Sunday, October 4, 2009

Why nodal pricing matters for generation and transmission?

This is another follow up post to Brian Fallow's articale in the New Zealand Herald on asset swaps. One of the things he suggests in the article is that nodal pricing doesn't matter for generation. This is not true. Arguments about the relevance of nodal pricing to the demand side are quite valid.

Nodal pricing in fact is highly relevant to efficient generation and transmission outcomes both in operational terms and investment efficiency. The locational issue relevant to generation is the fundamental question, (at its simplest) are we better to have a local generator only or a remote generator with transmission? If you don't have a locational signal in your pricing then you run the risk of using a remote generator every time and having an increasingly expensive transmission network. This is true of operating the electricity system as well. Sending power through transmission lines heats those lines up, which represents a loss of energy to the system. Without a nodal pricing system for dispatch then you could easily use a far away generator to meet demand even though it uses proportionally more energy (through losses) than the difference in cost of the local generator. Choosing which plant should run to meet the energy needs at least cost overall is a process called economic dispatch.

Brian Fallow makes a reasonable point in suggesting that nodal prices are meaningless because renewables are the first cab off the rank for investment and they have to be built where there is a resource. This oversimplifies the issue, however, because there are lots of renewable projects and they are distributed all over the country. The question isn't so much where should a project be built (it must be built where fuel is available - this is even true of thermals) but which projects should be built. Given that the main logistic challenge of the power system is getting power to Auckland from all over the country (but with significant other loads at very different places - such as 12% of the country's demand at the very bottom of the South Island) then deciding the least cost way of moving this energy around is a complex problem. The SPD model is the piece of technical magic that performs this task. Then, managing to signal to potential investors where they should invest (nodal pricing) is a neat trick that falls out of the New Zealand's electricity market pricing model (SPD - which is in fact an economic dispatch model).

Even under the heading of renewables you can choose hydro in the South Island (and other places), geothermal in the central North Island or wind (predominantly around the lower North and upper South Island). Each choice has a markedly different impact on how much the grid will cost to dispatch and how much transmission capacity will need to be built; and this can be very different at different times depending on weather (hydro and wind), outages and fossil fuel stockpiles. The pricing model for generation and transmission is, necessarily, complex.

Most of the solutions proposed to increase simplicity tend to be quite static. This is to say that they might work well if the electricity system didn't change much. The problems with electricity supply, however, are either very short term (I have to meet demand now) or very long term (which 40 year investment decision is best). Neither of these suits static solutions. The electricity system is very dynamic and dynamic means complex.

This does contribute complexity to the buying side of electricity and it is here that pure nodal pricing and economic dispatch has far less relevance. Here some form of simplification might make sense.

Mr Fallow has made some sensible observations but he does need to be aware of one of Albert Einstein's most useful quotes: "Everything should be made as simple as possible, but not simpler."