Guest essay by Chris Yakymyshyn
Vermonter Bill McKibben was recently quoted in Salon Magazine:
“The roof of my house is covered in solar panels. When I’m home, I’m a pretty green fellow. But I know that that’s not actually going to solve the problem.”
This is a very interesting comment. He had solar panels installed on his home, even though he knew it would not ‘solve’ the CO2 problem.
One goal of installing solar PV is to reduce CO2 emissions associated with generating electricity. Ideally, this would be achieved at a cost that is less than the social costs of CO2 emissions, estimated by the EPA to be somewhere between $12 and $117 per ton in 2015. To minimize the cost of avoided CO2 emissions, ideally a residential solar PV system replaces utility energy that is supplied by burning coal, since coal produces the highest CO2 emissions per kilowatt-hour. Likewise, adding solar panels in a location that already receives 100% carbon-free electricity will result in an infinite cost per ton of avoided CO2, since no CO2 emissions will be avoided. The reality at your wall outlet will lie somewhere between these two limits.
I decided to calculate how much it costs to reduce one ton of CO2 emissions by installing a residential solar PV system in Vermont. I then repeated the calculation in every other U.S. state and Canadian province or territory. This estimate assumes that electricity generated within a state, territory or province is consumed there, and that electricity imports constitute a small percentage of total electricity consumed within that state, province or territory.
The first step is to figure out roughly what percentage of today’s wall-plug power is provided by coal, natural gas, nuclear, wind, solar, hydro, geothermal, biomass, etc. The Energy Information Agency (EIA) tabulates, by year and by state, the total amount of electrical energy (in Megawatt Hours, or MWhr) delivered by each type of generating source. For the most recent year available (2011) in each state, the utility and IPP (Independent Power Producer) electrical energy generated by CO2 emitters (coal, natural gas, petroleum liquids) and non-CO2 emitters (nuclear, wind, solar, hydroelectric, geothermal and biomass) was extracted, with the assumption that biomass was carbon neutral. The ratio of fossil fuel to total electrical energy produced was then calculated for each state in 2011. The results ranged from 0.14% fossil electricity in Vermont, to 98.7% fossil electricity in Delaware.
The same tabulation was performed for Canadian provinces and territories using 2011 data from Statistics Canada. In Canada the results covered the entire range, from essentially 0% fossil electricity in Prince Edward Island up to 100% fossil electricity in Nunavut.
Next, the CO2 emissions per MWhr were calculated using the following emissions estimates: 1.4 tons/MWhr for coal, 1.0 tons/MWhr for fossil liquids, and 0.47 tons/MWhr for natural gas. The total CO2 emissions were estimated by multiplying the energy in MWhr produced from each source, by the CO2 emissions per MWhr for each source. The resulting CO2 emissions in 2011 ranged from <0.001 million tons CO2 in Prince Edward Island, 0.008 million tons in Vermont, up to 279 million tons in Texas.
The average CO2 emissions associated with electricity generation in each state, province or territory in 2011 was then calculated by dividing the total CO2 emissions by the total amount of energy generated. The resulting averages ranged from <0.001 tons CO2 per MWhr in Prince Edward Island, 0.001 tons CO2 per MWhr in Vermont, 0.567 tons CO2 per MWhr in Nevada, to 1.36 tons CO2 per MWhr (almost 100% coal) in West Virginia.
The amount of solar energy generated by a solar PV residential system was next estimated. The annual averaged hours per day of full sun for a South-facing fixed solar array tilted at latitude was extracted from the National Renewable Energy Labs (NREL) Renewable Resource Data Center. The values ranged from a low of 2.5 hrs/day in Yukon Territory up to 6.5 hrs/day in Nevada and Arizona. Assuming a 10 kW(AC) system with a 20 year service life and no aging, the total energy delivered by the rooftop solar PV system was estimated in Nevada to be (6.5 hrs/day)*(365 days/yr)*(20 yrs)*(10 kW(AC)*(0.001 MWhr/kWhr) = 475 MWhr of electricity. All of the generated electricity was assumed to be used somewhere in Nevada. This calculation was repeated for every state, province and territory.
The cost of the residential solar PV system was needed next. A recent article at Solar Panels Review gave 2013 price estimates for a contractor-installed system using several panel choices. The average unsubsidized cost was $5.57/Watt AC, or $55,700 for a 10 kWAC system. This unsubsidized cost is assumed to be the same everywhere.
The cost of CO2 emissions avoided using residential solar PV can now be estimated. The cost per ton CO2 avoided is given by the solar PV system cost divided by the total CO2 tonnage avoided over the 20-year life of the system. For example, using the previous estimates for Nevada, the avoided CO2 emissions cost is given by ($55,700)/(475 MWhr*0.567 tons CO2 per MWhr) = $207/ton CO2. This calculation was repeated for every state, province and territory and, as shown in Figure 1, plotted versus the fraction of generation that is free of CO2 emissions.
First, notice that the vertical axis is a logarithmic scale, ranging from $1/ton CO2 (well above the 5 cents/ton that traders at the now-defunct Chicago Climate Exchange determined was an appropriate price), up to $10,000,000 per ton CO2. Several horizontal lines indicate the California carbon exchange price of about $12/ton CO2 and one EPA estimate of around $60/ton CO2. A vertical line marks one widely discussed goal of 80% CO2-free electricity generation.
Note how the use of residential solar rapidly escalates the cost of avoiding CO2 emissions as the power grid moves towards a ‘low-carb’ diet. Also note that even in ‘high-carb’ states at the left side of the graph, residential solar PV is an expensive way to avoid CO2 emissions associated with electricity generation, never breaking below $100/ton CO2. Substituting DOE’s 2020 SunShot goal of $1.50/Watt installed cost for a residential system shifts the curve down, but retains the highly coveted hockey stick shape J.
So, Bill McKibben’s solar panels in Vermont are indeed avoiding CO2 emissions in Vermont (one of the data points at the far right side of Figure 1), at a cost of around $155,000 per ton CO2. This is equivalent to paying a carbon tax of $2.00 for one teaspoon of gasoline.
Figure 1- Semi-log graph showing the cost of avoiding one ton of CO2 emissions using residential solar PV, province or territory, as a function of the carbon content at the wall outlet. Several U.S. states and Canadian provinces are indicated. The two horizontal lines represent two official estimates of the social cost of carbon dioxide emissions.
References-
Salon magazine article-
http://www.salon.com/2013/09/15/bill_mckibben_being_green_wont_solve_the_problem/
Solar insolation data from NREL-
http://rredc.nrel.gov/solar/old_data/nsrdb/1961-1990/redbook/sum2/state.html
Electricity production in the U.S.-
http://www.eia.gov/electricity/data/state/
Electricity production in Canada-
Solar PV system costs-
http://solar-panels-review.toptenreviews.com/
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
30 years ago I did the math for a self-installed passive solar array in Kalamazoo, Mi. About 30 years. Average sunshine was 183 days per
year, most of it not in the winter. Insulation upgrades had a much quicker pay back. I’m not sure solar electricity is much different, unless it is heavily subsidized. Subsidies are nothing more than wealth redistribution schemes. We let the IRS and the utilities go to our neighbors and collect the money. If solar really made economic sense, then you would do it without the heavy subsidies.
…“The roof of my house is covered in solar panels. When I’m home, I’m a pretty green fellow. But I know that that’s not actually going to solve the problem.”
This is a very interesting comment. He had solar panels installed on his home, even though he knew it would not ‘solve’ the CO2 problem.
Um… he has said precisely what he means. The key is in the first two sentences.
– He has his house covered in solar panels.
– He is therefore ‘pretty green’. (this means that he is in tune with current social beliefs)
Whether or not this solves any problem is moot. That is not the point. The point is that he is fashionable, and the ‘right’ sort of person for Salon magazine readers to be reading. What he has said is much the same as if a celebrity in the 1970s had said: “I have lots of flared trousers…”.
A projected twelve years to pay off my solar panels here on the coast of southern Victoria – with no subsidies and minimal feed-in tariff. Just helps me avoid the worst of our Greens-inflated electricity prices. One thing I forgot to factor into the original budget was the annual cost of cleaning the panels, which is necessary given our salt-laden coastal air.
In Europe there area a number of Vermont-countries. Sweden is one through 45% nuclear and 45 % hydro. (he remaining 10 % are mainly the “by-product” of central heating). Never the less we are busy trying getting more green. And the bureaucracy in Brussels have ambitions for the European average and we are thus supposed to contribute to the greening of Europe by producing electricity that we do not need. Yakymyshyn’s calculations would do a lot of good by us, but will have probelms in reaching MSM ( as in the US).
You missed a major point: The storage of energy for nights and cloudy days. While using power from the utilities to replace lost solar power can be used, and as mentioned in one response, requires large backup capacity, at some level, stop/start of the utility backup system is impractical (once solar use becomes a major part of the daytime source for the system). In that case, local storage is the only solution, and the only current practical storage is battery. However, battery use is far more expensive than the solar cells. A combined battery and solar system fully independent from external hookup would be 3 to 10 time the cost of utility power, with no ability to ever recover the cost just due to battery cost alone (even if the solar cells were fully refunded).
Since they are closing Vermont Yankee, due to the competition from cheap natural gas. Yes, a nuclear plant is being shut down because of the advent of fracking, this calculation will no doubt change to a still risible, but marginally less ridiculous level.
What about inflation? Over 20 yrs inflation at 3% p.a. a £10000 installation will cost £18000 to replace.
Your biggest error was in assuming that a 10 KW solar panel array actually can produce 10 KW at standard maximum solar irradiance levels (one sun). It cannot. There are two ratings for a solar panel – what can be achieved in the lab, without the effects of heat on output and other environmental problems (advertised rating) and the real output ratings , which tend to be roughly 12% less. There are also inversion losses of 3 to 5 percent in converting to AC. There are also losses due to panels not being free from dirt or film. There are also losses due to non-optimal orientation of the panels (not pointed optimally).
Deterioration due to aging is also an issue. There is also the fact that producing a solar panel
requires lots of emissions. Some studies claim the carbon footprint of a solar panel power is roughly half that of natural gas. Also, natural gas emissions depend a lot on whether the generating plant is a closed cycle or open cycle type. Open cycle plants produce the most carbon emissions.
One cost factor for solar panels is NEVER mentioned. The cost when the roof needs to be reshingled and the solar panel array has to be removed and then reinstalled. Installation costs are the greatest cost component of a solar array
http://mobile.businessweek.com/news/2013-10-19/china-solar-energy-says-directors-detained-amid-fraud-probe
I would never, ever install solar in order to reduce my co2 output. We need to get atmospheric co2 up to at least 600ppm before I’m happy. I have my reasons.
I would install it (if I could afford it) for heating hot water / to perhaps save money / if I’m in a remote area and off the grid. Some people install solar in combination with a generator or small wind turbine for the last reason alone, which makes sense. Saving money or co2 reduction is not a bigger concern to them than simply getting power. I’ve seen a few in operation.
Some small wind turbines can be used to draw up water from a borehole if you are off-grid.
Another solar panel cost factor never mentioned is the fact that uncontrollable power inputs to the grid generate additional costs in order to accept that power. Backup generating capacity must be available , and regardess of whether the backup capacity has to actually generate power, it costs quite a lot to have it available. The only costs saved by using solar rather than conventional
generation is the cost of fuel avoided. But that cost is generally the smallest component of a power plant’s operating costs. In the case of nuclear, it is generally less than 10%. When a utility buys solar or wind power, it is paying for that power more or less twice – payment to the solar/wind provider, and payment to the generating plant for being available. CAlifornia is building pumped storage facitlities to store wind/solar energy when not needed, but this in no way allows for the closure of backup pants, since wind/solar power can be absent for far longer than the pumped storage capacity can handle. IT mostly acts to allow for wind/solar power to be used a dozen or so hours after it is generated, thus avoiding dumping that power into the ground. There are significant storage losses with this (expensive) scheme – pumped storage typically loses over 25% of power sent for storage.
Interested in why PEI has no fossil fuel use. Turns out it does have a lot of wind power but most of the islands power is imported from NB. http://www.gov.pe.ca/energy/js/chart.php list the power. This AM 24 MW wind power out of 170 MW total. Total wind generation is 176 MW, so they could be totally wind powered this morning with ideal winds.
I dont see the actual production of power from these panels compared to CO2, just the costs of installation. PV panels dont produce anything for 4-5 months in winter above the Arctic Circle.
Does this cost of installation also include the batteries, including their replacements every 6-8 years?
fred4d: “This AM 24 MW wind power out of 170 MW total. Total wind generation is 176 MW, so they could be totally wind powered this morning with ideal winds.” I guess for the few minutes its up there, but watch the output every hour, it’s all over the place. Jumps as much as 50% one hour to the next. That’s the problem with wind, they have no idea what wind will produce next hour, unlike every other normal power production.
Yeah, but – for some, at least – you can’t put a price on Feel-Good (TM).
Smug, self-important b@s*@**s.
“CAlifornia is building pumped storage facitlities to store wind/solar energy when not needed, but this in no way allows for the closure of backup pants, since wind/solar power can be absent for far longer than the pumped storage capacity can handle. ”
Germany already has that, but they are shutting them down because the costs are prohibitive.
http://www.spiegel.de/international/germany/high-costs-and-errors-of-german-transition-to-renewable-energy-a-920288-2.html
The BEST way to lower CO2 emissions in colder places with clean electricity (Canada, certain north east states), is to go geothermal. If you are unplugging a propane heating system to switch to geothermal you even save money, without any incentives.
If incentives were applied at the same level as solar, everyone would be on geothermal, as its useful for cooling too.
The rising cost of electricity in order to pay for ‘green’ energy like wind and solar is putting a damper on this actually useful technology.
Chad Wozniak says: “You are forced to chose between accepting that or having a humongously larger electric bill ”
But your FIT participation is what is causing those high power rates. YOU are the cause that everyone else is forced to pay.
Tom Andersen says:
The BEST way to lower CO2 emissions in colder places with clean electricity (Canada, certain north east states), is to go geothermal.
——-
Several years ago, when NG was 14c, I spent the money and switched my home to geothermal. Then fracking collapsed the price of NG, and the FIT program doubled my power rates. Geothermal is no longer cost effective.
I thought the sole purpose of installing solar panels on the roof was to make money, by taking it out of the bank accounts of all those electricity users who do not have solar panels, and transferring it into the bank accounts of those that do, via feed-in tariffs. At least that is the case in Europe.
A practice for which the term ‘daylight robbery’ is most apt: having solar panels on the roof is just a means to avoid it falling under the definition of theft according to Common Law.
JK says:
October 21, 2013 at 7:25 pm
The way I would approach the calculation is to first find out how much (unsubsidized) solar costs above and beyond what you pay at present.
The problem with this approach is that you are going to pay for peak generating capacity no matter what unless the solar has 100% correlation with peak load.
In the US coal/natural gas fuel cost is at most 5 cents/KWh and in some locality’s as little as 2 cents/KWh.
The value of any ‘intermittent’ energy source therefore should only be compared to the cost of the saved fuel unless their is very strong correlation with seasonal peak load.
Therefore, anywhere that is a ‘winter peak’ load state…the value of solar can never exceed the cost of power plant fuel.
Even in summer peak load states with solar having a high correlation with peak load the amount of solar value is limited to the difference between summer and winter peak loads.
IIRC The difference between summer and winter peak loads in Texas is about 1 or 2 GW out of 60 something GW total.
Chris, It’s an interesting study, but I see some substantial problems with the methodology.
1. With fossil fuel electricity, it’s almost always cheaper to move the electricity from a generating plant near the fuel than it is to move the fuel to a plant near the customers. Thus for small jurisdictions with no native fossil fuel resources like Vermont and PEI, the numbers probably aren’t very accurate. Averaging over all of the US and Canada might give a better picture.
2. Solar is assumed to be photovoltaic. Even in Vermont solar hot water and, in some applications, space heating would seem to have substantially better cost factors than solar PV. If McKibben can add, or has a friend who can handle basic arithmetic, it’s likely that some or all of his solar panels are for heating rather than PV.
3. As it happens, Vermont is a rather cloudy place — especially in Winter — with a significant amount of snow (80 inches a year in Burlington). In fact it wouldn’t be shocking for the output of any solar collector in December in Vermont to be pretty much zero. It isn’t clear that a solar availability correction is included in your computations.
You make a couple of simplifying assumptions that hurt your conclusions. First, that electricity is not traded between states and provinces. Vermont gets half its power from out of state.
Second, that solar would displace all other generation capacity equally. A smart utility would instead shut off the generator that costs the most to run, which will be first coal, then gas.
We’re getting exactly what one would expect if you leave the design of the electrical grid to the politicians. And now they’ve moved on to Obamacare….