KCEC using AI tools to strengthen wildfire mitigation

Tuesday (June 23) -- The Kit Carson Electric Cooperative (KCEC) is hoping to reassure the public of the "proactive measures" the utility company is taking to help reduce the threat of wildfires throughout its service territory, as stated in a press release.

The release goes on to explain some of these "proactive measures" in the form of two specific AI tools: AI Dash and Firescape.

KCEC described AI Dash as a real-time grid-monitoring platform that analyzes system performance, weather conditions, and equipment status. The utility Co-op described Firescape as a real-time wildfire mitigation tool that incorporates weather forecasts, fire behavior modeling, and satellite imagery. According to KCEC, the tools enhance situational awareness and help maintain the cooperative's electrical system.

"KCEC is committed to doing everything possible to reduce wildfire risk and protect the communities we serve," said Luis A. Reyes Jr., CEO of KCEC. "Many wildfires are caused by lightning and other natural occurrences that are beyond our control, but the more we can do together as a community, the better chance we have of preventing human-caused fires and minimizing their impact."

KCEC hopes everyone remembers fire safety through the Independence Day weekend.

The growing use of artificial intelligence has also raised questions about the environmental impacts of the technology, particularly the energy and water demands associated with large-scale data centers.

AI tools rely on data centers -- storage buildings specifically for housing and cooling computer servers -- to operate. With rising usage of AI applications like ChatGPT and Grok, tech companies are relying more on data centers, which consume water as a coolant. According to industry standards, a mid-sized data center (20-50 megawatts of electricity) uses about as much water as a small town.

The issue has local relevance for New Mexico. The proposed Project Jupiter in Las Cruces, is estimated to cost $165 billion and will include four individual data centers.

Researchers at the University of California, Riverside, have estimated that interactions with AI applications can indirectly consume water through data-center cooling and electricity generation. One widely cited estimate suggests that a series of dozens of ChatGPT prompts may be associated with approximately 500 milliliters of water use, though the amount can vary depending on the data center and cooling systems involved.

This reporter contacted CEO Reyes asking if KCEC used any water cooling systems for their computing servers.

"We use air-conditioning to cool all our servers," Reyes responded.

This reporter also asked Reyes about AI's broader context around water use.

"Does the environmental cost of AI infrastructure undermine some of the benefits of AI-assisted wildfire mitigation?"

"I think that’s the real dilemma," replied Reyes. "Many people that I’m dealing with on this hydrogen issue are using AI to formulate questions and responses, yet complain about the water that the hydrogen uses. AI from data we have probably uses 20 times more for an equivalent sized data center of 20 MW. The AI issue needs to be addressed, perhaps it will be centered around cooling systems. I believe AI is here to stay, so we have to figure out what the win-win solution is."