Data centers: Powering Australia’s AI boom, part 2

Alexandra Peace
Hello and welcome to Legal Outlook by Ashurst Perkins Coie.

You're listening to Here Today, Here Tomorrow, our podcast mini-series about data centers. This is the second in a two-part episode exploring the Australian market.
I'm Alexandra Peace, a real estate partner specializing in data center acquisition and development, and I'm joined today by Kate Muller and Mike Webb, partners in our projects practice. Kate and Mike advise across the full life cycle of energy and infrastructure projects, including data center projects. Today, we're going to cover energy procurement and risks for data center developers and operators in Australia.

So let's jump in, Kate, the government has said that data centers must secure new and additional clean energy generation and/or storage to offset demand. What does that practical procurement menu look like?

Kate Muller
So, “securing new and additional clean energy” – those are some quite powerful words that still lack definition in terms of what's been released, but I think the expectation is quite clear.

One thing that I am always at pains to point out is data center operators in Australia are already very active procurers of renewable energy. Data Centers Australia has a number of members, most of whom are committed to delivering on green energy commitments and really supporting the local renewable energy build-out that's underway.

So for the current data center load in Australia, about 40% of that power is coming from renewables already. That's either co-located or through virtual power purchase agreement.

Data center operators obviously have very high energy needs, and managing that energy load is sort of a twofold question:

• One is the physical electrons; so actually securing the connection and getting the energy into the data center.
• The other part of the equation is about what does that cost?
• (There's actually a third question, which is how green is that power, and does that matter?

So the way that you go about that energy procurement is usually a multi-pronged approach. You'll likely need a retail contract, which essentially gives you the electrons at your connection point. You might have what's called a “sleeved PPA”, which is a deal with a retailer and with some other renewable energy generation supporting that supply arrangement and giving you some green credentials as well. You can have a corporate PPA or a virtual PPA. Again, that doesn't give you your physical electrons, but it does give you some cost certainty and some green attributes, and you may also have a very active wholesale energy markets trading strategy where there's part of your load that may be exposed to market pricing.

So it's often a combination of those things, and obviously data center operators are sophisticated procurers of power. They know that they need power, and they know that there's these various models to get to get that power.

So, in terms of additionality, and Mike has alluded to this earlier, there are a very large number of proponents in the electricity sector that have projects under development that are seeking offtake, that would like to have contracts – nice long-term contracts – to offtake from their solar hybrids or their wind farms or their wind hybrids.

So there's certainly a market for data center operators to engage in and secure contracts for power and to give them that pricing hedge that they're looking for.

There is still that whole of system piece though that we do need to think about, which is the sum total of demand combined with what's actually existing in the system at any given time, and that's where obviously alignment of what's happening in the energy transition and the forecasting work that the market operator is doing, and how that feeds into investment decisions and prioritization, becomes a conversation that has to include this data center piece as well.

Alexandra Peace
Mike, can I ask what contractual evidence you think will become the market standard to providing additionality?

Mike Webb 
We've already got a reasonably good sort of certification system in place. So, for almost a decade now, we've had the Renewable Energy Target Scheme, which produces certificates for each megawatt of renewable energy, and that's essentially how you know any entity evidences its procurement of renewable energy.
That structure and architecture is already in place. That particular scheme is rolling off in 2030 and is being replaced by a new certificate-based scheme called the Renewable Energy Guarantee of Origin (REGO). So you'll have REGO certificates – evidencing that REGO certificates are actually more versatile and going to the granularity of exactly where that power was produced and at what time and and each day. So they they'll allow data centers to evidence that quite well.

Alexandra Peace
Looking at the NEM curtailment of utility scale wind and solar, it hit a record seven terawatt hours in 2025. Do you think data centers could actually be part of the solution, soaking up surplus daytime solar, or do you think they mainly worsen network headroom constraints?

Mike Webb
I think so. As you say, curtailment levels were at an all-time high last year – 60% up from the year before. Obviously a lot of that gets accounted for by solar. So, you know, in one sense, you would think a nice large amount of load during the middle of the day will soak up that spilled energy.

The problem is the data center load is not particularly favorable for that, so it's fairly flat. It's a large flat load, and so while it does soak up that energy during the middle of the day, it also sort of increases the demand during the peak hours, where there's potentially a sort of shortage in supply.

Things that could help with the situation are batteries. Whether they're co-located or contracted with the data center, they offer an ability to soak up some of that cheaper energy in the middle of the day and supply it to a data center during the peak year times, that could work where you've got a more flexible type of load.

And my understanding is that the data centers are introducing flexibility. So, not all of them have that large static load, that really sort of depends on the ultimate use of the data center. Where it's being used potentially by a hyperscaler, it might still be a little bit flat. But the smaller, more flexible use could allow for an ability to soak that daytime energy up and apply it at night.

Kate Muller
This conversation starts weaving us towards another key theme and another area of interest for data center operators, which is what degree of self-reliance do they need for their system security and their power security? What does that sort of backup or redundancy look like? What technologies can be deployed to support that?

And a thought I had in preparing for today was “gas peakers” at the moment are really difficult assets to stand up from a feasibility perspective. They have a very low capacity factor because they're not needed for significant tracts of the year, and batteries are compressing some of the peak times that they used to really be able to drive their returns from.
Perhaps interestingly, the business case for gas peakers with the data center demands and requirements will find a new place for this gas peaking support. And there's certainly a business case to see some existing infrastructure supporting these projects in the near term.

But it does also then start the new connection standards (along with the need for data center operators to have some embedded capacity or ability to respond to their needs in a self-service kind of way) and it does continue to feed into that thought about data centers really being very active loads, and the fact that data centers and data center operators may find themselves participating in our energy system in ways they haven't done before, through things like demand response system strength and FCAs, for example, which are revenue streams that haven't really been explored by data center operators before.

So data center operators are building out their energy teams, their trading teams, and starting to think about their assets as a multi-tool. They can obviously derive the revenues they're used to from providing their customers with services, but they may also be able to see revenue through participation in the electricity market in a way they haven't done before.

Alex
So shifting from the grid to behind-the-meter and co-location play, we're seeing more projects that combine data centers with co-located solar vest and backup generation. At what point, in your view, does this shift from a resilience play to a core commercial strategy, and what does that look like in practice?

Kate
This is still a nascent area, really, in Australia. As we've said before, historically, data centers were able to quite readily secure their connections, firmed connections for their capacity, and didn't feel the same pressures that are emerging now through the increased interest and pipeline of data center development. And all of that going on through the lens of an energy transition with substantial shifts in the way energy is generated, stored, and transported around our network. So what we are seeing is more pressure on data center operators, not only through the government expectations framework, but just from a genuine operations perspective of there being a need – a resilience need – to be able to support their own energy needs, or have sufficient backup or redundancy to continue their BAU operations in a world where their customers expect 99.99% uptime, having downtime that could be alleviated through self-serve behind-the-meter solutions is clearly the right direction of travel.

But those behind-the-meter batteries (or co-located batteries, or other generation and storage resources) are not only limited to serving the data center, but can actually perform roles for our energy market as a whole. They can provide backup power. They can be demand response engines. They can participate in the provision of grid services.

And in doing that, data center operators may well find themselves also being active energy market participants and providers of energy services in a way they haven't done before. So that creates new business models. It also requires new capability and competencies within those businesses in order to do that themselves.
But it may also lead to a higher incidence of joint venturing or joint development, where data center operators partner with energy sector participants to optimize what they can each bring to those infrastructure plays, feed into their strengths and benefit each other and the market in the process.

Mike Webb
Yeah, fully agree, I’d just add, there is a bit of complexity to the behind-the-meter co-location when it comes to if it's a battery or a gas-fired power plant or whatever the resources you typically expect those to be sort of financed equity and debt separately.

And particularly when it comes to that generation asset, things like what happens if the data center customer falls away? How do we export our power? You know, project-on-project risk structuring pieces around that obviously need to be thought through.

So, I think in the first instance, data centers will be sort of looking for “in front of the meter” power supply, but for all of the reasons and a lot of the issues we've already discussed, they may well need some form of backup supply behind the meter. And so joint ventures or s joint developments with those assets are an option.

Alexandra Peace
So, to that point, when we're looking at site selection and to make a project bankable and deliverable, are you seeing more interest in greenfield precinct scale developments or brownfield retro-fits?

Mike Webb
There's pros and cons to both of them. Greenfield developments we're probably seeing a little bit more of, particularly in the press. That's where you're hearing of the sort of 800 megawatts (or whatever it is) type developments, largely because those greenfield sites allow for that type of scale.

So they're new, and you are expecting to build this new network around you. But that's not to say brownfield sites don't have their place. They're potentially a lot quicker to power. There's already a connection in place. It just might be smaller, and you might not be able to stage that. A brownfield site would also be cheaper potentially, obviously without the network augmentation that's required.

Kate Muller
With our brownfield fleet of data centers, technology and customer demands are shifting. So quickly that there's a number of data centers now (you'd call them older generation of sorts, they're smaller capacity connections, largely connecting in at a distribution level, and even using the existing site footprint) but changing out the technology still would typically require an increased power load and an increased connection capacity.

Then you may still be looking at triggering augmentation for those sites. However, the benefit of those brownfield plays is they are established operators within the market. They've got a track record. They've got a customer base. So when we think about it from an investment risk or a bankability risk perspective, they're the types of players that people are comfortable transacting with.
And so in some ways the brownfield – while complex in their own ways – have other advantages, whereas greenfield have other challenges but other advantages.

So in the data center boom we're hearing a lot about greenfield, but there is equally an incredible opportunity in that brownfield space. And a lot of the existing operators will be looking at their strategy and working out where they're best to invest and what that footprint means for them.

Mike Webb
And again, when we talk brownfield, that could be “brown” existing data center sites, but also other industrial complexes where perhaps whatever the factory or industry load has moved on. So absolutely we're seeing really sort of interesting developments, kind of using that existing infrastructure and connection to bring in a data center.

Alexandra Peace
Thanks, Mike. Okay Kate, one final question for you: if you're advising a lender underwriting a data center project right now, what are the three energy-related risks you'd want to see addressed in the contract stack?

Kate Muller
So the first one, and Mike flagged this earlier: connection is absolutely a point of timing lag and bottleneck, and with regulatory reform on our radar as well, that's a very real issue that needs to be worked through and well understood.

The next one is just the role of data centers and how they exist within an already transforming electricity system, and the impact of things like curtailment.

And to the extent there's technical capability or the ability to look at more flexible ways of engaging with the electricity load, there's opportunity there to do things in different ways. But there's also a very real risk, particularly where customer contracts are so tightly wound to performance standards, and where for a lot of these new data center buildouts, it's the customer and the customer's credit that is underwriting those credit decisions.

And the final one is we're still working within a changing environment with respect to expectations around green attributes and how that sits alongside energy price and additionality, and so working through some of those issues from a sort of “future perspective” of what will the market require and how well adapted are these solutions? The power procurement solution, the duration of the offtake arrangements, the transformation that's happening in green attributes – lenders will be looking through that lens as well to understand what the longer-term strategy is for the assets as well. So those are a few of the key things I'd be looking at.

Alexandra Peace
Thank you Kate, and also thank you Mike. That brings us to the end of this two-part episode.

Thanks for listening to this episode of Here Today, Here Tomorrow, our Ashurst Perkins Coie podcast mini-series about data centers. You can find part one of this Australian discussion on Apple Podcasts, Spotify, or wherever you get your podcasts – and that's where you'll find the previous episodes in the data centers series too. Make sure you stay subscribed to this podcast so you don't miss any of our future episodes.

Until next time, thanks again for listening, and goodbye for now.