Prof. Severin Borenstein from Berkeley University makes a very readable micro-economic case against peak-time rebates (PTR). When PTR are the method to incentivise demand to become more efficient w.r.t. to the supply and the grid in peak times, consumers are paid to use less then some pre-decided amount on energy-intensive days. I share many of these concerns and I think some of them also work as a good foundation for an argument for explicit valuations of flexibility in energy consumption (which has become a theme to much of the work I did over the last 4.5 years).
Prof. Borenstein's arguments, in short, are as follows:
- Baseline consumption is set against past behaviour, which sets wrong incentives. "When my baseline for peak-time reduction is based on consumption during other high-demand days it undermines my incentive to conserve on those other days." It even lowers incentives to buy devices that use less energy in general.
- It is a very marginal payment. Just above the baseline, the incentive to use 1 kWh less is given, then just below the baseline it is gone. Only those consumers a little above the baseline will be interested, others will be too far above it to care. Not what you want as a mechanism designer.
- You'll also reward "free-riders", people who accidentally use less than the baseline. Wasted money.
- Consumers with stable consumption are hurt more than consumers with random and erratic behaviour, because you are not punished for going over, only reward for going under the baseline. He has a nice small example for this.
Borenstein, as well as the commentators on this article, agree that real-time pricing (RTP), which I mostly refer to as dynamic pricing in my work, is a far better option from an economist's point of view, but that domestic electricity consumers will probably not willingly accept it (even if RTP saves cpnsumers money overall, compared to PTR), as it is quite complex to use and there is a danger of suffering unforeseen price peaks.
Besides providing good arguments for RTP, this article also provides some support to the notion I use to define how we can/should identify tradable flexibility in energy systems. Especially the lack of a natural baseline against which to measure the contribution (of using less energy at a specific time) resonates with me. I argue that without a reference point, the economic valuation of flexibility becomes very questionable.
To trade flexibility explicitly, both parties should first agree on a reference point and possible deviations. Then, one deviation can be chosen. One could put a value on offering all the possible deviations or only put a value on the one realised deviation. I'd like to see market/pricing mechanisms that do it in this explicit way, which is mostly done in some sort of two-settlement procedure (one example is ABEM).
RTP, on the other hand, can bring flexibility into the trading of energy, but more implicitly. I cannot provide a full-fledged economist argument here, but the sketch would probably go like this: Let's assume that we can separate your normal behaviour (mostly independent from prices) from your possible flexible behaviour. The equilibrium in the RTP market which would be realised without your flexible persona implicitly serves as the reference point for the flexibility you can offer. The residual demand or supply curve for your flexible persona starts from there, and when you can buy or offer electricity on that curve, your flexibility is useful to the market, and in turn to you. In this way, market prices serve as a signal that brings out the necessary flexibility.
However, we know that equilibria are only a helping hand in imagining economic settings, and are almost never observed in the real world. Thus, this idea of a reference point is very implicit indeed. I believe that mechanisms, which can make the reference point for flexibility explicit, are actually easier to use in the end.