AEMC ponders reliability options

With summer approaching and an intense political focus on reliability of supply and power bills, particularly following the closure of firm power stations in South Australia and Victoria, how to manage reliability at lowest cost is again attracting attention.

This is shown by the 25 submissions the Australian Energy Market Commission’s (AEMC) consultation paper attracted on the proposed enhancement to the Reliability and Emergency Reserve Trader (RERT) rule change.  It attributes the “significant” level of active interest down in “large part” to the fact that the RERT imposes costs on consumers.

Following those submissions, the AEMC has now issued an options paper that sets out three possible approaches.

Background

The RERT is a safety net mechanism through which the market operator can obtain additional supply capacity to meet a forecast shortfall against the reliability standard (<0.002 per cent of demand not met in a financial year). This supply is then held out of the market and used as a last resort to avoid involuntary customer interruptions. The cost is then recovered across all customers in the affected regions.

AEMO has submitted a rule change request reflecting the growing proportion of renewables in the generation mix, an ageing fleet of firm generators and the unexpected retirement of capacity. The AEMC notes that the market operator has argued that an enhanced RERT will be needed as a stronger safety net and identified three key issues with the current framework[i]:

• The procurement lead time (maximum 9 months) and contracting durations are too short.
• There is a lack of a comprehensive risk assessment framework. It argues that there is inconsistency between the operational objectives of the current RERT to meet the reliability standard that allows some load shedding and maintaining a reliable operating state which in their view implies no load shedding.
• Bespoke negotiated contracts can be problematic because they make it difficult for AEMO to compare offers and can be very time consuming.

In their proposal, AEMO has also questioned the level and form of the existing Reliability Standard which is set by the Reliability Panel. They opened the possibility of other metrics, one of which is the “Loss of Load Probability” (LoLP), being the proportion of random simulation runs that have any load shedding.

AEMO also raised some other key questions in their rule change that are not addressed in the options paper. These were:

• Whether the contracting lead time should be extended to 12 months.
• Whether AEMO can contract several years into the future if low priced offers emerge.
• Whether AEMO should standardise its contracting structures.

These will presumably be dealt with by the Commission at a later stage.

The Reliability Panels Assessment

Aside from the submissions and consultation it has held on the proposed rule change the Commission has also received an assessment from the Reliability Panel on whether the reliability standard remains appropriate as the procurement trigger for the RERT.

The reliability standard is reviewed every four years by the Reliability Panel and its most recent review was earlier this year.

The panel also considered whether the standard needs to be tightened, the potential costs and benefits from any tightening of the standard, whether the Panel considered any alternatives to the reliability standard and the implications if the RERT procurement trigger was delinked from the standard.

The Panel found the reliability standard is a “crucial market standard” and reiterated its recommendation its 2018 review of the reliability standard and settings that the “materiality threshold for reassessing the level of the reliability standard has not been met at this time”. It did, however, acknowledge that the NEM is “transforming, and these materiality criteria may be met in the future”. It also noted that the AER must publish the results of a Value of Customer Reliability (VCR) survey by the end of next year and that this could be a trigger for a future Panel reassessment of the reliability “at or prior to the next four yearly review” it will also continue to monitor emerging trends.

It found that EY modelling that was undertaken for the Panel’s regulator review forecast a level of reliability in the NEM significantly better than the 0.002 per cent trigger in all regions for the review period.

The AEMC comments that it understands that AEMO is in the process of refining its views on the appropriateness of the reliability standard including incorporating more recent analysis and information and considering how it works in the current environment where tail events are becoming more extreme. This is expected to be available mid next month and AEMC notes it will be an important input in considering the issue through the rule change request.

The Options

Option 1

This would see changes made to the National Electricity Rules (NER). The market operator would only be able to procure reserves if the reliability standard is projected not to be met and it would only be allowed to procure an amount to just avoid a breach of the standard. During the 2017-18 summer, AEMO appeared to procure reserves well beyond this level (see Planning for Summer: What Have We Learnt?). This option is designed to address concerns about AEMO’s existing level of discretion through tightening the rules.

AEMO would continue to operationalise the reliability standard consistent with their Reliability Standard Implementation Guidelines (RSIG) which they may change from time to time. Other aspects of the procurement trigger and volume would stay the same.

The practical impact of this option would be that the procurement volume would be linked directly to the reliability standard, which would improve transparency of the amount of reserves that would be procured under the RERT and “providing stakeholders more certainty about the level and costs of out of market services”. The Commission notes that this could also result in an inefficiently low volume of reserves being procured, for example where low cost reserves were available to AEMO but not purchased as the power system remained slightly within the standard.

It would also mean that market participants would have the chance to collectively provide the desired level of reliability with AEMO making up the difference. Distortions would be minimised.

Option 2

This approach would combine the two separate processes of determining whether to procure reserves and how much reserves would be needed. The combining of these two processes would occur under a single economic assessment framework.

This framework would procure an amount of reserves which minimises the combined estimated cost of buying reserves and the estimated benefit of avoiding load shedding. As every additional MW of reserve is purchased, a model will simulate how much load shedding it is expected to avoid. This amount times VCR is the benefit. Every year, AEMO would keep buying reserves until its cost exceeds the modelled benefit.

This would remove the existing trigger from the NER and make the reliability standard irrelevant in establishing whether and how much reserves are needed.

The Commission comments that AEMO is currently proposing an economic cost minimisation model that aims to minimise the total cost of the RERT.

In the medium term a benefit of this approach is that it can make the trade-off between costs in a more dynamic way closer to real time and taking account of better information compared to the four-yearly review of the reliability standard and settings.

The AEMC notes that this may have a number of downsides, including:

• Shifting some of the risks associated with provision of some capacity from market participants to consumers;
• AEMO does not have financial incentives to make the difficult trade-offs as accurately as possible; and
• Incentivise market participants to leave the market with the hope of being better remunerated in the RERT market.

If applied in the short term, the options paper states, that it could be an improvement on the current approach, but if only applied in the short-term it could also increase the cost of procuring reserves.

Option 3

Under this option, like Option 1, changes would be made to the NER to remove the market operator’s level of discretion with the RERT procurement trigger and the amount of additional supply capacity. “Instead AEMO would only be able to procure if the reliability standard is projected to not be met, and would only be allowed to procure an amount related to just avoiding a breach to the reliability standard”.

This approach primarily addresses stakeholder concerns regarding the discretion of AEMO in determining whether to procure reserves and if so how much, but goes further than the first option by requiring changes to the way the market operator operationalises the reliability standard. The option also recognises the difficulty in implementing an annual standard for operational decisions on acquiring additional reserves.

It goes on to present some possible specific changes.

Under a possible approach outlined in the paper it says a monthly expected USE standard rather than an annual standard could be determined and purchases of reserves for periods less than one year could be made with reference to modelling of monthly expected USE against the monthly reliability standard.

It also proposes that the purchase of reserves be subject to the same sort of economic assessments as discussed in Option 2. This might mean that where reserves are quite expensive, that AEMO is permitted to remain outside the reliability standard.

AEMC’s paper notes that the main benefit of its third option is that it results in lower cost and/or more reliable outcomes, which are also more predictable. On the flipside, requiring AEMO to employ a particular methodology and putting it in to the NER may constrain the market operator and lead to inefficient outcomes. “It would likely be more appropriate for the discretion about how to operationalise the reliability standard to be limited through the Reliability Panel or AEMO guidelines.”

A forum is expected to be held in mid-November on the options paper and appropriateness of the reliability standard with a date yet to be announced. Submissions on the options paper close on 29 November and a draft rule is expected by end of January next year with the final rule to be finalised in April.

[i] AEMC Options Paper, Enhancement to the Reliability and Emergency Reserve Trader, 18 October 2018