Prediction of H2S Concentration Around Geothermal Power Plants Using Multiple Regression Analysis
- School of Civil Engineering, Department of Environmental Engineering, Istanbul Technical University, Istanbul, Turkey (aydinmer21@itu.edu.tr,burcak.kaynak@itu.edu.tr)
Investments in geothermal resources are growing worldwide with a view of being clean and sustainable, however it may also have negative impacts on air quality, ecosystem and health. Geothermal resources are used in different areas such as power generation and direct use including heating, thermal use, greenhouse and drying activities. The electricity production from geothermal power plants (GPPs) is closely associated with hydrogen sulfide (H2S) emissions negatively affecting ecosystems at certain concentrations and exposure. H2S is oxidized to SO2 after released to the atmosphere at a rate depending on temperature, sunlight and radicals.
Turkey has been ranked 4th worldwide in terms of electricity generation from GPPs with a total capacity of 1676 MW in 2021. With recent legal restrictions about GPPs, additional H2S measurements were started recently along with criteria air pollutants at selected air quality monitoring stations (AQMSs) in regions with GPPs in Turkey. Our preliminary result showed a significant correlation via exploratory data analysis between H2S and SO2 measurements in 2021 from one of these AQMSs. The wind speed and direction analysis showed these air pollutants were transported from the same directions coinciding with GPP locations. This study aims to analyze the relationship between H2S and SO2 in GPP regions in southwestern Turkey. The study area focuses on four regions specified according to GPP locations and H2S measuring AQMS locations along with SO2 measurements. Time period includes 2021-22 with ground-based H2S, ground and satellite-based SO2, ground-based meteorology measurements as well as other related parameters such as topography, GPP locations and capacities. There are peaks observed in H2S concentrations around noon at all seasons for three regions, at similar times SO2 concentrations usually peak as well. The Pearson correlations (R) between daily H2S and SO2 measurements are 0.76, 0.60, 0.46 and 0.42 for four regions. Correlations between H2S and SO2 measurements at lag times using 1-hr and 6-hr moving averages showed higher correlations with 6-hr moving averages indicating H2S to SO2 conversion. SO2 satellite retrievals are also investigated around these regions on these days when H2S was the highest. These findings strengthen our hypothesis of SO2 in the region being from the oxidation of H2S released from GPPs.
A linear model based on multiple regression analysis is developed using H2S as a dependent variable and other parameters as independent variables for understanding the levels of H2S in the region. Principal component analysis (PCA) approach is used to understand the importance and contribution of the selected parameters. This model will be used to predict H2S concentrations, because H2S measurements are not continuous and mandatory in the whole region. Moreover, the spatial distribution of H2S will be investigated for the region to understand the negative impacts on human health, ecosystems, and agriculture.
Keywords: Geothermal Power Plants, H2S, MRA
How to cite: Aydin, M. and Kaynak, B.: Prediction of H2S Concentration Around Geothermal Power Plants Using Multiple Regression Analysis, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-14483, https://doi.org/10.5194/egusphere-egu23-14483, 2023.
