Friday, November 6, 2009

Show Me the Money!

Before I get started talking about how much the 55% carbon emissions reductions I reported on here cost, I'd like to spend a little time discussing the conventional thinking about the cost of carbon reduction. Most writing about this on the Internet talks about "payback time", in other words, how long it will take for you to earn back the amount of money you invested in the renewable energy system, efficiency improvement, etc. The idea behind this thinking is that because these improvements result in a net reduction in fossil fuel costs, the purchaser of the carbon reduction product should expect some kind of financial gain on the transaction, at least, as measured over time.

I believe this kind of thinking is faulty. To see the problem with this thinking, suppose there were available a carbon reduction  product that required ongoing periodic fuel charges,  but the fuel didn't generate any fossil carbon. A car that runs on 100% ethanol is an example,  or a solar hydrogen generating system for home heating. The product may or may not be more expensive than a similar product that generates fossil carbon (ethanol generates carbon but it is recycled), and the fuel may or may not be more expensive than a fossil carbon based fuel. Nobody would ask about the "payback time" when buying one of these products. Of course,  people might make purchasing decisions based on the product cost itself and the cost of ongoing fueling. Some folks may even choose to buy the renewable product even if it is more expensive than a fossil fuel based product, or if the fuel is more expensive, just because they want to help contribute to carbon reduction.

Furthermore, consider a consumer trying to decide whether to buy a consumer electronic good or a renewable energy technology device. Nobody asks "what's the payback time on a plasma TV?". Or "what's the payback  time on a Lexus?". I suppose if you rented the TV out to your neighbors or showed movies on Saturday night and charged admission, or used the Lexus as a taxi  you could answer the question, but people don't even expect any payback time for these kinds of goods. There is a net financial loss when you buy them, what you get from these is  the utility derived from using them,  which you otherwise wouldn't have (well, OK,  there are also the intangible benefits of looking like a cool guy who Made It when  you drive the Lexus).

The point I'm trying to make here is that "payback time" isn't a useful primary criterion when talking about the cost of consumer-financed carbon reduction products. Environmentalists and some economists have sold the public on the idea that they'll save money over time if they buy these products as a way to counter the fact that the products themselves are generally more expensive than the equivalent fossil fuel-based products. But while that's a nice extra, the real reason for buying carbon reduction products is that if we don't start urgently deploying them now, the planet is simply going to cook. Weighting a devastated Planet Earth, in which the mid-latitudes have become desert and agriculture no longer works, against the extra cost of the renewable device certainly indicates where my buying decision would come down. Now, this doesn't mean that economic factors shouldn't be taken into account, it just means that the primary criterion for a consumer buying a carbon reduction product should in my mind be: which product can I afford that pulls the most carbon out of my lifestyle the fastest?

So my favorite figure of merit for measuring the economic effectiveness of carbon reduction products is the "one year cost per unit weight carbon eliminated". That is, the cost of the product divided by the unit weight of carbon eliminated as measured year over year from before the device was installed until after. Some folks like to talk about the "lifecycle carbon reduction v.s. cost"  but I feel that the problem is so urgent that products which result in more carbon reduction now are preferred (kind of like the time value of money, a dollar now is worth more than one in the future). The "unit weight" I use is the kilogram (kg), since carbon emissions are typically measured in metric tons (mt) (1000 kg = 1 mt).

The graph above shows the unit cost per kg of one year carbon reduction for the various treatments we've applied to our house and car. The costs for renewable energy technologies (solar PV and solar hot water) reflect rebates and tax reductions. The two hybrid car costs ( Model I and Model II Prius) reflect a hybrid premium (around $3500) above the equivalent fossil fuel powered car (a Corolla in this case). In other words, we could have bought a Corolla, we chose a Prius and payed a little extra. Finally, some of the costs reflect aggregate treatments. For example, in the solar PV case, we not only installed solar PV,  we also removed two huge pool pumps and the pool they went with, plus bought an energy efficient washer and dryer and CFLs, so I can't break out the contributions of each individual product. Similarily with the plug-in hybrid, the cost consists of both the hybrid premium and the cost of buying and installing the aftermarket battery pack.

Notice something peculiar? Most commentators (for example, here) claim that efficiency improvements are far more cost effective than renewable energy technologies in reducing carbon. There is even a well known study from McKinsey and Associates (not free unfortunately) backing up this claim. Yet, my direct experience here is exactly the opposite. The two efficiency treatments (double pane windows and a new refrigerator)  had a per kg carbon reduction cost almost 5x the renewable energy technologies. The most cost effective reduction products were the hybrid cars.

This is quite puzzling. My speculation is that for a reasonably well insulated house, such as ours, with reasonably energy efficient appliances, the reductions obtained from "low hanging fruit" type efficiency improvements are simply not available. Some of the studies on the advantages of energy efficiency assume a house with no insulation at all, leaky hot air ducts, and a refrigerator from the 1950's. Our house isn't in that category.

The other possibility has to do with where the money goes for renewable energy technologies v.s. efficiency improvements. As anybody who has done home remodelling knows,  labor makes up half to three quarters of the cost, depending on the job. The high labor costs in northern California mean that anything involving remodelling is likely to be costly regardless of how cheap the materials are. In contrast, the renewable  energy  technology products generally take very little time to install (it required around 4-5 hours to  install the booster battery in my Model II Prius) or can even be used out of the box. There are exceptions - solar hot water for example requires lots of labor to put in and must be custom designed for the site.

A real eye opener is comparing these costs to the cost of carbon offsets. PG&E offers billing for carbon offsets as part of our monthly energy bill. They charge around $0.012 per kg carbon (based on the cost of offsets for fossil gas), almost three orders of magnitude less than the cost of solar hot water system.

My conclusions from this study are:
  1.  If you can't afford anything else, sign up for carbon offsets with your utility if they have them, or go to a Web site like (a nonprofit) and buy carbon offsets.
  2. You get the most carbon reduction  for your dollar by buying a hybrid car that gets substantial reductions (not a hybrid Lexus) like the Prius
  3. Solar hot water or a pluggable conversion for a Prius come next.
  4. PV and reduction in large electricity hogs like pool motors are next.
  5. Unless you have a really poorly insulated house or are committed to making your lifestyle net zero carbon, efficiency improvements like new double pane windows and insulation are last (of course there are other reasons why you might want to do these improvements).


  1. What an interesting analysis.

    Another factor might be the climate you live in. If you try to keep your house at 68, and the temperature outside is much below zero all day (or over 100, for all that) then insulation is a big factor. If the temperature is 46 at night and 60 during the day, it's not quite as difficult.

  2. You are completely right about "payback time". This is a simplistic measure. It especially has no significance, since we are really using "payback" to decide whether or not to defuse a bomb.

    With regard to the metric "dollars per ton of CO2", we have used this also with regard to our product (whole house fans):