I wish to make a few specific and personal comments on the approach laid out in the draft determination, supplemented by impressions gained during the September consultation workshop on this topic.
A lack of cost reflectivity in consumer tariffs generally and network tariffs in particular has been a feature of the Australian market from the beginning. Air conditioning has been a growing source of demand, imposing additional strains on the network for a long time, and yet there was no great push before to reform network pricing; rather, networks simply gained approval for and built the capacity needed to accommodate the growth. Those costs have gone into increased network charges across the board.
The current interest in tariff reform seems to have been driven by the related issues of declining demand and the increasing viability and penetration of embedded rooftop solar PV technology, with a further technology boost in prospect from improving battery technology and costs. Notably, such technology poses little or no extra cost burden on the network. However, under existing tariffs high levels of PV penetration appear to threaten the current regulatory and business models that sustain the profitability of distribution businesses.
A key element of the rule change proposal is to establish long run marginal cost (LRMC) as the prime criterion for setting network tariffs. In practice, however, most distribution network elements are now a long way from becoming constrained, not least because of the extensive investment in recent years, coupled with general decline in demand. As the AEMC notes in its draft determination, this means that LRMC is in most cases very much lower than the historical cost of the network at each point which. Under current regulatory arrangements, this historic cost is allowed to be recovered. This in turn implies that tariff setting will be dominated by a strategy of cost recovery, defined by the revenue cap imposed on distribution businesses by the AER as a separate process.
With most retail customers still on simple accumulation meters, coupled with a likely jurisdictional requirement for some level of tariff uniformity, it seems likely that retail tariffs will retain a relatively simple structure. Under the proposed rule change, the exercise is then to recover the difference between the revenue requirement and what can be recovered with a (low) LRMC, in a way that least distorts the economic outcome. In the absence of time of use metering, this leads almost inevitably to reliance on a much larger fixed cost element in network tariffs, through an actual fixed charge or a declining block structure with a similar intent and outcome. How a network charge on retailers filters through to customers is an important issue here.
This element of any tariff proposal should be scrutinised very carefully, or perhaps even ruled out as an option beyond a certain point, for the following reasons:
- As argued in a number of submissions, a fixed charge is inequitable as it hits lower level and likely less well-off consumers to the same level as larger users.
- Taking a system wide view, the case for fixed charge in the network tariff, which would have the effect of discouraging solar PV, is not compelling. While PV may or may not be “paying its fair share” for the distribution network, it is almost certainly being underpaid for the long run contribution it makes to shaving the whole of system peak, which is substantially driven by industrial and commercial loads and associated air conditioning. There is a brief acknowledgement in the supporting NERA studies of this fact, but it is effectively ignored beyond that. More careful examination could well show that the distortion in network tariffs is matched by a distortion in customer retail tariffs that undervalues the pattern of PV energy to the system as a whole. A reform of network tariffs without a commensurate reform in general retail tariffs (and associated dynamic load management) could well lead to a worse economic outcome.
- Pervading the draft determination is an assumption that the technical requirements for, and cost of, an efficient network is somehow knowable by the regulator, even with expert advice. An efficient cost seems unknowable given the huge asymmetry in knowledge and information between the regulated entity and the regulator. So by allowing fixed charges whenever other cost recovery methods look difficult, the regulator could well be allowing easy passage of investments and practices that are not efficient. After all, an important overriding consideration, regardless of long run and short run calculations, is that customers are prepared to pay for the service they receive. With an unavoidable fixed charge, this element of choice is removed.
If PV is acknowledged as a technology gaining momentum as appears to be the case, a greater emphasis on fixed charges is a potentially dangerous and certainly inadequate response. If more careful study shows the perceived distortion is indeed real while viewing the system as a whole, a move to-a-time of use metering and tariffs would be a much superior response. Further, if costs must be recovered, some proxy of an LRMC structure would be better than a fixed charge, even if investment in that part of the network might be a long way off.