THE SCOPE FOR INTEGRATING THE ENERGY
MARKET AND NETWORK SERVICES
Summary draft report
October 2000
The scope for integrating the energy market and network services
Summary draft report
Introduction
We were required by the Code to review the adequacy of the existing criteria for the determination of regions in the national electricity market and whether the addition or substitution of different principles would better meet and facilitate the Code objectives. We considered a range of options from minimal change to the existing criteria all the way to full nodal pricing. Some form of risk management mechanism is essential in order to achieve the underlying objective of the market of minimising impediments to trade between regions. We therefore looked as part of our review at the performance of the existing settlement residue auction arrangements, at the potential impact of any new criteria on those arrangements and at ways of refining the existing arrangements better to achieve the underlying market objective.
This review also encompasses the assessment we were required by the Code to undertake of the financial impact of distribution losses on market participants. That review was required to consider in particular whether marginal loss factors could be used to calculate distribution losses. We have also, as part of this element of our review, sought to resolve the current uncertainties over the application of the existing loss factor arrangements, in particular to new market participants.
We have adopted the same three-part test of theoretical and practical soundness, and the need for the outcomes from the review to deliver a demonstrable overall improvement compared to the existing arrangements, that we developed for our transmission and distribution pricing review.
We have also been concerned to ensure that our proposals are consistent with the existing broader design of the market, and with planned and prospective future developments.
Criteria for determination of regions
The existing criteria for the determination of regions are set out in figure 1. We have examined the scope for gains to be made on all three measures of efficiency:
productive, obtaining the maximum output for a given set of inputs;
allocative, devoting resources to their best use; and
dynamic, ensuring the most efficient future investment
as a result of refinements to those existing criteria.
Figure 1: Existing criteria for determination of regionsThe boundary of a region should be closed and enclose at least one significant load and/or generation centre.
Where practicable, significant generation and load centres separated by network constraints should be located in separate regions where those network constraints are likely to influence the optimal despatch of generation and/or scheduled load in the order of 50 hours or more in the financial year for which the intra-regional loss factors are pre-determined.
The regional boundaries should be located so that transfer limits can be clearly defined, and transfer flows across regions easily measured, at the regional boundary.
The application of pre-determined static intra-regional loss factors within a region, and the application of a pre-determined inter-regional loss factor equation, should not significantly impact on the central despatch of generation and/or scheduled load that would result from a fully optimised despatch process taking into account the effect of losses.
NEMMCO must aim to minimise the variation between the set of pre-determined loss factors and the resultant averaged intra-regional loss factors, and any errors in the inter-regional loss factor equation across the trading in the financial year for which the intra-regional loss factors were determined.
Where a connection point can be assigned to more than one region such that the criteria set out above can be met in either region, then the transmission network connection point will be assigned to the region such that the variation between the set of pre-determined intra-regional loss factors and the resultant averaged loss factors is minimised.
Within those requirements, the number of regions should be minimised.
Electricity markets are extremely dynamic even in the very short term. Currently, however, little of this dynamism is reflected in the regional structure of the market. The use of annual average marginal loss factors can be an extremely poor representation of actual despatch. The variation is typically 3 per cent off-peak and can rise to as much as 8 per cent or more. The welfare gain – utility less costs - from increases in productive efficiency as a result of better representation of losses is therefore potentially significant.
There is also scope for productive efficiency, and therefore welfare, gains as a result of better representation of constraints under a refined regional structure. This would also allow the current distortions to market outcomes implicit in some of the constraint formulae to be removed. The volumes traded mean that the potential savings here too are significant. Just a $10/MWh difference between the bids of currently constrained-off and despatched generators, multiplied by 1,000 MW of generation for 100 hours which is the extent of the current constraints between the HunterValley and Sydney, would value the productivity gain at $1 million.
Improvements in allocative efficiency will result from changes to the market clearing price as a result of more efficient despatch and the market’s response to those refined price signals. Especially industrial customers in the Latrobe and HunterValleys and central Queensland can expect, on average, lower prices as a result of a refined regional structure. Customers in Melbourne, Adelaide, Sydney and Brisbane on the other hand are unlikely to see increases.
Overall, the short and medium-term gains in productive and allocative efficiency from improved representation of losses and constraints within a refined regional structure are potentially worth tens of millions of dollars a year, depending on the precise criteria adopted.
The longer-term gains in dynamic efficiency as a result of improved investment decisions are potentially even more significant. The potential savings in the single example of the need to augment the transmission capacity from the LatrobeValley to Melbourne, identified in VENCorp’s latest annual planning review, are some $6 million to $8.5 million. With annual investment in the electricity industry of some $2.4 billion, the potential dynamic efficiency gains as a result of a refined regional structure might be of the order of hundreds of millions of dollars.
Managing market power
Market power exists as a function of the underlying nature of the electricity system, regardless of the regional or even the broader market structure. Nonetheless, it is an essential prerequisite of any refined regional structure aimed at capturing the potential welfare gains that it must not entrench, and if possible should mitigate the negative effects of, market power.
Our proposals will make the abuse of market power more transparent, and therefore easier to detect and rectify. They will encourage more investment in areas where market power currently exists. Our proposed improvements to the settlements residue actions will also make it easier for competitive retailers to enter markets.
As an added safeguard, however, in order to ensure that any new regions are of a sufficient critical mass and to prevent the creation of small isolated markets conducive to the abuse of market power, no region under the refined structure should encompass less than, say, 100MW of generation and/or load. A minimum size of region is also likely to make sense on grounds of practicability and in order to ensure that the resulting regional structure meets the requirement of the three-part test that it must deliver a demonstrable overall improvement compared to the existing arrangements. We should, however, welcome comments in particular on the most appropriate precise minimum size of a region.
A refined regional structure
We have considered in detail five options for the basis for a refined regional structure for the national electricity market, ranging from no change to the existing criteria for the determination of regions all the way to full nodal pricing. We have not considered options that step back from the existing criteria, for example by reducing the tolerance of constraints, because this would represent a disbenefit compared to those existing arrangements. Figure 2 summarises the basis for each of those options. Figure 3 sets out, for illustrative purposes only, the indicative regional structure that might emerge. NEMMCO will remain responsible for undertaking the further detailed analysis necessary to develop the precise regional structure, including regional boundaries, based on the revised criteria that emerge from our review.
Option (1) represents in effect no change, and option (2) minimal change, from the existing criteria. Option (1) would remove the current Snowy region which would result in a significant loss of efficiency. No change to the existing criteria is not, therefore, a realistic option.
Option (2) would create an expanded Snowy region encompassing Canberra and Yass, a second and possibly third region in Queensland, and a new HunterValley region. This would represent a significant potential gain in particular to end-use customers in Queensland. Prices might in some circumstances be almost 2 per cent lower than now in Cairns, and 1.5 per cent lower in Stanwell, under this structure.
Crucially, however, options (1) and (2) both retain a backward-looking approach to the treatment of losses and constraints. Continued reliance on historical information will pose increasing problems for efficient despatch once QNI, and subsequently the additional generation currently under construction and planned in Queensland, come on stream.
Options (3) and (4) therefore adopt a forward-looking approach. They also establish alternative quantified bases for tolerance of constraints and to replace the existing discretion over the variation between pre-determined and actual losses. They establish a range for the duration of constraints of between 25 and 50 hours and would require losses to be accurate to within ±2.5 to 3 per cent for between 90 and 95 per cent of all transmission exit points.
Option (5) represents full nodal pricing. Its net benefits, especially in the particular circumstances of Australia’s transmission network, are arguable. Without a firm hedging mechanism, which would be difficult if not impossible to devise, it would expose participants to largely illiquid markets and therefore unacceptable risks. Moreover, nodal pricing that would allow the co-optimised despatch of active and reactive power is currently incompatible with five-minute despatch and pricing.
A refined regional structure along the lines of options (3) or (4), on the other hand, would cost very much less to implement: the direct one-off costs might be some $5 million. The indirect costs are equally likely to be small. The results of analysis of such a refined structure by NEMMCO and PHB Hagler Bailly are presented in the appendices to our report. It would yield potential productive and allocative efficiency gains of at least $150 million over the next ten years. The dynamic efficiency gains over the same period are likely to be much more significant: of the order of $500 million to $1 billion. Such a refined regional structure would also result in reductions in prices to many end-use customers, in particular in the bush.
The criteria for the determination of regions should therefore be revised to:
encompass forward-looking treatment of constraints and losses;
quantify tolerance thresholds for the duration of constraints and the accuracy with which loss factors are represented. The threshold for the duration of constraints should be within the range 25 to 50 hours. Losses should be required to be accurate to within ±2.5 to 3 per cent for between 90 and 95 per cent of all transmission exit points. We should welcome comments in particular on the precise levels at which these thresholds should be set for the purpose of determining the final criteria; and
limit the minimum size of a region to include 100MW of generation and/or load.
1
Figure 2: Summary of options for revised criteriaOption 1 / Option 2 / Option 3 / Option 4 / Option 5
Existing criteria, in particular: / As for option 1 but improved representation / Revised criteria including: / As for option 3 but: / Full nodal pricing at transmission level
backward-looking loss factors;
50 hour constraint limit
minimum number of regions / of loss factors / forward-looking loss factors;
variance in loss factors limited to 3 per cent for 90per cent of nodes
50 hour constraint limit / variance in loss factors limited to 3 per cent for 95per cent of nodes;
25 hour constraint limit
minimum size of region to encompass 100 MW of generation/load
1
Figure 3: Indicative regional structuresOption 1 / Option 2 / Option 3 / Option 4 / Option 5
South Australia / South Australia / South Australia / South Australia / 340+ nodes
Victoria / Victoria / MountGambier / MountGambier
New South Wales / Snowy/ACT / Victoria / Victoria
Queensland / New South Wales / Snowy/ACT / Latrobe
South East Queensland / New South Wales / Snowy/ACT
Northern Queensland / HunterValley / New South Wales
South East Queensland / HunterValley
Central Queensland / South East Queensland
Northern Queensland / Central Queensland
Northern Queensland
possibly / possibly / possibly / possibly
HunterValley / HunterValley / Latrobe / South West Queensland
Central Queensland / South West Queensland / Riverland, SA
Western/Northern Victoria
4 – 5 / 6 – 8 / 9 – 11 / 10 – 13 / 340+
1
Figure 4: Revised criteria for the determination of regionsThe regional structure of the national electricity must:
ensure that the boundary of a region is closed and encompasses at least [100] MW of load or generation;
restrict the expected error between the dynamic loss factors and the resultant averaged intra-regional loss factor determined to apply for the following financial year to less than [2.5-3] per cent for [90-95] per cent of transmission exit points within each proposed region; and
conform as closely as practicable with significant forecast network constraints, where those network constraints are likely to influence the optimal despatch of generation and/or scheduled load of the order of [25-30] hours or more in the following financial year.
So far as practicable consistent with the principles set out above, the structure should also ensure that:
regional boundaries are located so that transfer limits can be clearly defined, inter-regional loss factor equations determined and transfer flows across regions easily measured; and
elements of the proposed regional boundaries align with existing regional boundaries.
Figure 5: Implementation timetable
final report published / December 2000
Code changes to implement report’s
conclusions and recommendations
forwarded to ACCC / February 2001
revised criteria for determination of
regions incorporated in Code / August 2001
NEMMCO publishes report recommending refined
regional structure based on revised
criteria / December 2001
new regional structure approved by NECA / April 2002
forward-looking loss factors introduced / July 2002
first settlement residue auction incorporating refined regional structure / September 2002
revised regional structure implemented / July 2003
1
The criteria should be recast to identify those which it is essential that any refined regional structure based on those criteria must meet and those where an appropriate balance between conflicting objectives will need to be struck within the decision-making process. NEMMCO should also be required, in drawing up its recommendations for a refined regional structure based on those criteria, explicitly to have regard to the same three-part test of theoretical and potential soundness, and the need for the outcomes to deliver a demonstrable overall improvement compared to the existing structure, that we have adopted in developing the revised criteria themselves.
Revised draft criteria for the determination of regions are set out in figure 4. An implementation timetable for the introduction of a refined regional structure based on those criteria is set out in figure 5.
Beyond the initial refined regional structure
The regional structure of the market will need to continue to evolve. The existing arrangements for reviews of regional boundaries, however, are at the same time unstable and introduce a lag between the occurrence of a constraint and the creation of a new region to reflect that constraint. They also involve very significant and unacceptable regulatory risk.
In order to reduce that risk in the future, the outer boundaries of the new regions established under the refined structure that emerges from our review should remain fixed, at least until the first five-yearly review. There should, however, be scope to sub-divide a region where a new constraint emerges. The criteria for sub-division should be based on those used for determining the initial refined regional structure. Where a proposed sub-division is based on a forward look rather than historical information, however, it should require forecast breach of the criteria by a threshold of one-and-a-half times the underlying criterion.
Any subsequent sub-division of existing regions should be implemented, with a minimum six months’ notice, on 1 July each year. This will coincide with the annual recalculation of loss factors and therefore minimise disruption.
In order to help the market to prepare for likely sub-divisions of existing regions, NEMMCO should publish a regional assessment as part of its annual Statement of Opportunities.
There should also be regular five-yearly joint NECA and NEMMCO reviews of the regional structure, the first to be completed by July 2008, using the Code consultation procedures.
Improved risk management
The existing settlement residue auction arrangements have more than proved their worth. Overall, the proceeds from those auctions so far have totalled some $75million, compared to total actual residues of $114 million. Proceeds from the first two rounds of auctions represented less than 50 per cent of actual residues. Proceeds in the third quarter, however, were almost at par. Proceeds in the latest quarter actually exceeded residues by almost $2.5 million. Increasingly, therefore, the auctions represent good value for end-use customers, to whom those proceeds are returned through reduced transmission use of system charges. The survey we conducted as part of our review also overwhelmingly confirmed the value of the hedge instruments those arrangements allow them to construct to market participants.