CleanTechnica is the #1 cleantech-focused
website
 in the world. Subscribe today!


Clean Power PG&E Time of Use Rates and Solar Output

Published on February 13th, 2012 | by John Farrell

10

How Distributed Solar Can Reduce Electricity Prices

Share on Google+Share on RedditShare on StumbleUponTweet about this on TwitterShare on LinkedInShare on FacebookPin on PinterestDigg thisShare on TumblrBuffer this pageEmail this to someone

February 13th, 2012 by
 
 

What if installing more solar could reduce electricity prices? It’s already happening in Germany, world leader in solar power, and it’s likely to happen in the U.S., too.

Right now, the idea of solar reducing electricity prices seems silly.  After all, when subsidies aren’t factored in, the cost of residential solar will be higher than residential retail electricity prices in all but 3 states until after 2016.  But solar has two key factors in its favor:

  1. Electricity, like many things, costs more when in high demand.  And while many U.S. ratepayers on are flat rate electricity plans, the truth is that their utility pays more to deliver electricity on those hot, sunny afternoons in the summer when air conditioners are running like mad.  Utilities call these times “peak periods,” when electricity use spikes and they have to turn on every last power plant.
  2. Solar PV arrays tend to produce at their best during these peak periods.

The following chart shows how PG&E (a California utility) charges significantly more for electricity during the afternoon hours when demand is high, and how a south-facing, fixed-tilt solar array can produce a lot of electricity during those peak hours.

Solar output can actually match this peak curve better, if the panels are angled toward the southwest rather than due south, resulting in more late afternoon output.

Either way, however, solar adds electricity to the electricity system when it needs it most.  And when that happens, it supplants electricity that was previously supplied by the dirtiest and most costly fossil fuel “peaking” power plants.  In Germany, the sharp growth in solar power output (from 3 terawatt-hours in 2007 to over 18 terawatt-hours in 2011) reduced the cost of electricity during their mid-day peak period by 40%, almost completely eliminating the price differential between peak electricity and the base cost.

The process where solar supplants expensive peaking power is called the “merit order effect.”  It works because utilities buy solar power on long-term contracts and there is zero marginal cost to take solar electricity at any particular time (they’ve already paid for it).  The peaking plants, on the other hand, tend to sell their power on the spot market.  Therefore, every megawatt of additional solar on the grid during a peak period supplants a megawatt of peaking power, eventually putting those costly plants out of the picture.  Ultimately, it means that during periods of high solar PV output, there won’t be peak power events with higher electricity prices.

Admittedly, the U.S. has a ways to go.  Solar produced enough electricity for as much as 17% of peak demand in Germany in 2011, while one of the U.S. leaders in solar per capita – Gainesville, FL – only serves about 1.5% of its peak demand with solar.

But solar is growing at an exponential rate in the U.S., just as it did in Germany.  And since solar can provide power when the grid needs it most, there’s a lot more to its cost than cents and kilowatt-hours.

This post originally appeared on Energy Self-Reliant States, a resource of the Institute for Local Self-Reliance’s New Rules Project.

Keep up to date with all the hottest cleantech news by subscribing to our (free) cleantech newsletter, or keep an eye on sector-specific news by getting our (also free) solar energy newsletter, electric vehicle newsletter, or wind energy newsletter.

Print Friendly

Share on Google+Share on RedditShare on StumbleUponTweet about this on TwitterShare on LinkedInShare on FacebookPin on PinterestDigg thisShare on TumblrBuffer this pageEmail this to someone

Tags: , , , , , , , , , , ,


About the Author

directs the Democratic Energy program at ILSR and he focuses on energy policy developments that best expand the benefits of local ownership and dispersed generation of renewable energy. His seminal paper, Democratizing the Electricity System, describes how to blast the roadblocks to distributed renewable energy generation, and how such small-scale renewable energy projects are the key to the biggest strides in renewable energy development.   Farrell also authored the landmark report Energy Self-Reliant States, which serves as the definitive energy atlas for the United States, detailing the state-by-state renewable electricity generation potential. Farrell regularly provides discussion and analysis of distributed renewable energy policy on his blog, Energy Self-Reliant States (energyselfreliantstates.org), and articles are regularly syndicated on Grist and Renewable Energy World.   John Farrell can also be found on Twitter @johnffarrell, or at jfarrell@ilsr.org.



  • Pingback: 10 German Solar Energy Myths Bjørn Lomborg is Spreading

  • http://cleantechnica.com/ Zachary Shahan

    Note that we’re discussing TWh above, not TW.

  • http://k.lenz.name/LB Karl-Friedrich Lenz

    For some reason I can’t reply to the last comment by Zachary Shahan, so I’ll post the answer here.

    Solar_power_in_Germany at Wikipedia has 3,075 GWh for 2007, which would be 3.075 TWh. What would be the page where it is 3 GWh?

    • http://cleantechnica.com/ Zachary Shahan

      Yeah, I updated it. I don’t know where I was looking before, when I wrote that comment,.. or if my eyes were crossed. But I checked again and updated it,.. and now just saw your comment as well. Thanks for the catch. :D

  • http://k.lenz.name/LB Karl-Friedrich Lenz

    *Ahem* That would be 18 *TWh* of solar in Germany 2011, or 18.52 TWh to be exact.

    • http://cleantechnica.com/ Zachary Shahan

      Thanks! Corrected it here — will make sure John sees and corrects it anywhere else it’s published, too. Looks like the 2007 figure was actually, gigawatt-hours, though — probably where the slip-up originated :D.

      • http://k.lenz.name/LB Karl-Friedrich Lenz

        Both figures are TWh. The error comes from the fact that in German people write 18.500 when in English you would write 18,500, so it is easy for someone with English as a native language to misread the graph in the IZES study you cited.

        Growth from 3 TWh to 18 is impressive; from 3 GWh to 18 TWh is impossible even in Germany :)

        • http://cleantechnica.com/ Zachary Shahan

          Hmm, my source on wikipedia has it wrong too, then… it’s got it as GWh

          • KellyEO

            Zach,

            What’s the resolution on this? What are the proper terms John should be using?

            P.S. – doing some research and need to be accurate!

            Thanks!

          • http://cleantechnica.com/ Zachary Shahan

            The post above is updated with the proper terms.

Back to Top ↑