• Home
• Energy & Resources
• Electricity
• Archive
• Reports
• Dry Year Security Reports
• 2004/2005-2007/2008
• 3. Electricity Commission

• Electricity

• Archive

• Reports

• Dry Year Security Reports

Since the first publication of the New Zealand Electricity Outlook: Dry Year Risk, much has changed in respect of the electricity industry. Since December 2002:

1. the Maui Gas Redetermination concluded that Maui's economically recoverable reserves available under their contract were considerably less than previously thought and this led to a reduction in the amount available for generation from early this year;
2. the nation has faced another electricity "crisis" on the heels of winter 2001, exacerbated by the reduction in Maui's ability to support essentially unlimited off-take;
3. the electricity industry made arrangements for alternative fuel supplies for plant previously burning predominantly gas, notably Huntly (coal) and New Plymouth (oil);
4. the industry's efforts to meet the government's goal of self-governance ended in failure;
5. the government announced the formation of the Electricity Commission (EC), taking over industry governance from November 2003;
6. the government also announced that the EC would assume responsibility for ensuring the industry achieves 1-in-60 dry year security without the need for calls for widespread voluntary savings;
7. the government announced a scheme for the EC to procure and manage reserve generation capacity to be used only infrequently in extreme dry years;
8. Transpower has introduced a new scheme for under-frequency load shedding, and has developed changes to the way the HVDC link is modelled in its SPD dispatch model, that will allow the HVDC link to sustain transfers of 626MW⁷ southward in a dry year, such changes becoming operational from 12 November 2003.
9. Transpower increased the night load limits on the BPE-HAY line from 800MW south to 930MW south.⁸ This will enable more electricity to flow southward on the HVDC Link.

These are the most dramatic changes seen in the industry since 1998/1999 when ECNZ was split into three SOEs and government mandated the ownership separation of the lines and energy businesses. They have lead to a number of changes in our assumptions for dry year security modelling, in particular:

• monitoring and achieving dry year security of supply will become a more carefully planned and co-ordinated effort lead by the EC, but note that we do not assume that the EC will intervene in a heavy handed or anticompetitive fashion in the electricity industry;⁹
• information concerning security of supply will become more widely and readily available, including, but not limited to, projected fuel stocks, plant availability, and plant maintenance schedules;
• the EC will co-ordinate dry year security of supply to a standard of 1-in-60 without planning for any voluntary savings other than those that can be expected from the normal response of consumers and producers that are exposed to the electricity spot price;
• all major thermal plant that is serviceable will be available to run to capacity in an extreme dry year except for planned outages. Unplanned outages have not been modelled;
• the recently announced 155MW diesel-fired plant at Whirinaki will be available for reserve duty from 1 June 2004 - the role of reserve generation is discussed in section 3.1;
• the EC will monitor and co-ordinate hydro storage so that the SI heads into winter with at least some minimum volume of water, set at 2,100GWh for this security update - this is discussed in section 3.2;

3.1 The Role of Dry Year Reserve Generation

Dry year reserve generation, as a ring-fenced service, is a new development in the electricity market. In our modelling we assumed that the 155MW at Whirinaki would be offered at a price of $250/MWh,¹⁰ chosen because this is approximately equal to the highest monthly average price seen in the market to date. This value is therefore just at the upper limit of what the market might be expected to deliver on a sustained basis and therefore this price should not overly impact on the dispatch of other plant, so that reserve generation becomes the "plant of last resort."

Regardless of the mechanism for deciding when to run reserve generation, it will not affect the results of our modelling since the offer price is selected only to control the timing of dispatch of the Whirinaki reserve generation.

While all generation plays a role in securing supply, dry year reserve generation is a new service which appears to have been developed because we have had two electricity "crises" in close succession. The electricity spot market never had an explicit expectation that security of supply would be co-ordinated or delivered by the market, other than by the actions and interactions of producers and consumers.

Under the EC, the government requires a specific dry year security standard to be met, the so-called 1-in-60 criteria, without the need for calls for widespread voluntary savings. This criteria is considered further in section 6.

All of our modelling shows that the supply-demand balance is relatively tight at present, at least compared to the last few years since TCC and Otahuhu B plants were commissioned, so the addition of capacity, be it reserve or not, will help achieve the 1-in-60 standard. The reserve plant is not expected to run often and indeed our modelling shows that this is the case under the modelling assumptions. However, these assumptions cannot hope to take account of all contingencies, for example unexpected outages of hydro or thermal plant or to provide spinning reserves¹¹ when all other plant is running to provide energy, and it is clear that in a tight supply situation the reserve generation could be valuable in backing up other plant as well as being the plant of last resort if things do not go to plan.

As discussed in section 5.6, other types of reserve, such as frequency keeping reserve, spinning reserve generation, tail water depressed reserve and interruptible load reserve are currently required to cover the risk sudden generator outages and/ or one pole of the HVDC Link failing. This reserve requirement has not been explicitly modelled in the scenarios.

Because of this, our modelling may understate the amount that dry year reserve generation would run since we can not hope to model all possible scenarios under which it might be required.

3.2 End of Year Storage

Our modelling was conducted for the year ending March in each year considered and we refer to storage at the end of each run as "end of year storage." Storage at this time of year is important because this is a critical time before the onset of winter and the reduction of inflows that occurs on average in the SI at this time.

In our previous security of supply report in December 2002 we assumed that end of year storage could end up at a range of values from as low as 900GWh in the SI.¹² Under the EC, we assume that the risk of shortage would be too great if this occurred and that the EC will take action to achieve end of year storage that is consistently high.

We looked at historical storage back to 1979 for Lakes Pukaki and Tekapo and also undertook some initial analysis of what could be achieved in the modelling. In the end, we chose a value of 2,100GWh target for end of year SI storage as a conservative figure. This is achieved in the modelling by setting controlled lakes, Pukaki, Tekapo and Hawea, to 67% full and all other lakes to 50% full.

However, there are a small number of inflow years that consistently fall below this level in our modelling, notably 1950, 1955, 1960 and 1977, all of which feature relatively low inflows over spring and summer, the period when SI lakes are usually topped up. Clearly an event consisting of a dry spring-summer followed by a dry winter, could be particularly difficult to manage through.¹³

At a recent workshop at the Electric Power Optimization Centre (EPOC¹⁴) a presentation from NIWA described a phenomenon known as the Interdecadal Pacific Oscillation (IPO) which affects the inflows into the southern hydro lakes. The IPO tends to drive inflows higher or lower for decades at a time. The years 1945 - 1977 are in a drier period of the IPO while 1978 - 1999 are in a wetter period, the difference being of the order of 15% in total inflows. There is a suggestion that the IPO could now be swinging back to a drier period.

The years 1950, 1955, 1960 and 1977 are all from a drier period of the IPO and also out of recent experience apart perhaps from 1992, which was to a degree similar as storage was low at the end of 1991. If the IPO is swinging back to a drier period then we may see more dry spring-summer events and end of year storage management could become even more important than it has been in the last two decades. The modelling shows that it is actually quite difficult to get SI storage back up to 2,100GWh in these years. This is evident in the charts in Figure 8 and Figure 9 in section 7.