Data centres are power hungry – but they don’t have to be a burden on NZ’s grid
AI data centres are contentious for consuming large resources. But with the right incentives and regulation, can they make for more flexible electricity systems?
Data centres, known for their significant power consumption, are not merely a burden for New Zealand's electrical grid. They have the potential to serve as assets, especially with the increasing demand for Artificial Intelligence (AI) in the global market. Datagrid, a Singapore-based company, has recently secured approval for a NZ$3.5 billion data centre near Invercargill, which will double the country’s data processing capacity once operational in 2028.
This development, however, has sparked concern over its impact on the grid, especially since it will become the second-largest electricity user in the country.
Governments worldwide, including New Zealand, are grappling with the economic advantages of such centres against the environmental and grid pressures they bring. In Ireland, data centres already account for nearly a quarter of the country's metered electricity consumption, raising questions about future grid capacity. In New Zealand, electricity shortages have already occurred during dry years with low hydro lake levels, making it vital to explore how data centres can be integrated into the grid rather than viewed as a threat.
Data centres could potentially become valuable to the grid by adjusting when and how they use electricity. Since many computational tasks in these centres are not time-critical, they could schedule their operations for periods when renewable electricity is abundant, prices are low, or the grid is under less strain. This approach could help absorb surplus wind and solar energy that would otherwise go unused, making new renewable projects more financially viable.
Moreover, data centres already possess sophisticated electrical infrastructure like uninterruptible power supply (UPS) systems, batteries, and backup power systems. These assets could stabilize the grid by providing fast frequency response and reserve capacity. With appropriate market rules, they could temporarily reduce demand during network congestion.
The existing battery and UPS systems within data centres are technically capable of providing these services, but market arrangements need to evolve to enable them to do so more widely.
Additionally, data centres' large and predictable electricity demand could support more renewable generation. Long-term power purchase agreements could improve the financial viability of wind and solar projects, particularly smaller community-owned ones that often struggle to secure finance. Datagrid’s existing power purchase agreement with Mercury covers 140MW of its electricity demand, leaving potential for the remaining demand to support new community-owned renewable energy projects in Southland.
However, not all data centres are designed to serve as grid assets. Poor planning can lead to increased peak demand, costly transmission and distribution upgrades, worsened local network congestion, and competition with other users for renewable electricity. Therefore, the key to turning data centres from a potential grid burden into a solution lies in their design and the market incentives they receive.
In conclusion, data centres can indeed be grid assets if designed and operated appropriately. By shifting their energy consumption to times when renewable energy is most available, they can help stabilize the grid, support renewable energy projects, and enhance grid resilience. As more data centres like Datagrid’s Southland development come online, New Zealand’s strategy should focus on integrating these facilities as grid assets rather than viewing them as grid constraints.
Written by urgent.news from The Conversation AU's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.