① Media has discovered that more than half of the households experiencing the most severe "Electrical Utilities distortion" (Harmonic distortion) are located within 20 miles of a major Datacenter. ② The media pointed out that this "distorted" electrical supply could ultimately damage appliances, increase vulnerability to electrical fires, and even lead to power restrictions and outages.
On December 30, Caixin reported (Editor Huang Junzhi) that as the AI boom propels the US stock market and related fields to soar this year, investors are also very enthusiastic. However, looking forward to the future of AI, there are still mixed feelings. A recent report by the media stated that in order to meet the computing demands of AI, Datacenters in the USA are rapidly increasing, which could be bad news for the local electrical grid.
Using 1 million Residence Sensors tracked by Whisker Labs and market intelligence data from DC Byte, media discovered that more than half of the households experiencing the most severe "Electrical Utilities distortion" (Harmonic distortion) are located within 20 miles of a major Datacenter.
In other words, there seems to be a correlation between the distance to the Datacenter and the "bad harmonics." It is understood that "harmonics" refers to a distortion of the normal current waveform, where, in addition to the fundamental frequency, there exist various non-integer multiples of the voltage or current components at frequencies higher than the fundamental frequency. Bad harmonics can affect the normal operation of electrical utilities and even cause equipment damage.
The media pointed out that this "distorted" electrical supply could ultimately damage connected appliances, increase vulnerabilities to electrical fires, and even lead to power restrictions and outages. The issues surrounding AI Datacenters could be more significant because their energy demands are inherently unstable.
Aman Joshi, Chief Business Officer of Bloomberg Energy, said: "No electrical grid is designed to handle load fluctuations from multiple Datacenters simultaneously."
Concerns about the collapse of the US electrical grid have long existed.
The fact is that as the rise of AI triggers a surge in electrical utilities demand, worries about whether the "US electrical grid will collapse" have been around for a long time.
Firstly, to meet the intensive computing needs of AI, datacenters will consume more and more Energy. According to data from the US Department of Energy and the Census Bureau, a datacenter with peak demand of 1 gigawatt consumes power equivalent to the annual electricity usage of approximately 0.7 million households, or the total electricity consumption of a city with a population of about 1.8 million.
Arm's CEO Rene Haas has predicted that by 2030, the power consumption of AI datacenters may account for 20%-25% of US Electrical Utilities demand. Mizuho Securities projected in August that by 2030, computing centers powering the Internet Plus-Related and AI may require up to 400 terawatt-hours of power. According to data from the International Energy Agency, this exceeds the total generation capacity of the United Kingdom in 2022.
Earlier this month, the North American Electric Reliability Corporation (NERC), the US Electrical Utilities regulator, also warned that the North American grid faces "serious reliability challenges" as generation cannot keep up with the growing demand for AI. Simply put, more than half of North America will face the risk of Energy shortages in the next decade, potentially leading to power outages in the USA.
In fact, the reliability of the US grid is already in question, to the point where it can be said to be "overwhelmed."
Data shows that the average lifespan of a transformer in our normal grid is 25 years, while statistics show that 70% of the Electrical Utilities access and distribution facilities in the USA are aging. According to a report from the US Department of Commerce, the average lifespan of transformers in the USA has exceeded their expected lifespans, reaching 30 to 40 years.
In 2020, California faced power shortages due to extreme heat waves, leading to many households losing their power supply. This marked the first large-scale rolling blackouts in the USA in nearly 20 years. In the following years, incidents of power shortages triggered by extreme weather became more frequent, with the latest occurring in the winter of 2022, when consecutive snowstorms caused power outages in multiple states, and grid operators in the Southeast and Midwest also issued warnings of possible rolling blackouts.
Editor/lambor
