Friday, August 6, 2010

Carbon Neutrality with PVs

Photo credit: Dow Chemical

Yesterday, I explored the carbon captured by a straw bale building as a factor in getting to a carbon neutral/negative household. Today, I want to look at how far building-integrated photovoltaics can get us.  Let's start with a breakdown of typical residential energy demand.  I found figures from the EIA Residential Energy Consumption Survey (Table 14, 2005) as follows, based on region.

To this, I would like to add vehicle usage and personal air travel.  Slightly out of date figures from the 2001 National Household Travel Survey show household gasoline usage at 1053 gallons/household/year, and at 115,000 btu/gallon, that's 121 million btu/hh/yr.

I couldn't quickly find good household airline usage statistics, but overall, airlines used 8.5% of transportation energy versus 76% being used on highways.  So we can guess that household airline usage is about one tenth of household gasoline usage, overall.  That gives this:

These numbers are a bit rough and ready, but will get us in the ballpark.  So, the biggest chunks, in order, are household driving/gasoline usage, then heating (except in the South), then hot water, and then it gets into the smaller pieces.  Over time, climate change is going to tend to drive down the heating piece, and increase the air conditioning piece.

Before we think about the conservation options here, let's just briefly calculate how much PV is required to generate this much energy.  For that, we need a map of the solar resource, like this one:

For my purposes, I'm going to take the NorthEast as having 4.5 kWh/m2/day, the Midwest as 5.0, the South as 5.25, and the West as 5.75 (with the understanding that these are rough regional averages - there are places in Arizona where you can get up to 7.0).  We have to multiply this by a solar panel efficiency.  For my purposes, I'm going to use the Dow integrated roof shingle pictured at the start of the piece, which uses a CIGS cell with a 13% efficiency.  Doing the math, to get from kWh to btu, days, to years, etc, we get the blue bars in the following graph for the PV requirement:

(Note that there are a variety of issues with this - in particular, you can't power a gasoline car with PV in any sensible way - so this is just a rough indication).

The y-axis is in square meters.  You can get to square feet by multiplying by 10 (pretty close, anyway).  Clearly, these totals are a bit of a stretch to fit onto the roof of a residence.  For example, our house from yesterday had a 48' x 32' exterior footprint.  Positing a 45 degree south-facing shed roof (not the most practical or appealing roof design), with a 2' overhang all round and covered in PV, we get 52 x 36 x 1.414 = 2600 square feet - not quite enough in the north-east.

The red bars show what happens if you make the following assumptions:
  • Super-insulation saves you a factor 4 on the heating and cooling energy usage
  • Use of ground-source heat pumps saves you another factor 4 on the heating usage.
  • Preheating of water via the ground source, or solar hot water, saves you a factor two on the hot water energy usage.
  • Use of electric cars saves you 3X on the driving energy usage (roughly the ratio of electric motor efficiency to internal combustion engine efficiency)
  • No other changes.
With those assumptions, I think you can get in the right ballpark.  Now you need 1000 - 1300 square feet  of rooftop PV, depending on region, which seems like something one could potentially do (at a price in $$, of course).

Finally, this is an operating energy calculation.  Embodied energy will have to wait for another time.


Anonymous said...

I checked this out and found a web site where I live that estimated one would need 24m² (maybe 250 square foot) for a low energy house. I will have about 50 m² so that is maybe enough depending on which figures are are right. In the end I would have ot read a lot on that and get professional advice but it sounds workable at any rate- depending of ocurse on the costs and if tit takes a couple of years on increasing efficiency.

Stuart Staniford said...

Do you have the link? One thing that is quite likely is they are not including vehicle energy use, which I'm choosing to include in my boundary. Of course, one could also be a lot more aggressive about appliance energy use (which I didn't assume any reduction in above, mostly for the sake of understanding the ballpark quickly).

chapter1 said...

For the Northeast in particular, where much of the energy goes to space heating, I wonder if it would be appropriate to use an efficiency greater than 13%. The heating efficiency of direct sunlight has got to be far higher, even if an active, energy driven system is used to capture it. Seasonal factors will hurt efficiency (most sunlight arrives when heat isn't needed), but 13% still seems awfully low.

On another note, I don't think there's anything magic about roof square footage. Other sources of space are available (e.g., solar panels built into or over driveway or nearby slice of road).

Stuart Staniford said...


Fair points both. Any actual design would very likely employ a mix of technologies in a more complex way. I'm just trying to understand the magnitude of the main issues.

Anonymous said...

2600SF is a LOT of PV - about ten times the size of a typical 3-4KW household installation - and a massive expense that probably exceeds the cost of home/land.

You'd be back-feeding the gird whenever the sun is up, essentially acting as peak power generator. You'd likely be into an whole different classification regarding state/fed incentives, and how you'd interface with your local utility, permitting, pricing, etc...

I understand it's just a ball-park sizing exercise, but it's pretty far out!

Stuart Staniford said...


With care on everything else, I'm more thinking we could get it down to something like 1000 sq ft. Obviously it's going to cost plenty, but then we'd never have to pay for any fossil fuels ever again (and who knows how high they might spike at points in the future). I'll have to do more work on the economics at some point in the future, but I doubt it's a loser over the long term.

Nick G said...


An EV is about 6x as efficient as the average US vehicle (which gets 22MPG): a gallon has 35 kWh of energy, which will propel an EV 140 miles.

That includes charging losses: the Volt uses .2 kWh/mile from the battery, and .25kWh/mile from the wall. Of course some, like the Aptera, use only .07kWh from the battery.

porsena said...

The projected cost seems to be $10/sq ft with availability coming next year.

Stuart Staniford said...


Your link is broken, but this story provides the same figure for the after subsidy price.

Mr. Sunshine said...

The "Dow Chemical" photo is amusing, since the ABS plastic thin-film solar shingle obviously forms the waterproof membrane on that roof, but the job also obviously lacks the Code-required fireproof barrier beneath the solar component... so, there's a "Red Tag," not legal anywhere in the USA job that will have to be torn down and re-installed entirely....

Mike Aucott said...

Nick G.,

Your assertion that EVs are six times as efficient as gasoline cars is not correct if you count the entire process, which has to include the efficiency of electricity generation. In the U.S. today, that's about 35%, which bumps the factor of six down to about a factor of two. If you correct your calculation further to represent the mpg from a gasoline car with a passenger capacity similar to the Volt, which could be close to 40 mpg, the EV is only marginally better. From a carbon emissions standpoint, full-battery EVs and plug-in hybrids are actually worse in some parts of the country than gasoline-only hybrids, because much of the electric generation in some regions comes from coal combustion. See Moyer, M., 2010, The dirty truth about plug-in hybrids, Scientific American, 303, 54-55.

assystems said...

From a climate change perspective, I always like to compare solar PV (essentially a black body) to what would be achieved by simply painting the roof surface white.

Keeping in mind that the first-order effect of installing a black body is to absorb more heat, and all the knock-on effects of manufacture, installation, and maintenance of these complex systems made of exotic materials, I remain totally unconvinced by solar PV.

Just to be clear, I've lived with PV for 30 years in several countries, and currently have both a solar PV system and a solar hot water system. I consider them a luxury in my off-grid home, not an eco-friendly choice.

Joel F said...


Why did you choose a relatively awful CIGS product as your PV example? Why not choose something more like industry average? Or, for a little more work, the range of commercially available product? The most efficient PV products are almost twice as efficient as your example.

@Mike Aucott, who said:
"[Nick G's]assertion that EVs are six times as efficient as gasoline cars is not correct if you count the entire process, which has to include the efficiency of electricity generation."

Not if you are comparing the heat energy of gasoline to PV, in which case there is no further efficiency reduction beyond what Stuart has already included. If Stuart's numbers here are looking at the heat energy of gasoline instead of the work energy, then indeed an electric car may need up to 6 times less heat energy equivalent from PV than from gasoline. Stuart, don't you think you ought to at least mention this?

@assystems, who said:
"From a climate change perspective, I always like to compare solar PV (essentially a black body) to what would be achieved by simply painting the roof surface white."

Have you really done the comparison? This is one of those concerns that sounds intuitively important but turns out to be trivial. The climate effects of albedo are insignificant compared to CO2. See here:

Joel F said...


Sorry, I should read before I speak. I see that you did indeed mention a reduced energy need for electric cars. Care to weigh in on why 3x instead of 6x though?

Stuart Staniford said...

On the CIGS issue - I was specifically looking at building integrated PV (I have to sell my wife on this on aesthetic grounds ;-), and the building integrated stuff I've seen so far for residential is all thin-film.

Stuart Staniford said...

"Stuart, don't you think you ought to at least mention this".

That was the "Note that there are a variety of issues with this - in particular, you can't power a gasoline car with PV in any sensible way - so this is just a rough indication" caveat.

3X came from comparing the 88% efficiency that Tesla quotes with typical auto efficiency of 20-30%. You're probably right that the comparison is a bit generous to the ICE vehicle.

Mike Aucott said...

Joel F. and Stuart,

Certainly electric cars need less heat energy to turn their wheels, and the 3X or better range seems realistic for this nominal efficiency when compared to ICEs.

It's the net efficiency that counts, however, and so you have to include the efficiencies of electricity generation and charging/discharging the battery. Electricity gneration efficiency is currently about 35%. Charging/discharging might be in the range of 80%. Multiplying 3 by 0.35 by 0.8 indicates that the net efficiency of electric vehicles is on a par with, and perhaps worse than, conventional hybrids and diesel vehicles.

Stuart Staniford said...


The context here is running an EV off PVs, so the coal power plant efficiency is not relevant. Tesla claims 88% for plug to wheels efficiency. I don't know about the other EVs that are about to come on the market, but I assume they will be similar or a bit less.

Mike Aucott said...

In situations where it's feasible to power EVs from PV, of course you are right that the plug to wheel efficiency is what's relevant. My concern is that powering EVs with PV isn't likely to be cost-effective and so won't happen on a large scale, and they'll end up contributing to continued combustion of coal. I tend to rant and rave on the subject and indeed got beyond the context of the original discussion.

Joel F said...


As someone who installs solar, I currently regard the Dow shingles as a mostly a gimmick, until I'm shown otherwise. (How do you do maintenance on the system? How do you flash around vents and stuff? What if you want to use micro-inverters or similar options? I see a lot of disadvantages compared to traditional modules.)

Also, there are other approaches to integrated PV that don't involve thin film. Particularly well developed are panels that integrate with tile roofs. Try a google images search for "BIPV".

I must sympathize if your wife is more concerned about looks than saving the planet. I truly don't understand the 'aesthetic' objections to solar power. (Asphalt shingles are prettier than solar panels? Wha?) I think that it has almost nothing to do with actual opinions of what beauty is, and everything to do with a fear of standing out and looking different (or, with some chairpersons of the HOAs, the power trip of preventing others from standing out.) I suspect that the aesthetic objection will fade away if solar becomes more common.


If I were you I would temper your concerns about EVs with the fact that there is pretty much no where in the country that driving one of the new EVs will produce more emissions than the 25mpg gasoline car which is currently typical. In fact, for the national average for emissions (1.31 lbs CO2 per kWh), even a 40mpg car is still significantly worse (~30%) than what the Leaf or Volt is supposed to get. If you live anywhere were emissions are much lower than the national average, switching to an EV has even greater benefit. There is a difference between saying "The best thing for the planet is to drive less" (absolutely true) and saying "EVs would be just as big a problem as gasoline cars." (They wouldn't be.) See here for a comment I made on this subject at The Oil Drum.

Mike Aucott said...

Your numbers look correct to me Joel. A future for EVs may exist (but such a future almost certainly includes a lot of coal combustion). In the meantime, the U.S. could produce little diesel cars that get 80 mpg like they do elsewhere in the world, and avoid the investment in infrastructure needed for EVs. And we could get serious about powering vehicles with natural gas.

I totally agree that the best approach is to minimize driving. It's incredible that so many of us still have to commute daily to a job that can mostly be done from a computer.

Going Green said...

Stuart, if you are thinking solar, don't throw away the bulk of it which comes in as heat. You can boost your efficiency from the 15-20% level up to 60-70%. While not readily available today, the round numbers can be improved dramatically if you use the heat. With solar thermal already you can get 70% of your water heated for you (I don't have a good number for the space heating).

Once you turn down this path, you start seeing other potentialities such as absorption chilling for air conditioning (search yazaki 5 ton chillers).

While the chiller is not economical perhaps, the CHP solar payback will certainly beat a pathetic 20% solar PV payback.