The Commonwealth Energy Minister, Chris Bowen, has lodged two requests with the Australian Energy Market Commission (AEMC) to amend the National Electricity Rules (Rules) to ensure data centres contribute to network costs they cause or accelerate and to ensure these costs are not borne by existing electricity consumers.
The requests identify 7 gaps in the Rules which are split across two rule change requests. While the Minister prefers the two requests to proceed concurrently, it may be the second request (gaps 5 to 7) will take longer as it involves more complex issues.
The changes are significant for data centres looking to get grid connection for their projects, both in terms of the charges payable (which will be higher) and the negotiations that will be required with networks.
The key takeaways are:
- Data centres connecting to the distribution network will be required to pay for the costs of new or accelerated transmission augmentations required to meet their load requirements.
- Data centres will likely be required to provide prudential support to cover, at least, their charges for network augmentations required by that data centre which are at risk of stranding. Networks are likely to require bank guarantees or cash (eg. capital contributions or prepayments) so they do not have the risk of bearing any under-recoveries.
- Networks will be restricted from rolling in data centre funded assets, which are stranded, into their regulated asset bases (which are recovered from consumers). This will mean networks are more likely to require full bonding of the relevant charges for augmentations discussed in items 1 and 2 above.
- Networks must publish detailed information on existing and prospective data centre connections including proponent, site, timing, capacity, connection status and planning status. This does not extend to customers, or potential customers, of the data centre.
- At the transmission level, the current mechanism which allows applicants to fund upstream augmentations through a “funded augmentation” will be clarified to ensure (a) it applies to data centres (particularly clustered connections) and (b) future users of those funded augmentations contribute to the costs paid by the original funders. However, unlike gap 1, this does not appear to be a mandatory mechanism requiring transmission connected data centres to fund transmission augmentations but see gap 7. Query if the intention is that a funded augmentation could be agreed in place of formulaic charges discussed under item 7 (referable to gap 7) below.
- Transmission connected data centres will likely be required to bear the costs of jurisdictional schemes which are currently recovered from distributors. Charges will likely be based on their contracted demand rather than consumption. This includes schemes like the NSW Roadmap and NSW Climate Change Fund. It will not be possible to avoid these costs by negotiating bespoke services.
- Similar to distribution arrangements, transmission connected data centres will likely be charged for network augmentations according to a specific guidelines which may be formulaic (eg $/MW or $/km). These charges may change over time to reflect the actual cost of the transmission augmentations.
The AEMC will need to initiate these rule changes and process them in accordance with the standard rule change process (which could take 6-12 months). In the past, existing connections have been grandfathered from new rules by reference to where they are in the connection process, eg. enquiry, application or offer stage. The Minister’s requests do not refer to the transitional or grandfathering arrangements but we expect the preference would be to apply the changes broadly, including to data centres currently in the application process. This could create uncertainty for data centres if the scope and level of network charges are uncertain and could materially increase post-financial investment decisions. Alternatively, given the direction given by the jurisdictions, networks may be able to start implementing some of the changes in substance under the existing rules.
Gap 1: Pricing upstream transmission augmentations into distribution connections
Gap: Rule 5.3AA sets out a framework for connection applicants and distributors to negotiate “distribution network user access” and in particular, power transfer capability levels. This can include charges for upstream transmission augmentations to achieve the capability.
The request states this provision does not apply to loads such as data centres. Even if it did, a "negotiate in good faith" requirement is considered insufficiently robust to protect small customers from cross-subsidy risks as there is minimal incentive for distributors and loads to ensure that upstream costs are not socialised.
Request: The proposed change involves amending Rule 5.3AA to ensure (1) large loads are covered by the rule and (2) that the costs of required transmission augmentations are paid by the connection applicant, rather than passed through to other customers. The AEMC is also asked to consider whether a mechanism is needed to recover actual costs from large loads if augmentations exceed their forecast costs.
Impacts: Data centres connecting to the distribution network will likely be required to pay for upstream transmission augmentations required or accelerated by their connection. The request notes as an example, that NSW data centre load will accelerate the Sydney Ring South project by three years at an additional cost of $1 billion. While the Rules may enshrine the principle of causer pays charging, work will likely be required to determine the level of causation, and allocation of costs, at an individual data centre level.
It is also worth noting that the equivalent “transmission network user access” rules were removed in 2018 so this does not apply to transmission connected data centres. However, see discussion in gap 5 and 7.
Gap 2: Prudential arrangements for asset stranding
Gap: Where a data centre's expected demand does not eventuate, there is a risk that network assets built to support the connection will be stranded and costs recovered from the existing consumer base. The NER provides optional prudential tools (capital contributions, prepayments), but there is no obligation for network providers to use them, and the Rules provide more detail for distribution (Rule 6.21) than for transmission (Rule 6A.28).
Request: The proposed change would amend Rule 6A.28 to align transmission prudential arrangements with the distribution framework in Rule 6.21. Network service providers (NSPs) would be required to implement specific prudential arrangements (e.g. pre-payment, bank guarantees) for connections above a certain threshold.
Prepayments or bank guarantees are preferable as they do not require legal action to enforce them but where legal action is required, NSPs should be compelled to take legal action to recover the money to ensure the stranding costs are not transferred to consumers. The request notes that the prudentials should cover the actual capital expenditure noting that cost overruns on augmentations are not uncommon.
Impacts: Data centres will likely be required to provide prudential support, in the form of bank guarantees or cash equivalents, to cover their charges. This is presumably limited to the augmentations required by that data centre which is at risk of stranding. The prudential support may initially reflect the full charges over the term with a step down over time.
Gap 3: Safeguards for reallocating the costs of network assets description
Gap: Under clauses 6A.19.2(8) (transmission) and Schedule 6.2.1(e)(8) (distribution), a NSP may roll part of the value of a user-specific negotiated service asset into its regulated asset base, shifting costs to the broader customer base. While three safeguards exist (the asset must provide shared services, capital costs must not have been recovered, and the AER must approve the roll-in), these provide "strong, but not complete" assurance against cost-shifting.
Request: The proposed change would further restrict reallocation of specified assets under clause 6A.19.2(8) and Schedule 6.2.1(e)(8), or alternatively provide the AER with enhanced guidance for assessing roll-in applications. A classification for customer-specific assets subject to these enhanced protections (eg. a capex threshold or specified asset types) should also be introduced.
Impact: Networks will be restricted from rolling in data centre funded assets, which are stranded, into their regulated asset bases (which are recovered from consumers). This will mean networks are more likely to require full bonding of the relevant charges for augmentations discussed in gaps 1 and 2.
Gap 4: Collection, sharing, and publication of information on large loads
Gap: Transmission network service providers (TNSPs) are required to publish a register of large generator connections (nameplate rating ≥ 30 MW) and large bidirectional units (> 5 MW) under clause 5.18A.2, but there is no equivalent register for large loads.
Request: The proposed change would extend the clause 5.18A.2 requirement to loads over a specified size, requiring both distributors and TNSPs to provide information on all prospective and existing loads with a rated capacity of 5 MW or more to a new public register hosted by AEMO.
Impacts: The information must include the proponent (and where acting on behalf of another, the other party), the site, site owner, the requested connection capacity and voltage level, the bulk supply point, the ramping rate and timeframe, connection process status and its network planning status eg proposed, anticipated, committed or operational.
Gap 5: Barriers to the Use of Funded Augmentations Description
Gap: Rule 5.18 permits TNSPs to negotiate with customers to deliver shared transmission upgrades as "funded augmentations," whereby the customer fully covers design, construction, and ongoing costs, but this mechanism has rarely been used. The rule is limited in scope and primarily concerned with notice to other parties. There is also a "first-mover" disadvantage - the first connection does not get priority access and may pay for an asset subsequently used by later connections.
Request: The proposed change would amend rule 5.18 among others to ensure it can apply to data centres funding transmission augmentations and resolve the first-mover issue – potentially through a funding pool mechanism or a "pioneer scheme" approach analogous to that used in distribution, which allows refund of part of the cost of a dedicated distribution connection asset where it is subsequently used by other customers within 7 years.
Impacts: Unlike the request for gap 1, it appears that the funded augmentation would not be mandatory for data centres but a matter of negotiation. Data centres who fund an augmentation would be compensated if future users “use” the funded augmentation.
Gap 6: Jurisdictional scheme costs can only be recovered from distribution-connected customers
Gap: Clause 6.18.7A of the NER enables the cost of jurisdictional schemes (such as the NSW Roadmap and NSW Climate Change Fund) to be recovered via distributors, but there is no corresponding provision for TNSPs. This means large transmission-connected customers do not contribute to funding these schemes.
Request: The proposed change would introduce a new rule in Chapter 6A, analogous to clause 6.18.7A, to enable pass-through of jurisdictional scheme amounts to TNSPs and their customers.
A new transmission service charge category may be required. This would allow jurisdictions to allocate costs to particular classes of connections (eg. distributors would be excluded to stop double charging) and for new tariff structures, in particular capacity based charges. The request also suggests the new charging category should apply to both prescribed and negotiated services to prevent a data centre avoiding the charge by negotiating a negotiated transmission service (eg. batteries do not pay TUOS charges as they are on a negotiated service).
Impacts: Transmission connected data centres will likely be required to bear the costs of jurisdictional schemes which are currently recovered from distributors based on their contracted demand rather than consumption. This includes schemes like the NSW Roadmap and NSW Climate Change Fund. It will not be possible to avoid these costs by negotiating bespoke services.
Gap 7: Guidelines for transmission connection policies and charges description
Gap: Clause 5A.E.1 and clause 5A.E.3 of the NER provide connection charge principles and require the AER to develop connection charge guidelines for DNSPs, but there is no consolidated guidance for transmission connection pricing. Distribution guidelines require distributors to use unit rates to calculate and allocate the costs of shared network assets to new customers. For example, in distribution, there is an incremental cost-revenue test included in the AER Connection Charge Guidelines. One of the components is an “Incremental Cost Shared Network costs” being costs incurred by the DNSP for standard control connection services which are not solely used by the connection applicant (ie. any augmentation attributable to the new connection) which is charged at AER approved rates.
In contrast, transmission connection contributions depend on negotiation between the parties, subject only to the bounds of the negotiating principles in Schedule 5.11 which are very high level.
Request: The proposed change would require the AER to publish a guideline for transmission connection pricing similar to the distribution connection charge guideline, or amend Schedule 5.11 (noting Victoria's unique framework may require updating the AER's negotiated transmission service criteria). The guideline should consider how transmission unit rates (eg $/MW or $/km) can account for changes in estimated and actual costs over time, given that actual major transmission project costs have far exceeded initial estimates.
Impacts: Transmission connected data centres will likely be charged for the shared transmission network according to a specific formula. These charges may change over time to reflect the actual cost of the transmission augmentations.
