Map showing highlighted areas of the Smackover Formation and the sampling area. (Image Credit: USGS)
The USGS used water testing and machine learning to discover approximately 5 to 19 million tons of lithium beneath southwestern Arkansas from the Smackover Formation. This finding could supply sufficient lithium for EV batteries several times over on a global scale by 2030. Study results have major implications for US-based lithium production and the worldwide shift to clean energy.
The Smackover Formation is beneath Arkansas, Louisiana, and other southern U.S. states. It's well-known for the large volume of oil bromine deposits. Research has pointed to it as a potential site for lithium extraction from brines.
According to the study, U.S. lithium imports could eventually be replaced by lithium deposits in the southwestern region of the Smackover Formation. The International Energy Agency's data suggest that five million tons of lithium has the potential to fulfill the estimated demand for lithium in EV batteries in 2030 nine times over. Researchers involved in the study say these values aren't indicative of what we can extract using current tech and depict untapped lithium reserves.
"Our research was able to estimate total lithium present in the southwestern portion of the Smackover in Arkansas for the first time. We estimate there is enough dissolved lithium present in that region to replace U.S. imports of lithium and more. It is important to caution that these estimates are an in-place assessment. We have not estimated what is technically recoverable based on newer methods to extract lithium from brines," said Katherine Knierim, the study's principal researcher and USGS hydrologist. New lithium extraction techniques are essential to estimate the percentage of these reserves that may be commercially feasible to develop.
This is the first time AI estimated lithium concentrations across the Smackover Formation brines. USGS Brine Research Instrumentation and Experimental lab analyzed the samples and compared them with the USGS Produced Water Database. Combining lithium concentrations in brines with geological data enabled the AI model to predict lithium concentrations in regions that didn't have any extracted samples. By using this predictive model, lithium resources can be mapped across a wide region.
"The USGS – and science as well-- works best as a partnership, and this important research was possible because of our strong partnership with the Office of the Arkansas State Geologist," said Knierim.
KoBold Metals, a tech-driven mineral exploration startup, also uses AI to find critical minerals, including lithium. The company discovered potential lithium deposits in South Korea, Australia, Namibia, Quebec, and Nevada. The AI tool integrates geochemical, geophysical, and satellite data to produce predictive models. Using these techniques accelerates discoveries and improves success rates compared to standard exploration methods.
Have a story tip? Message me at: http://twitter.com/Cabe_Atwell