Hot spring binary cycle power generation is one of the important renewable energy system, and the effective utilization of surplus hot spring water has been demanded in Japan. However, the stable operation of these facilities is difficult because the adhesion scales of calcium carbonate and silica/silicate cause pipe blockage and hindrance to heat exchange. Though present practical inhibition technique is an injection of chemicals to hot spring water, there is a demand for scale countermeasures that are more environmentally benign. The material surface treatment or modification is expected to be able to suppress the formation of scales without causing an environmental harm. However, there are few s reports about such materials because it is due to the lack of knowledge about the microstructure of the scales formed in actual geothermal hot water environments. In this study, we analyzed microstructure to understand the initial scaling process in the hot spring water. Specimens were immersed in hot spring water from Honda-Yudayuu source (Obama town, Unzen city, Nagasaki pref., Japan). Immersion sites were the aeration tank and the drain after it. Austenitic stainless steel (Type 316) sheets were used as specimens. To observe time variation of microstructure, specimens were immersed for 3 days. SEM with an EDS detector was used to characterize the microstructure of the scales. The scales of calcium carbonate and magnesium silicate were confirmed in both the tank and the drain. Saturation index (SI) of the hot spring water gushed from Honda-Yudayuu source was analyzed. In the result, calcium carbonate and magnesium silicate showed super saturation. The formed scale species were in accordance with the scales predicted by thermodynamic equilibrium calculation. In all immersion time, calcium carbonate and magnesium silicate were observed. The amount of calcium carbonate didn’t change, and there are a few particles on the sample. On the other hand, the amount of magnesium silicate increased with duration, and magnesium silicate overall formed on the samples immersed for 3 days. Therefore, initially formed scale is not calcium carbonate but magnesium silicate. There are two types of the precipitation morphology of magnesium silicate. One type is particle, and another is net-like appearance. In the drain, only particulate magnesium silicates were observed. In the aeration tank, not only particulate magnesium silicate but also net-like one were observed after the immersion time of 1 day. The particulate magnesium silicates grew by the aggregation of particles, and net-like magnesium silicate grew with the formation covering on the surface of specimen. Adhesion mechanisms can be classified as either chemical interaction (oxidation reduction reaction and condensation one etc.) or physical interaction (van der Waals force and gravity force etc.). Generally, bonding force by chemical interaction is larger than that of physical one. Then, the adhesion strengths of particulate magnesium silicate and net-like one were evaluated. The adhesion strength of net-like magnesium silicate is larger than that of particulate magnesium silicate. Particulate magnesium silicate is considered to adhere by the physical interaction, and net-like magnesium silicate may adhere by chemical interaction.