Michael Giberson
Rich Sweeney, at Common Tragedies, raises the question “Is dynamic pricing green?”
Riffing off of Lynne’s article in Smart Grid News and a complementary post here on Knowledge Problem, Rich acknowledges that dynamic pricing for retail power can encourage load shifting away from peaks and may even reduce consumption overall. But, he suggests, even if dynamic pricing reduces consumption a little it may not reduce overall emissions from power generation.
The reason, he says, it that much peak load generation is natural gas fueled, while baseload generation tends to be mostly coal fueled. All the load shifting that dynamic pricing encourages will move consumption away from gas and toward coal. Given that coal generation tends to produce much higher levels of pollutants than gas, load shifting can increase emissions.
As a general matter, I think Sweeney is right, but a more precise answer could be had looking at the question on a region-by-region level.
- Some areas have more baseload hydro, for example, or nukes, in which case the shift from peak to off-peak could reduce emissions even absent any net conservation.
- In some cases — fewer and fewer with oil prices the way they are — peak generators run on fuel oil, and some small generators run on diesel. Coal typically produces higher emissions than fuel-oil powered generators, but the contrast isn’t as great as with gas.
- In the longer run, reducing peak consumption helps delay investment in new generating plants and transmission lines, thereby helping to avoid the environmental costs associated with that investment.
The real lesson here is that there is a difference between economizing on the consumption of electricity and economizing on the use of environmental resources. Putting a real price on retail electricity will bring about more efficient use of electric power, but if you want more economical use of environmental resources — such as, for example, clean air — then we need to put a real price on it, too.
It’s not really as simple as gas-vs-coal. As you rightly point out Michael, there are plenty of places where baseload has sizable hydro and nuclear components. It should also be noted that peaking gas plants are cheap simple-cycle machines that typically run with extremely low efficiency and high levels of emissions-per-kWh. This is pretty different from combined-cycle natural gas plants, which are expensive and efficient enough (high capital cost, low operating cost-per-kWh) that they need much longer run times to be economic, and are thus run as something much closer to baseload.
Furthermore, (as I think Lynne hints at in the Smart Grid article) concepts of resource makeup based on the time of use will shift. In the short (and probably medium) term the grid operation is built on stable baseload generation augmented with peak generation to meet predictable peak demand. In the long term this may well shift toward a grid where the baseload is supplied by a vast array of intermittent generators supported by a set of dispatchable generation and much deeper demand-side responsiveness/storage than we currently see. Properly equipped with technology and information, people’s energy demand has a much greater capability to actually follow intermittent supply.
Furthermore, coal tends to run as baseload not because of some fundamental law of physics but because of the economics – coal plants have high capital cost, low running costs, so once you’ve built your coal plant it makes sense from an accounting point of view to run it flat out until the maintenace costs start killing the profits.
Gas plants’ accountants also want to run their plants as long as possible, but since not every plant can run at a very high load factor they get displaced up the merit order. Old coal plants also get displaced up the merit order to lower load factors as new efficient ones get built.
In the short-run, if dynamic pricing flattens the load duration curve, no new coal or gas plants can be built, so whether the existing gas plants or the existing coal plants are run for longer depends on the relative fuel prices. We could easily see a situation where gas runs more than the old coal plants. To do a proper analysis would require a specific market and information about the plants’ individual start up and run-down times.