How to mislead with graphs: solar edition

Posted on

One of the great mysteries of debates around climate change is why so many people actively exaggerate the growth and current levels of renewable energy. If you really cared about climate change  you should be motivated to accurately appraise where we are, and such appraisals are always sobering. A belief that renewable energy is supplying more energy than it really is should be counter productive. Yet, any observer of environmentalist’s attitude to German solar power will realise this is not the case. Hype and exaggeration is all that we get.

Consider these two facts: solar panels provided 50% of German electricity for a couple of hours one Saturday afternoon last May, but in total solar provided 4.6% of German electricity last year. The second statistic is what matters, yet any survey of green reporting of the issue shows that peak solar power is all that gets reported. Many people even believe that German now gets 50% of its electricity from solar all year round, not just on the occasional sunny afternoon.

And this leads me to one of the more egregious examples of this. CleanTechnica is not what you would consider calling a site that provides sober analysis. The general attitude is probably “read that press release, and hype it up some more.” So this recent report on German solar is not exactly surprising. But the graph below is particularly awful, even by the stands of CleanTechnica.

Those things on the right are the output from the Fukushima power plant, and a new EPR nuclear reactor. Clearly the author thinks the average reader is stupid enough to draw the conclusion that solar power produces far more electricity than the Fukushima power plant.  Yet, if you redrew this graph to show average output of Germany’s 33 GW of solar things would not look so impressive. But I guess an average of 3.4 GW lacks the impact of a peak of 24 GW.

Now, as I fear I always need to point out, these observations are not “anti-renewables” or “pro-nuclear” they are merely pro-arithmetic.


9 thoughts on “How to mislead with graphs: solar edition

    Thomas Gerke said:
    July 15, 2013 at 5:13 pm

    Hi Robert, we are a little paranoid & patronizing today aren’t we?

    There is absolutly nothing misleading about this graph & article. The whole point of the article was giving showcasing more data & information about the impact of solar on the German power mix on one (sun-)day. What relevance does your nagging have to that subject?

    Since you are pro-arithmetic, could you point out the misleading error in the graph?
    It shows the pv solar output on an hourly basis. The net generation capacity of an EPR and Fukushima have been added as a reference. (because nuclear is commonly known to have high capacities)

    It would be misleading if I would have only put the 24 GW peakload output next to the 1.6 GW of a (working) EPR, but I didn’t do that did I?

    In reality there is nothing wrong with the information presented, but it does not follow your narrative. In reality the world is not filled with idiots who think that windturbines & solar panels are dispatchable power generators… that is your condescending worldview.

    The world is however full of so called “experts” that told & tell the public that solar can not possibly have any significant impact.
    Yet generating more than 200 GWh / 20% of the daily electricity demand is a significant contribution. The same goes for covering more than 10% of the electricity demand during summer months in both Italy and Germany.

    BTW: 200 GWh was more than what the nuclear fleet with a nameplate capacity of 12 GW produced that day… Unlike your unscientific hero David McKay, I am pro-reality… and in reality there are more factors to be taken into account than over simplified & misleading theories that don’t take real world data & scientific studies by leading academic experts of renewable energy systems into account.


      joffan7 said:
      July 15, 2013 at 8:02 pm

      So, Gerke, can you explain why there wasn’t a red bar for every hour of the day? Can you explain why the entire country’s solar-generated electricity should sensibly be compared against one nuclear reactor?

      In fact, if you had supplied an hour-by-hour comparison of solar output – that one sunny summer’s day – against nuclear output, both totalled for the entirety of Germany, that might actually have been a useful graphic.


        Thomas Gerke said:
        July 16, 2013 at 8:33 am

        It’s a output per hour graph.


        joffan7 said:
        July 16, 2013 at 4:56 pm

        Obviously. That doesn’t answer why you didn’t put a red bar in every hour of the day. And it doesn’t answer why you compared the whole-country solar output with anything less than the whole-country nuclear output of that day.


      Donough Shanahan said:
      July 16, 2013 at 8:19 am

      Yes to add why did you not use a solar graph for winter; or an average capacity factor for the year for the two plants? Why not compare the two power generators per unit cost of build and operation; surely that is the correct way to normalize the data for comparison?
      But your own reply shows that you are being disingenuous.
      “It would be misleading if I would have only put the 24 GW peak load output next to the 1.6 GW of a (working) EPR, but I didn’t do that did I? ”

      Actually you did. You just did it over one day.


        Thomas Gerke said:
        July 16, 2013 at 8:43 am

        -Because it’s an article about solar energy on July 7th 2013.

        -No, I did put 24 hourly outputs from solar next to the maximum output of a nuclear reactor at a given hour.

        The graph clearly shows that solar generation is zero at night and peaks at about noon. Everyone can fill the solar output with the nuclear blocks and reach the conlusion that the solar output is the equivalent of 8.75 GW of 24h baseload power.
        I don’t call the readers stupid by suggesting they somehow reach the conluusion that the noon peak means that solar is equivalent to 24 GW of baseload. I dunno how retarded the energy debate is in your mind or in your home country, but it’s definatly not the case where I am from.


      jmdesp said:
      July 16, 2013 at 1:38 pm

      Gerke, 200 GWh was a significant contribution, but maybe not as positive a one as you’d expect. This is covered in the previous blog entry “German solar power exports”.
      During the peak of production, almost all the solar production was exported instead of locally consumed (or more precisely, instead of “erasing” the high carbon fossil production as one would wish, the solar production caused it to be exported at low price to neighboring countries, even those which electricity mix is much lower carbon than Germany).
      Meanwhile when we look at the production curve of nuclear, it matches exactly 1 for 1 the curve of lignite coal, which means that 1 GW of nuclear capacity can exactly and directly substitute for 1 GW of lignite and allow to shut it own (yes, sometimes nuclear plants have a technical problem and are unexpectedly shut down, but so do also coal & gas plants, just as frequently). I feel that you are in favor of a partial and skewed reality where.

      BTW: I took the *exact* numbers published by, and they show 197.4 MWh produced by solar, against 198 GWh produced by the nuclear reactors. The bar representing that should have been at 8.22GW. But this very small difference is what’s important. What is significant is that if nuclear had one that day produced all day long at full power, the around 8.65 GW it was producing at 3 and 4 in the morning as well as at from 8 to 11 PM, it would have produced 207.6 GWh of power instead over the day.
      What we see instead is that when solar start producing faster than demand grows between 7 and 9 AM (it’s already producing 7GW at 8 AM where people on a Sunday morning are not really active) and when it produced most at 1 PM, nuclear reduces it’s production down to 7.7 GW to accommodate for it (exports can’t always do everything), And that’s this reduction caused by solar that you’re using to claim nuclear did no better than solar on that day !
      It probably takes all the expertise of the leading academic experts of renewable energy systems to skew and manipulate the true meaning of data this much.


        Donough Shanahan said:
        July 17, 2013 at 8:16 am

        Your claim
        ‘It would be misleading if I would have only put the 24 GW peakload output next to the 1.6 GW of a (working) EPR, but I didn’t do that did I? ‘

        My claim is that this is disingenuous as you simply picked the best day for solar instead of showing the overall contribution to the grid on a annual basis. You then continue to further dig a hole by saying

        ‘Everyone can fill the solar output with the nuclear blocks and reach the conlusion that the solar output is the equivalent of 8.75 GW of 24h baseload power.’

        To be not disingenous you must add in the basis for this claim and that this claim is not valid for other times than this day and that this is impossible during winter months for example. Picking a single day and holding this up as an example of how good a technology can be while not showning the flip side does not give a true representation of the abilitiy of said technology to deliver.


        joffan7 said:
        July 19, 2013 at 3:12 am

        Thanks for doing the work TG should have done, and that effect was precisely what I was expecting to see – that nuclear had to throttle back because of the uncontrolled flood of solar onto the grid.


Comments are closed.