How current GHG policy distorts capital allocation

As we think about how to price GHG emissions, it’s often (and accurately) cited that having a meaningful conversation about GHG pricing first requires that we remove all the existing subsidies so that we can stop irrationally allocating capital. Clearly, we can’t provide insurance liability waivers to nuclear and ratepayer guarantees to regulated utilities and then conclude that Monty Burns’ access to capital represents the action of unfettered markets.

Let’s ask the obvious question: how much do various technologies cost us, per ton of CO2 reduced, on an all-else equal basis? If we are already rationally allocating capital between our alternatives, then the differential addition of an actual price on CO2 ought to only help the good technologies proceed a bit further over the line. As you might imagine, we aren’t. And as I alluded to here, the environmental community deserves some of the blame.

Here’s my math. I’m a heat and power guy, so I did my math on those sectors of the economy. I’d welcome anyone with expertise in transportation and other sectors to number-crunch accordingly.

First, the baseline. Average retail power prices in the U.S. today are about $90/MWh. Average U.S. CO2 emissions per unit of electricity consumed are about 0.6 metric tons/MWh. Both values are on a delivered basis so that we can accurately compare central and local generation technologies.

Let’s say we had a technology that could produce power with 0.3 metric tons/MWh of delivered power and that the resulting power would cost $120/MWh. That gives us a $30/MWh “cost” and a 0.3 metric ton/MWh benefit, for a net cost per ton of benefit of $30/0.3, or $100/ton. How do the usual suspects stack up?

Assumptions

If I knew how to embed spreadsheets in posts, I would. But I don’t. (If anyone wants the spreadsheet, email me and I’ll send it to you.) Suffice to say though that I had to make some simplifying assumptions.

Most importantly, any level playing field has to assume that all market participants have equal access to capital. This of course isn’t true, but energy markets are highly regulated; to the extent one party has access to cheaper capital than the other, it is almost certainly because of a regulation that tilts the existing playing field (see the earlier comments on nuclear). So I assumed that every technology can find investors who are willing to recover their capital with 12 percent interest over 20 years, on an unlevered basis. This is faster capital recovery than regulated utilities get, but lower than private equity demands. Moreover, to the extent that our regulatory environment is allowing anyone to recover cheaper capital (whether through loan guarantees, ratepayer underwriting or some other mechanism), it is implicitly creating a public subsidy. By holding our capital recovery assumptions constant, we can ask ourselves whether the subsidy is worth paying per ton of CO2 reduction, rather than simply accepting the difference as a persistent reality.)

Other assumptions of note:

• Central generation requires $1,400/kW transmission and distribution to connect to load; local generation avoids this capital cost.
• Transmission and distribution, if required, adds 9.5 percent line losses, thus requiring 1.095 MW of installed upstream generation per MW of delivered load.
• Natural gas costs $8/MMBtu, Coal costs $2/MMBtu.
• Plants that recover waste heat displace natural gas-fired boilers

The T&D capital and loss numbers are based on U.S. averages. Natural gas and coal prices are roughly representative of current U.S. markets. The cogen assumption is pretty typical, for the simple reason that no one who is installing a modern boiler can get much besides natural gas permitted. All of these assumptions could be varied of course, but they set the baseline.

Comparative costs per ton

Let’s start with coal with carbon capture and storage. For the sake of argument, we’ll assume the technology works. (No small assumption, but at least consistent with the assumptions made for all other technologies.) I assume $6,500/kW for capital costs, per FutureGen. Add in T&D costs and assume appropriate line losses, and you need to recover $1,158 per kilowatt, per year in order to pay off your capital. Add in fuel and labor costs and you’re up to $1,398 per kW-year. Meanwhile, an average U.S. coal plant today only runs about 72 percent of the time, meaning that this kilowatt of capacity will generate 8,760 x 72 percent = 6,307 kWh per year.

Note to non-power wonks: There are 365 days in a year, 24 hours in a day, or 365 x 24 = 8,760 hours in a year. Thus, one kilowatt can generate a maximum of 8,760 kilowatt-hours in a year.

Back to our math. $1,398 / 6,307 = 22.2 cents/kWh for power from coal with carbon-sequestration, or $222/MWh. Compare this to current retail electric rates of $90/MWh, and this implies that coal plus CCS would raise power costs by $132/MWh. Meanwhile, it would save 0.6 metric tons/MWh, giving us an overall cost of $132/0.6 = $219/ton.

In other words, if we took away every subsidy in the system and compelled coal plus carbon sequestration to compete on a level playing field, it wouldn’t compete until CO2 prices exceeded $200/ton, or about 10 times what they are currently trading for in Europe. It is no exaggeration to conclude that a rational market that fully factored in the cost of CO2 emissions would never invest in coal plus CCS (at least not until it had exhausted lots of other, much more cost-effective options).

So how do other technologies compare?

Nuclear is only slightly better, requiring $157/ton. (The causes are very similar to coal plus CCS: high capital costs plus the costs of transmission and distribution.)

Combined cycle gas turbines can achieve 50 percent fuel-to-electric efficiencies, giving an overall carbon signature of 0.4 tons/MWh. That’s a bit better than the grid, but they use expensive natural gas and still require wires to connect them to the load. In aggregate, they shake out at $287/ton — even worse than coal + CCS.

Now let’s look at traditional renewables.

• Central wind farms (e.g., those that still require T&D) cost $131/ton. Better than coal plus CCS, better than nukes, and better than CCGT — but still above where most are estimating CO2 markets will settle out. (Note to wind wonks: I assume $2,500/kW for the turbine and a 40 percent annual capacity factor. I suspect the latter may be generous.)
• Sustainably harvested (e.g., CO2-neutral) biomass in a central power plant costs $108/ton.
• Solar PV requires a whopping $1,047/ton. (This assumes locally generated solar power, thus avoiding T&D costs; centrally generated solar is much worse.)
• Geothermal is the cream of the crop in the renewable space, requiring just $11/ton.

The lesson of this list? It’s all about capacity factor. Intermittent renewables simply have a hard time recovering capital, because they don’t run very often. (Put another way, a solar plant that runs 20 percent of the time has to cost one-fourth as much as a geothermal plant that runs 80 percent of the time to be competitive per ton of CO2 reduced.) Free fuel certainly helps (and explains why geothermal compares so favorably to biomass, where I assume $3/MMBtu fuel), but ultimately, capacity factor constraints are hard to get around.

Now how about locally-generated sources? All CHP is locally generated, since you can’t move heat very far:

• 60 percent efficient, natural-gas fired cogen costs $60/ton.
• 90 percent efficient, natural-gas fired cogen saves $54/ton. (Or, if you prefer, costs negative $54/ton.)
• Biomass with waste heat recovery saves $135/ton.
• Recycled industrial waste heat without a cogen cycle (e.g., recovered only for power) saves $91/ton.
• Recycled industrial waste heat with a cogen cycle saves $103/ton.

Conclusions

This list is not exhaustive, and I certainly don’t mean to imply that we now know which winners to pick. But in a world that really cares about CO2 reduction and that must operate under fiscal constraints, any decision to preferentially shift resources towards the highest cost approaches is a decision not to maximize the rate of GHG reduction.

And the temptation to pick winners is no less present among those lobbying for coal plus CCS than it is for those in the environmental community who insist upon a 100 percent solar grid. Both paths have the potential to reduce CO2, but neither does it in an economically-responsible way, and therefore neither does it in a way that will maximize the rate of GHG-reduction.

A few final observations:

1. To the extent that our current policy conversation is beginning to take CO2 reduction seriously, it is focused almost exclusively on the most expensive approaches. This is irresponsible.
2. The technologies being deployed in response to current RPS markets are almost universally the most expensive routes to CO2 reduction.
3. We’ve got an awful lot of playing field to level to get this right.