Blackout reality check: Coal and extra link may not have changed anything

Source: The Advertiser

2909_sastorm_sp

As reported in the August issue of CEDEX® Electricity Update, SA peak demands this year were 3.0 GW in summer and 2.5 GW in winter. AEMO states that the state currently has 1.70 GW of combined cycle and steam turbine gas generation and 0.92 GW of open cycle gas turbine generation.

In addition, SA is connected to the other mainland NEM states through the AC, i.e. synchronous, Heywood interconnector, the capacity of which has recently been upgraded to 0.65 GW.

Several years ago AEMO established its Future Power System Security program. In its August 2016 program Progress Report, it explains:

“AEMO established the FPSS program to formalise and accelerate the work it has undertaken in the last few years to address operational challenges arising from the changing generation mix. If left unaddressed, these challenges will test the efficiency and adequacy of current operational and market processes. The FPSS program focuses entirely on power system security. It aims to adapt current processes to address immediate risks, while promoting solutions to maintain power system security over the next 10 years. To date, AEMO has not identified any NEM-wide power system security concerns during normal operation.”

The Report goes on to state:

“Initial challenges are more acute in South Australia, due to the combination of its generation mix and risk of separation from the rest of the NEM. The risk of separation has itself not changed, however, the potential consequences have.”

Related to, but distinct from, this program AEMO has published a number of reports under an agreement with the government of South Australia on issues specific to that state, called South Australian Advisory Function reports.

The August 2016 South Australian Electricity Report contains an extensive discussion of the reliability of supply under circumstances where supply through the Heywood interconnector is lost, including, in the Executive Summary, the following statement:

“In the rare event of the unexpected concurrent loss of both Heywood Interconnector lines, there is a high risk of a region-wide blackout in South Australia. South Australia has separated from the rest of the NEM due to such non-credible contingency events four times since 1999. The likelihood that a region-wide blackout would follow a non-credible islanding event has increased as the region has become more reliant on energy imports, and wind and rooftop photovoltaic (PV) generation, to meet demand.”

It is possible that this, or another statement like it, may have provided a pretext for commentators, unaware of AEMO’s work on system security, to voice doubt and speculation on renewable generation.

Loss of the two Heywood Interconnector lines was not what occurred on 28 September. What did occur, as clearly set out in the preliminary report by AEMO, released on the morning of 5 October, was loss, over a period of 40 seconds, of sections of three of the four major transmission lines running between Adelaide and Port Augusta.

Two of the lines were a double circuit, i.e. two lines on the one set of towers, and this set of towers is the one which most of the published photos show. These lines provide the link between the now closed Northern and Playford coal fired power stations, in the north of the state, and the main load centre, in and around Adelaide.

About two thirds of the state’s wind generating capacity is also linked to major load centres through these lines, as is the Murraylink DC Interconnector.

Should a new link to NSW be built, geography means that it would almost certainly be linked to Adelaide through these lines, or a new line following a similar route. The lines which were damaged constitute a large part of the spine of the South Australian transmission system (the other part being the lines to the south east of the state which form the SA section of the Heywood Interconnector).

The National Electricity Rules (Rule 4.2.3) describe an event such as occurred as a non-credible contingency, defined in the following terms

“(e) A non-credible contingency event is a contingency event other than a credible contingency event. Without limitation, examples of non-credible contingency events are likely to include:

(1) three phase electrical faults on the power system; or (2) simultaneous disruptive events such as:

(i) multiple generating unit failures; or
(ii) double circuit transmission line failure (such as may be caused by tower collapse).”

Thus what happened is explicitly anticipated in the Rules as the most severe possible (and by implication very rare) type of event.

AEMO reports that immediately prior to the event, power was being supplied to consumers in the state from the following groups of generators and interconnectors:

wind generators north and west of Adelaide                                                  579 MW 32%

wind generators south of Adelaide                                                                   304 MW 17%

gas generators around and to the south of Adelaide                                    330 MW 18%

Interconnections                                                                                                  613 MW 33%

AEMO states that “pre-event … there was no credible risk of separation of SA from the National Electricity Market (NEM)”. However, as described above, a non-credible contingency event occurred. The immediate consequence was that 315 MW of 579 MW being provided by wind generators north of Adelaide, and connected to the damaged transmission lines, i.e. 17% of total demand at the time, went off-line within 42 seconds of the first line failure.

The near instantaneous loss of 17% of supply would be an extremely challenging emergency for any electricity supply system. The rate of change of frequency was too fast to allow large scale controlled load shedding to avert overloading.

Consequently the automatic protection systems caused Heywood, the main interconnector to Victoria, to trip out, followed within 0.2 seconds of all gas and wind generators still connected. The result was what is called a region-wide blackout.

Why this wind generation went off-line is not known and AEMO does not speculate. Others have suggested that the protection settings on the relevant wind generators may have been “too conservative”, i.e. unnecessarily sensitive, meaning that the generators were tripped out when they did not need to be.

If this were correct, the trip would not have been caused by the wind generators per se, but on the way in which they were being operated. The correct answer will not be known until AEMO completes its full incident report.

This time last year the now closed Northern power station was generating on average an output of around 250 MW most days. It is highly likely that it would have been affected by, and responded to, a sudden large loss of load in exactly the same way as the windfarms, with the same ultimate consequence.

Furthermore, contrary to some uninformed comments, there was significant gas fired generation capacity available but not being used at the time of the failure. Three of the four units at Torrens Island B were generating 246 MW in total, but had full capacity of 595 MW available, while two of the fast response Open Cycle Gas Turbine plants, also located close to Adelaide, were not generating but had 334 MW available.

There are well defined, but very seldom used procedures for restoring supply after a total blackout. The AEMO report describes in detail how these complex and difficult processes were implemented. The preferred procedure was not followed because the gas turbine generator being paid to provide black start services was unable to supply the full contract power level. This meant that first supply was established through the Heywood Interconnector and the first customers were reconnected less than three hours after the blackout.

Throughout the night gas generators were progressively brought on-line, followed by windfarms unaffected by the transmission line failures. AEMO estimates that by midnight on the Wednesday the load was about 60% of its day-ahead forecast level expected prior to the event, and by 9.00 the following morning it had reached over 80%.

Given the magnitude of the transmission line failure, not to mention ongoing local storm and flood damage to the low voltage distribution network, the speed with which electricity supply was restored across most of South Australia is a tribute to the skill and hard work of AEMO and all participants in the electricity supply industry in the state.

Hugh Saddler is an energy analyst with Pitt&Sherry and his work is supported by The Australia Institute.

Comments

7 responses to “Blackout reality check: Coal and extra link may not have changed anything”

  1. trackdaze Avatar
    trackdaze

    Politics at its lowest.

  2. Rod Avatar
    Rod

    It is very disappointing that no one else has highlighted what a fantastic job they did in getting people back on line in under four hours.
    People seem quick to blame but hesitant to recognise those who would have been under immense pressure at the time.

    1. Jane Hilton Avatar
      Jane Hilton

      I agree Rod, in most places the power was returned with incredible speed considering the situation.

      1. Dudley Hunter Avatar
        Dudley Hunter

        Yes. Once they turned Pelican Point CCGT on. It takes 4 hours to fire up. Hey presto power.

    2. Dudley Hunter Avatar
      Dudley Hunter

      All they did was switch Pelican Point on. It should never have been off.

      1. Rod Avatar
        Rod

        Just read the summary of the 2nd report. Apparently the interconnector provided enough power to get TIPS started. As some units were still warm I guess they could have been up and running in four hours.
        No mention of Pelican Point in it I read unless you know something more?
        I can assure you there were lots more grey hairs in the Electranet and TIPS control rooms after this event.
        https://www.aemo.com.au/Media-Centre/-/media/BE174B1732CB4B3ABB74BD507664B270.ashx

  3. Dudley Hunter Avatar
    Dudley Hunter

    The blackout happened at 4.16 pm. By 11pm, power was 90% restored across SA. Are you saying the transmission towers were repaired in just 7 hours? I don’t think so.
    What happened is Pelican Point CCGT was ordered on at 6.36pm (it was offline). It takes 4 hours to fire up. By 11pm almost all power is restored.

Get up to 3 quotes from pre-vetted solar (and battery) installers.