AI Data Centers Drive Surge in US Electricity Demand and Bill Increases
AI data centers are significantly impacting electricity demand and costs in the U.S. this summer.
Key Points
- • PJM reports a sharp increase in electricity demand due to AI data centers.
- • Electricity bills could rise 20%, with average increases of $25 or more per month.
- • Data centers consume up to 30 times more power than traditional centers.
- • Peak usage could exceed 154,000 megawatts, causing potential supply issues.
This summer, the U.S. is facing an unprecedented surge in electricity demand driven primarily by data centers that support artificial intelligence (AI) technologies. According to a report by PJM Interconnection, the largest power grid operator in the country, electricity demand is expected to reach peak levels, impacting electricity bills significantly.
PJM has noted that some regions could see electricity bills rise by up to 20%, translating to average increases of $25 or more per month for households within its service area. For commercial users, price hikes could approach 30% due to the soaring demand brought about by the deployment of AI systems such as ChatGPT, which necessitate massive data processing power. Data centers are now consuming as much as 30 times the power of traditional data centers, leading to a staggering projected increase in power costs by $9 billion between 2024 and 2025.
As the peak electricity usage is anticipated to exceed 154,000 megawatts this summer, PJM has warned that record levels might occur during extreme heat waves when demand spikes. The urgency is exacerbated by the retirement of fossil fuel plants, with over 9,000 megawatts of coal capacity set to be shut down or converted to gas by 2025. This wave of closures leaves the grid underprepared to cope with the heightened electricity need, especially as clean energy sources such as wind and solar are experiencing delays and increasing development costs.
To counter potential rolling blackouts, PJM is implementing demand response programs that encourage large businesses to cut down on electricity usage during peak consumption periods. However, if demand surpasses 166,000 megawatts, certain regions may be unable to guarantee sufficient reserve power, raising concerns about grid reliability. Even individuals who do not directly use AI technology are likely to feel the financial strain as the costs associated with the expanded energy demands are distributed among all grid users.