One of the common man-made causes of earthquakes is the operation of geothermal power plants. Normally, their magnitude does not exceed 2 points, but there are exceptions. Thus, the magnitude of the most powerful earthquake in Europe, caused precisely by the operation of the station, reached 3.4. And in South Korea in 2017, when the geothermal power plant was launched, it shook up to 5.4!
The technology of operation of the stations is based on the use of two wells: cold water is pumped into one, and boiling water rises from the other outside from underground sources. Sometimes, in the process of pumping fluid into wells, small earthquakes are observed in the vicinity. Scientists found out new details and once again clarified the reasons for such phenomena.
Experts from Columbia University (New York) to study such earthquakes have proposed using machine learning algorithms (one of the artificial intelligence methods for which mathematical statistics, numerical methods, optimization methods, probability theory, and techniques for working with data in digital form are used). They analyzed for 3 years (2012 - 2014) seismic data in a geothermal field, which is located near San Francisco. There are 22 stations and more than 350 wells in the field. During this time, geophysicists tracked 46,000 weak tremors with magnitudes ranging from 0.3 to 1.5, Science Advances reports. Signals were later used for machine learning algorithms.
Having gone through the whole complex process - from processing each seismic signal to drawing up general patterns of the occurrence and development of earthquakes, the experts ultimately identified 4 groups of earthquakes. Two groups were dependent on seasonal changes in the volume of water injected into the well: the peak of stronger earthquakes occurred in the winter and the maximum volume of injected water, and vice versa, the least shocks were observed in the warm season with a minimum volume of water injection. The third group of earthquakes was noted when the region was abnormally dry weather. The rocks reacted perceptibly to the rate of water injection - this is the reason for the fourth type of earthquake.
The results of the study turned out to be an invaluable tool for correcting the operation of geothermal stations. The process of pumping water into the wells of water stations must be clearly regulated. Scientists suggest using machine learning methods not only for describing and explaining earthquakes, but also for predicting them.