Wholesale price/demand sensitivity

  • Last post 25 May 2016
Alice Saville posted this 17 May 2016

Electricity users and generators can make better business decisions if they have access to tools that show, for example, how changing demand levels can impact wholesale prices.

We observed high demand and high wholesale prices on 23 June 2015. This is typical behaviour during winter in New Zealand and I used the tools in EMI to see exactly what the situation looked like.

High demand on 23 June 2015: www.emi.ea.govt.nz/r/0uydw and www.emi.ea.govt.nz/r/zdnjp
High wholesale prices on 23 June 2015: www.emi.ea.govt.nz/r/wwpx5

If I was a large-scale industrial consumer, I might wonder if it was practically reasonable and financially beneficial to decrease my demand during the winter months, in order to avoid high prices.

An industrial consumer could use vSPD-online to plan scheduled outages because they can test their tolerance for changes in price due to a change in demand. The change in demand may be as low as 1% or as high as 500MW. In vSPD-online, they can set the scale or range that’s applicable to their situation, in order to plan ahead.

The industrial consumer may discover that a small change in demand has a significant impact on prices and that it’s worthwhile to make changes in their production schedule. Conversely, a change in demand may have a minimal impact on prices, so they do not need to alter their production schedule.

These are the steps I took to determine demand/price sensitivity on 23 June.


I logged into the vSPD-online dashboard and clicked ‘Create an override’ at the top right.

I named the override ‘Demand increase by 10%’ and gave it a description of ‘Increase demand in both islands by 10%’.

I clicked ‘Create a new override component’.

I chose ‘Demand’ from the type of override, clicked ‘Use trading period/s’ and then clicked ‘All’.

On the next screen, I ticked ‘North Island’ and then ‘scale’

I typed ‘1.1’ into the field, to scale by an increase of 10%.

I repeated this process again so that I could also add ‘South Island’ because I wanted to see the change in both islands.


After I’d finished creating the override I clicked ‘Create a job’ at the top right.

I named the job ‘Price/demand sensitivity’ and gave it a description of ‘Pricing on 23 June 2015 between trading period 15 and 20.’

In the drop-down list, I chose the override that I’d just created.

I set both of the dates to 23 June 2015 and clicked ‘Range’ of trading periods so that I could view activity between TP15 and TP20.


I got an email to confirm that the job was finished. I downloaded the results from the vSPD-online dashboard but there was also a link in the confirmation email.

I analysed the data and discovered the following, as you can see in the attached image below:

- if demand increases by 1%, the price per MWh doubles
- if demand decreases by 1%, the change in price is barely noticeable.

To note: I could have created an override that only applies to a specific day. But by creating my override more generically, ie, an increase in demand of 10% at all locations in both islands, I can use this override in future jobs based on any day.

I was initially motivated by what I observed on 23 June 2015, but the same override or experiment can be applied to any trading day.

Attached files

  • Liked by
  • msouness
baker_tony posted this 25 May 2016

I think this might go hand in hand with EMS's new demand forecasting service for GXP and RCPD (which uses TESLA Forecasting's load forecasts).