Why is Wind Power Even Considered?

There is all manner of interesting information available online these days. Transpower who are responsible for the NZ national power grid has a most informative website.

On their site was a paper about forecasting wind generation and, after reading the article, the words from a 70s song sprang to mind. “You don’t know which way the wind blows, so how can you plan tomorrow?”

Having the best possible forecast of New Zealand’s likely electricity demand over the coming hours and days is very important, both to maintain cost-effective energy security, and to provide accurate price signals to the market.

Wind generation is one component of load estimates which can impact forecast inaccuracy. Wind is effectively treated as negative load in the market and is currently required to offer at or below $0.01 /MWh. This means it will nearly always be dispatched, and remaining NZ generation must meet the balance between load and the generation from wind.

I have no idea how the intracies of the wholesale electricity market work but selling at or below $0.01/MWh does not seem like a great return.

When forecasting generation two hours in advance of any point in time, ‘persistence’ forecasting is used for wind offers. In other words, the wind generation from a site is assumed to continue at its current measured level for the next 2 hours. Beyond this timeframe, generator offers are used to provide a forecast.

Conforming and non-conforming load are the other two contributors to estimates of load. Conforming load follows predictable levels and daily patterns, and includes residential, commercial and light industrial load. It is forecast centrally by the system operator, using a load forecasting tool which has weather and historical load as inputs.

In other words, the grid operators expect a peak in the evening when dinner is being cooked, and this peak is pretty consistent day to day and varies with the season.

Non-conforming load is demand from large, mainly industrial,users at 12 sites around the country and is estimated using bids. As shown in Fig 1, wind contributes around quarter of the total forecast inaccuracy, despite only being responsible for around 4% of the load balance. Improvements in wind forecasting accuracy could potentially have a significant impact on overall load estimation.[…]


Here wind generation is considered a “load” and a very unpredictable one at that.

The website also provides a weekly market report (only the last 5 are available). What can we find in there?

Report for 7 July:

Wind generation increased to 5% of total generation […]

The previous three reports did not mention wind and this one said “increased to” so clearly wind generation was less than 5%.

Report for 14 July:

Wind generation increased to 6% of total generation this week […]

There is also a ‘live data’ page.

So glad that Dr Woods is building us 4.5 windfarms per year. The future for wind just blows me away.

How long is this stupid charade going to play out?