Mark Gongloff: Data centers are stewing in their own pollution
This article discusses how data centers, while contributing to climate change, are also increasingly vulnerable to its effects, such as extreme weather. A study by XDI Pty Ltd. highlights that many planned and existing data centers face high risks of damage, leading to rising insurance costs. The author advocates for building resilient data centers and implementing careful regulation to mitigate their environmental footprint and adapt to a hotter climate.
The Daily Gazette columnist Mark Gongloff argues that data centers, while significant contributors to climate change through their energy consumption, are also increasingly vulnerable to its severe impacts. A new study from risk-analytics firm XDI Pty Ltd. reveals that approximately 6% of nearly 3,000 planned global data centers are in locations immediately susceptible to high damage risk from extreme weather, potentially making them uninsurable without significant hardening. This follows an earlier XDI report estimating similar risks for existing data centers by 2050, with an additional 20% facing moderate risk and higher insurance costs.
The article notes a widespread public opposition to data centers in the U.S., with a Heatmap poll indicating 71% of Americans are at least somewhat opposed to having one nearby. Despite this, the imperative for artificial intelligence (AI) is driving a rapid expansion of data center construction, with an estimated $7 trillion investment by 2030, according to McKinsey. Gongloff cites examples like Nvidia CEO Jensen Huang and Ford Motor Co.'s AI initiatives to illustrate the economic reliance on this boom.
However, XDI founder Karl Mallon emphasizes the urgent need for a "climate ready" data ecosystem, as current preparations are insufficient. While the global data shows specific "hot spots" like China's Jiangsu province, the U.S. also faces significant risks, with 20% of New Jersey data centers in flood zones and half of planned U.S. data centers in areas prone to severe convective storms, according to a reinsurance firm MS Amlin. Extreme heat and indirect impacts like infrastructure disruption further compound these challenges, making the financial toll of climate disasters potentially tenfold worse.
Gongloff concludes that data center operators must reduce their substantial environmental footprints, as their high energy demand, reliance on fossil fuels, and water usage in already-strapped regions exacerbate climate risks. He acknowledges AI's potential in climate adaptation but stresses that its development must be thoughtful and carefully regulated to ensure its benefits outweigh its environmental drawbacks.