Technology and the Environment DD's Eco Notes

Environmental Leader has a good article on the report by IEA that says that $45T of investment will be required by 2050 in order to reduce CO2 by 50%. In particular, it highlights the graph which is very interesting, showing 15 of the 50% reduction coming from efficiency gains, and the next 20 of 50% coming at relatively low cost from the power sector. So combined, these two are 70% of the overall reduction, and looking at the graph, can be achieved with near net 0 cost (the efficiency savings cover the added cost of power switchover).

So what's left is a massively expensive effort to convert our industry, cars and trucks to an alternate fuel source. The conclusion of the IEA report is that we would need $500/ton tax or price of carbon to achieve the reductions required in the industry and transport sectors. But wait a minute - the power industry needs a price of carbon under $100 to meet its goals, and the efficiency gains can be met without much of a price of carbon at all. At $500/ton, we'd be looking at an additional cost for electricity of around 30 cents per kWH - are we really going to do that just because another sector is harder to deal with than electricity?

What this graph so nicely points out is that we have three different problems, and because of the economics, maybe we shouldn't treat them the same way. First, we have the need to increase our efficiency. Since this is generally cost neutral or better, this should be able to be done through some kind of incentive program. Second, we need to convert our electricity production, and there's increasing consensus that we can do this at a reasonably low cost.

Finally, our big problem is that we need to wean ourselves of our oil dependence for transportation and industrial use. For better or worse, we also appear to have a massive disconnect in supply and demand in this sector, so we are doubly incented to address this problem. And you can see by the large range of predicted cost in the graph, especially compared to the small range in the power sector, that we don't really yet know how to do this.

So I draw optimism from the left 2/3 of the chart, and am very concerned about the right third. But more importantly, I'm increasingly worried about us using carbon pricing as a big hammer that we're going to fix everything with. We have at least three distinct problems here, so maybe we should be looking at at least that many mechanisms to drive change.

The Environmental Leader reports on an analysis from Carnegie Mellon claiming that a $35 price for CO2 would results in a 10% decrease in emissions.

I haven't read the report yet, but just at the face of it I'm skeptical when I look at the impact on the pricing and usage of gasoline and electricity. First, lets do gasoline. Roughing out the math, gas produces just under 20 lbs of CO2e per gallon of gas, or 110 gallons per metric ton. So the $35/ton pricing would add $0.35 per gallon. Given the current gas prices, you'd be looking at an additional increase of less than 10%, and much less than the increase over the last 12 months.

Electricity shows a higher increase, but I'm still not sure its enough to change behavior that much. Using 1.3lbs of CO2 per KWh as a US average, that yields around 1,700 KWh per metric ton of CO2e. So a $35 price for CO2 would add about 2 cents per KWh. Average retail electricity price is around $0.10, so this would be a 20% increase in retail electricity prices, and higher for commercial electricity which tends to be lower. This is enough of an increase to cause some changes in behavior, but 10% seems like a stretch. For electricity, however, it would help with the economic case for emerging green alternatives, so would probably do some good on the generation mix side over time.

So when I read the report I'll find out how they get there, but I'm very doubtful for gasoline and transportation, and mildly skeptical for electricity usage at $35/ton. This analysis definitely makes one thing clear: at CO2 prices below $5/ton, there will be no change in behavior.