最新研究显示,为满足人工智能数据中心的巨大电力需求并规避电网接入延迟,开发者计划在美国、英国及欧洲大陆大规模建设现场燃气发电设施 1。据BloombergNEF发布的研究报告指出,这些拟建项目将创造高达181吉瓦(GW)的天然气产能,其中美国占据了计划中现场燃气电站总量的90% 1。预计这些设施每年将排放约3.45亿吨二氧化碳当量,这一数值相当于近四倍英国峰值电力需求所产生的碳排放,或约等于7280万辆化石燃料汽车的年排放量 1。
随着人工智能技术的扩张,相关能源消耗正在急剧攀升。报告预测,到2035年,全球AI数据中心的电力需求将几乎翻两番,达到2354太瓦时,届时将占全球电力总需求的7% 1。为了缓解电力供应压力,英国财政大臣John Healey已批准数据中心在等待正式电网连接期间建造和运行临时的现场化石燃料发电站 1。具体而言,包括Wapseys Wood和Quest Park在内的多个英国项目因面临漫长的电网接入等待期,正积极寻求通过自建现场燃气电站来解决供电问题 1。
尽管行业方声称自建电厂仅是过渡性的临时措施,但专家警告此举可能带来严重的长期后果。这种模式可能导致“化石燃料锁定效应”,即基础设施一旦建成便难以转向清洁能源,从而加剧气候变化风险 1。
New research reveals that developers of artificial intelligence data centres are planning to construct on-site gas power stations on a huge scale in the United States, the United Kingdom, and continental Europe 1. According to BloombergNEF, these planned facilities aim to create 181GW of natural gas capacity 1. The initiative is driven by the need to satisfy the enormous electricity demands of AI operations while circumventing delays in grid connection 1.
The environmental implications of this strategy are significant, with the projected infrastructure expected to emit approximately 345 million metric tons of carbon dioxide equivalent annually 1. This volume of emissions is comparable to the annual output of nearly 72.8 million fossil fuel-powered cars 1. Although the developers describe these power stations as temporary measures, experts warn that such projects could result in long-term lock-in effects for fossil fuels, thereby exacerbating climate risks 1.
Geographically, the United States accounts for 90% of the planned on-site gas power stations 1. In the UK, Chancellor John Healey has approved regulations allowing data centres to build and operate temporary on-site fossil fuel generators while waiting for grid connections 1. Specific British projects, including Wapseys Wood and Quest Park, have sought self-built gas plants due to prolonged wait times for grid access 1. By 2035, global electricity demand from AI data centres is forecast to almost quadruple to 2,354 terawatt-hours, constituting 7% of total worldwide power consumption 1.
评论
还没有评论,欢迎留下第一条。