Initial Thoughts about the AEMO’s Electricity Statement of Opportunities (ESOO) for 2009

Today I attended the launch of the 2009 Issue of the ESOO at the Brisbane venue.

I admit to arriving a little late, by which time they had run out of the copies of the full document to give away (though Robyn assures me mine is in the mail – thanks!) and so my comments below must be read on the basis that I have not even had the chance to open the cover on the full document.  Also keep in mind that this is written on the same day as the launch of the ESOO, and towards the witching hour (not that that is much of an excuse, in reality – as my thinking would not be much better tomorrow).

Especially this year, I will look forward to reading the document in detail, for reasons both internal and external.

General Comments

By my reckoning, this was the 11th SOO published since the NEM started on 13th December 1998.  We have already noted the 10th birthday of the NEM less than a year ago.

Over the period of the past 10 years, my impression is that we have seen the SOO grow from a relatively small document (201 pages in 1999) and track upwards to become a truly gargantuan effort in more recent times.

Hence, it was refreshing to see it slimmed down, somewhat, for the 2009 issue.  This was due, in part, to the excision of the Annual National Transmission Statement (ANTS) that will be produced in a separate document (with a new acronym) in December each year.

Being a company that has set its goal as making the market understandable to everyone, we certainly see the need for such information to be widely (and easily) available.

2)  In the presentation of the ESOO, it was also pleasing to note that Tim George (EGM Planning) kicked off discussions of the ESOO with a general overview of the structure of the ESOO, and the way in which it was put together – basically he “set the scene”.

Such an introduction may have been given in previous years – but (if so) I don’t recall it.  Certainly, my view is that Tim’s introduction will have helped the general level of understanding that people gained of the document on the day – and probably headed off a number of questions that would have happened afterwards, otherwise.

What did I observe with the Questions Asked?

No doubt you will hear much about the ESOO in the coming weeks, and have a chance to peruse it yourself.  Hence there’s nothing I can say here about the document itself (especially not having seen it) that would be of value to you.

Tonight I’d just like to note a few things I observed about the questions asked after the forum.

1)  In general terms, there seemed to be two types of questions:

(a)  Questions that go to the core of the methodology used by AEMO in the production of the document.

(b)  A number of other questions were asked that indicated that the AEMO could probably still do more to clarify (up front) how the ESOO fits into its other suite of documentation, reports and data (e.g. the drought studies, and the MT PASA outlook, for instance) – and of the difference between possible energy and capacity shortfalls.

2)  In particular I have recorded the following (if, for nothing else, so we have a record we can return to at a later date).

(a)  Questions were asked about the assumptions underlying the demand forecasts, such as the impacts assumed of:
i.  The GFC – note we already looked at this here, to some extent
ii.  The increase in MRET
iii.  The CPRS, if/when it is implemented.
iv.  Assumptions about specific major load projects (turns out there is a similar process for deeming them “committed”)
Specifically, people wanted to know more about what had been assumed in terms of price-responsiveness of demand.

(b)  Questions were asked that were really about the logic AEMO used to both identify potential generation projects (to list them as “Announced”) and then to determine if they are to be deemed as “Committed”.

(c)  Questions were asked about the approach taken in the modelling, such as:
i.  To account for the much-talked-about increase in Forced Outage Rates that might occur in brown coal units over the coming decade in response to reduced maintenance expenditure under a CPRS scenario.
ii.  To discount installed wind capacity to an effective capacity amount at times of peak demand.

Improvements to the ESOO?

During the presentation, it was noted that the AEMO would be (within a month or so, I think it was) starting a review process of the ESOO document and, through this, seeking feedback about how the document could be improved.

Hence, the following thoughts are my initial two cents worth:

Record of the Event

For some people, it may well be that an A/V recording of the ESOO presentation itself (including the questions asked, and answers given) would be useful.

It’s probable that the questions will be transcribed for the minutes of the Market Information Forum and stored on the AEMO website – I think this was done last year, but the history on the AEMO site does not go back that far.  However, some might still prefer an A/V record.

Focus on NEM-Wide Demand

One of the driving principles for the development of the NEM was to encourage inter-state competition between competing sources of generation – and hence deliver a more efficient (and sustainable) outcome for energy users.  This has largely been achieved, most of the time.

For this reason, I find the habit of retaining a state-based approach to the development of demand forecasts somewhat anachronistic.   I hold this view for a couple of reasons, including:

1)  The method: At a state-based level, I understand that the Jurisdictional Planning Bodies use both a top-down and bottom-up approach to the forecasting of demand.  On a NEM-Wide basis, this ends up being a middle-down, and bottom-up approach (though AEMO does coordinate the approach somewhat by imposing uniform assumptions about the economic growth rate).

2)  The message: As has been noted by many people, the incidence of inter-regional constraints is still higher than some people might like – and, as many know, the incidence of transient “Economic Islands” is the primary driver of spot price volatility in the NEM.

Hence, we believe that it is important for AEMO to clearly identify, when they forecast a possible shortfall in reserve, whether this is due to an imbalance between supply and demand across the NEM, or whether the issue is more a function of localised issues (particularly transmission constraints, or just undersized links).  Note, here, that I am not necessarily advocating a “gold-plating” approach to transmission augmentation, just pointing out that documents such as the SOO should, ideally, make it clear to everyone what the possible solutions might be (including transmission expansion).

Hence, I believe it would send an important message to every reader of the underlying nature of supply & demand across the NEM if the document (and the launch presentation) were to open with a whole-of-NEM focus in terms of installed capacity, peak/average demand forecasts, and hence capacity shortfalls.

That won’t stop journalists taking the information and applying it with respect to state issues (such as Brian Robins’ piece in the SMH today) but it will help, over time, to change people’s frames of reference.

Indeed, it is for this reason that we have ensured (since version 4 and up) that NEM-Watch has provided a clear view of the “Whole of NEM Demand” and the “NEM-Wide Instantaneous Reserve Plant Margin”.

10 Years of History

The NEM now has 11 years of history on which people can start to see underlying trends in terms of demand growth – the ESOO should make use of this data history.

As we found out ourselves, reference to only a limited amount of history can result in an incomplete/inaccurate frame of reference.  Particularly, charts highlighting forecast growth rate of demand should be presented in conjunction with 10 years of demand history so readers can clearly see any variations.

Sanctity of Load Forecasts

In my (albeit limited) experience, it seems that participants, observers (and their consultants) will be fairly flexible in terms of questioning the details included about supply-side options in the NEM, but will be much less likely to question the underlying assumptions for demand growth (or make their own load forecasts).

Indeed, some seem to treat an M50 forecast as “gospel”.

Given the propensity for this to happen, it would be worthwhile for the AEMO to include some review (possibly as an appendix) to highlight how “accurate” these load forecasts have been over the years, and whether this accuracy is increasing or decreasing.

I did this in July 2008 – just for NSW winter peak demand, and so I can understand that it will take a fair amount of time.  However, I do believe that it would be worthwhile.

Trend the Past History of NEM Reliability?

In the same way as I would be suggesting that the ESOO provide a historical trend of NEM demand, it would also be worth considering whether the SOO should also provide an overview of the level of reliability that has been delivered by the NEM each year to date.

I’m not thinking that this review would be as detailed as the Reliability Panel reviews conducted reasonably regularly (such as this “Comprehensive Reliability Review” from December 2007) – I am more thinking that the ESOO could perhaps form the role of a central long-range review and planning document that would then reference other more detailed reviews, for those who want more information.

It certainly should not be AEMO’s role to change the reliability standard, nor (of course) to assess its own performance in delivering on that standard.  However, it does seem logical to me, at least late at night, for the AEMO to take a lead in incorporating a summary view of past reliability (i.e. statements of fact) into the ESOO to assist a reader’s frame of reference.

Given that AEMO takes a probabilistic standard from the Reliability Panel (i.e. the 0.002% number) and converts it to a deterministic standard as a guide (the Minimum Reserve Levels) it would also make sense to highlight what the reserve levels were, in each region (and NEM-Wide), at times of peak demand in previous years.  Again, the purpose of doing so would be to enhance the reader’s frame of reference with respect to the forecasts.

More ideas later, but that’s enough for tonight…