Congestion pricing for electricity networks


Electricity network pricing has been a hot topic in the last six months. And there has been a variety of discussions on demand-side management for electricity. My earlier comments on these are here and here. But there is a disconnect. Network pricing makes up more than half of our electricity bills. And network investment, to cater for peak demand, has been one cause of rising electricity prices. However, there is no serious discussion of congestion pricing for the network.

Congestion pricing for transmission would mean that, if the network becomes congested, the price that users pay for ‘the wires’ goes up. It also goes down (in theory, to zero) when the transmission system is not congested.

Network congestion is likely to arise when there is a strong demand for power. So spot electricity prices (which currently do vary) will be positively correlated with network congestion charges. But currently network charges (generally) do not vary with demand. So introducing real time pricing for electricity networks would create a double incentive to conserve power at peak times. This would lead to more efficient investment in generation, transmission and transmission substitutes.

What is an example of the latter? Well – batteries. Apparently battery research is moving ahead (albeit the good old lead-acid battery still leads the pack). Potentially, batteries provide a substitute to peak usage of power and peak transmission investment. If cheap enough and reliable enough, households or businesses could invest in batteries that charge when the network is not congested and/or the spot electricity price is low. Then, when the network is congested and/or the spot price of electricity is high, the battery substitutes for grid-based power.

Now I am not suggesting that this technology will be the alternative, but the best way to encourage the development of such alternatives is to let the market work. But to do this, electricity network prices have to reflect congestion.

Are network congestion charges possible? In a ‘full nodal’ network every part of the transmission and distribution system would be priced in real time. But this can be pretty hairy stuff (I will let the engineers argue about its feasibility). But even in the absence of ‘full nodal pricing’ we can do better than ‘flat’ pricing for networks. Even rough multi-part tariffs based on congestion in a region are likely to be better than flat tariffs.

Of course, the key for congestion pricing for electricity networks, as with real-time pricing for the electricity itself, is to have the smart meters in place so customers can react to the prices. But, as I discussed before, I think that can be solved by letting the market deal with the smart meter problem as well.

6 Responses to "Congestion pricing for electricity networks"
  1. Network businesses invest to meet consumers potential demand, not their actual demand. The relative subdued maximum demands observed in a number of states last summer reflected cooler than normal weather. Had weather conditions been hotter, maximum demand would have been considerably higher. Network businesses needs to meet what might have occurred, not what did occur.

    True cost-reflective pricing should include a ‘(maximum) demand charge’; this is a charge based on the maximum load a consumer requires from a system. The charge reflects what consumers might wish to consume at most and is typically levied $ per kW or kVA . The fixed charge is paid even if actual demand is zero or does not reach the agreed or nominated maximum demand requirement. If actual demand is higher than the agreed maximum, a considerably penalty would be applied for that billing period and the higher maximum demand requirements would apply for the subsequent billing periods.

    A demand charge provides a similar incentive to consumer as a time-of-use tariff (which is charge on what is actually consumed, not what is potential consumed). It differs from a fixed supply charge which is levied on all customers (of the same class/type) equally. These charges are commonly levied on very large business customers by network businesses. This charge is also paid by retailers on behalf of residential and smaller business customers; the cost is passed on to these smaller customers irrespective of individual customer’s maximum demand. In effective, customers with a low maximum demands (i.e., low number of appliances and no air-conditioner) are subsidising high maximum demand customers (i.e., with many air conditioners).

  2. I think SM has it right. The spot price should not be the consumer price. The expected spot price should be, which would be estimated by some kind of average historical spot price, broken down by TOD and DOY. But it is easy for us to pontificate about this on an economics blog. In the real world, insurance companies still charge double the premium for a driver who is 25 years and 11 months old compared to 26 years. Pricing in Australia is just so far from the theoretical optimum it is scary.

  3. Obviously some of the commentators here are not familiar with what is happening in Victoria, the only state to have mandated the installation of smart metres.

    Time of use tariffs are already being charged in Victoria. The first to see this are those customers who have elected to take advantage of the generous subsidies legislated for those who are prepared to lay out some capital and instal PV systems. This is quite clearly an attempt by the retailers to minimise their own costs that they incur because of this regulated, non-market driven policy. Thus far they are not doing very well.

    Maximum demand tariffs are already being implemented (please note SM). In fact my own domestic contract already includes a MD tariff. At this stage it does not affect me because there is a total demand limit that triggers this tariff and my demand is well below that. But there are some domestic consumers who will get an enormous surprise when they see the effect of their profligacy on their electricity bills.

    My point is that if we leave this alone, remove al the subsidies and the silly sustainable energy targets, the market will eventually sort this out.

    Stephen’s article only disturbs me if he is suggesting a congestion charge that varies in real time. In such circumstances consumers have no ability to react, even with smart meter technology.

  4. There are a variety of ways to improve pricing. At one end is the simple low-shoulder-peak for power and network capacity. Slightly more sophisticated is to make it time and temperature variant. The peaks occur on very hot days and, to a lesser extent, cold days. That is when you want the incentives to curb electricity use (or if you want to have your air-conditioning going full blast, make sure the power prices adjust so you are not being cross subsidised as SM mentions). And there can be even more sophisticated solutions with remote control of appliances.

    A rising block tariff (if you use more than a certain amount of power then the marginal price rises or there is some other penalty) does not necessarily reflect the peak use and congestion unless it takes into account when you use power. It is better than nothing, but it is not clear how much better (e.g. may encourage less airconditioner use on moderate days as people ‘save their allotted power’ for really hot days. This would shift demand the wrong way).

    I think the two key issues are:
    1. even to the degree consumers face (limited) variability in power charges they essentially face flat network charges. But the peak hits power and network capacity.
    2. the degree of potential innovation on smart meters is only limited by the poor incentives in the current one-size fits all model. But that will persist if governments insist that the network companies own the smart meters (even if consumers pay for them, like in Vic) and that there is a single accepted meter that is rolled out to everyone.

%d bloggers like this: