Comparative analysis of costs and economic returns when using traditional and digital training formats for personnel of gas producing enterprises

Abstract

The article examines the economic feasibility of transitioning from traditional personnel training formats to digital solutions in gas production enterprises by comparing total costs and measurable production effect based on a representative empirical dataset. The aim of the study is to substantiate an optimal investment model for training, based on a comprehensive accounting of direct and indirect costs as well as labor performance indicators. Materials and methods include the analysis of data from three large enterprises (2021-2023), more than 2,500 trained employees, the formation of two groups (traditional format, n=210; blended with a predominance of digital technologies, n=245), decomposition of costs into capital and operational, examination of financial and production documentation, calculation of ROI over a three-year horizon, and statistical processing (comparative and correlation-regression analysis). The obtained results show fundamental differences in the structure and level of costs per employee: direct – 114,10 thousand rubles (traditional) versus 38,60 thousand rubles (digital); indirect – due to time away from production – 79,52 thousand rubles versus 9,88 thousand rubles; total cost – 193,62 thousand rubles versus 48,48 thousand rubles. The pedagogical and operational efficiency of the digital format is higher: reduction of operational errors by 31,2% versus 18,6%, reduction of time to reach target productivity by 9,8 days versus 5,4, final certification score 92,7 versus 81,3. The integrated economic evaluation over three years shows NPV of −126,5 thousand rubles and ROI of – 36,98% for the traditional approach, compared with NPV of 422,8 thousand rubles and ROI of 643,53% for the digital approach, indicating multiple superiority of digital solutions in efficiency and scalability. The discussion emphasizes the scale effect, the key role of VR/simulators (strong positive correlation with reduction of incidents), the robustness of advantages under variability of model parameters, and the practical significance for geographically distributed companies; limitations (heterogeneity of content, need for updates) are noted, and directions for further research (predictive performance models, integration with production data) are outlined.