a series of articles on core climate solutions, that taken together, will do what it takes to mitigate climate change and prevent the worst case scenario.
    Joseph Romm is a physicist and former assistant secretary of Energy for renewable energy and efficiency, during the Clinton administration.
    This is a site I highly recommend. Romm is a prolific writer and has knowledge of climate science, energy and how government works.

  3. We need both PV and thermal solar.

    Solar thermal and transmission lines:

    A study by the Western Governors Association found that 300 GW of solar thermal could be built near existing transmission lines.

    Solar thermal and Water:

    Solar thermal (or CSP) can be water or air cooled. There are several configurations for cooling.
    Water cooled plants can be combined heat and power, hot water and electricity. Or they can desalinize water, either exclusively or in conjunction with electric power production.

    We need the additional transmission lines anyway. Some of it can be shared by solar and wind farms. HVDC will allow long distance transmission with far less line loss than AC lines. HVDC does not have the big electromagnetic fields that worry many people with AC transmission lines.

    Solar thermal with heat storage produces dispatchable power that is more valuable than base load power, especially when trying to intergrate and balance various power sources like wind and PV solar into the grid. In fact, a grid with less base load power and more dispatchable power from CSP would be easier to balance, thus facilitating more PV solar and wind.
    This gives CSP the ability to displace the base load power of coal plants, which is a top priority for mitagating climate change.

    NREL says a parabolic trough CSP system with 6 hours heat storage has a capacity factor of 40% and that up to 70% capacity factor is feasable with more heat storage.

    The potential is huge for solar thermal. The California deserts alone have the potential for 660 GW of solar thermal, using only land with a 1% slope or less, and avoiding environmentally sensitive areas. Land with up to 3% slope is good enough.
    Compare that with California’s present total generating capacity from all sources- 58 GW.

    Using less than 1% of the U.S southwest desert areas for CSP could power the whole country.

    Opposing solar thermal in the desert for the sake of the desert ecosystem, is short sighted. If we don’t drastically reduce our CO2 emissions as soon as possible, like yesterday, climate scientists now say that 30% of the earth and the entire U.S. southwest will go into a drought that lasts 1000 years. Soil moisture content in the southwest would go down by 20-50%. A dustbowl would extend from California to Kansas.
    This would turn the deserts there into something like the Sahara. The damage to the desert ecosytems would be completely devestating. We will only use a tiny percent of the deserts to achieve the desired end. The roughly 100 mile by 100 mile area in which CSP could power the whole country would be stretched from California to West Texas. And I don’t believe we will build enough to power the whole country on CSP alone. We don’t need to. With PV and wind power added to CSP, we can power most of the country.

  4. Solar in the desert makes sense to me if you believe you can shade large swaths of land without unintended consequences.
    PV has the added green advantage of covering and already developed ecosystem. It is essentially an in fill solution.

  5. Pardon me, but what’s wrong with you (E2T) people?

    First, who cares about efficiency? We are talking about harvesting a resource that is pouring down at 1000 Watts per square meter on a sunny day. Efficiency is not the issue, overall cost is.

    And that’s the second point. PVs are really expensive. Not because of the install, or the separate sites, or whatever, it’s because the panels themselves are really expensive. 5 to 6 dollars per watt. Compare with 10 cents per kilowatt hour, and you see you need
    60000 sunny hours to recoup the cost of your panel. Assuming on average 6 sunny hours per day, that works out to more than 27 YEARS to recoup the cost of the panel itself. (Yes, the state of California offers $$$ credits for the panel, but that doesn’t mean that someone isn’t paying for all of this.)

    One should add “local thermal solar” to the list. 25kW units based on Stirling engine technology, deployed at the residential or neighborhood level. Less cost than PV, less transmission problems that the big thermal plants.

  6. Rob Bernier

    Given the pros and cons of each side, including the costs for the PV processing plant, and the turbine plant, how does the cost benefit analysis work out?

  7. George Hefferon

    Do we need to have solar panels on every roof? What about small community facilities to spread the initial investment. I also think that PPAs may emerge as the enabler of PV solar for residences and small commercial applications. I’d be interested in reactions to this latter thought.

  8. DS in NJ

    “Solar thermal plants need a lot of water”
    WTF?? The water is not consumed, just repeatedly heated & cooled. It doesn’t even have to be potable water.

  9. Real energy independence will come from individual solar roofs not central systems.

    This is recognized in Germany and Japan.

    What about footprint? Why do we want to continue using up land when there are already millions of square miles of existing roofs? What about terrorist attacks on a centralized power plant? Terrorists would find it difficult to blow up millions of roofs, unless we continue to allow Iran to build nuclear weapons.

  10. I’d add a couple more cons to the PV side.

    1. The distributed systems require a smarter grid, which is happening, but from what I can tell, pretty slowly.

    2. PV, to make a difference, will require individual consumer acceptance and even given all the services cropping up to make the decision easier, a basic monetary payback of more than 5 years is going to be a tough sell for many people.

    And a pro:

    1. Your PV setup will still work if the electrical grid goes down. It might be less efficient, but it’s more resilient.