Jeongsik Oh

Asset management of power assets is a decision-making process that manages risk by considering asset condition and the consequences of failure while ensuring supply reliability and cost-effective maintenance within a limited budget. However, when maintenance plans are developed and budgets are allocated separately for each asset class, it is difficult to compare the investment effects across different types of assets and to establish an investment plan that reflects organization-wide asset management objectives. To address this limitation, the maintenance effects of heterogeneous assets must be evaluated on a common economic basis to prioritize investments, and the assets to be maintained and the timing of maintenance must be determined jointly within the available budget. This paper proposes a Deep Q-network-based integrated maintenance optimization method for heterogeneous assets, namely power transformers and gas-insulated switchgear, using risk and investment value assessments. The proposed method reflects condition changes caused by asset deterioration and maintenance, and it learns a maintenance policy that accounts for long-term investment effects under a shared budget constraint. In the simulation, the Benefit–Cost ratio is calculated from the present values of the benefits and costs estimated using synthetic data for power transformers and GIS, and it is incorporated into the Deep Q-Network algorithm as the investment value to derive a maintenance strategy. The proposed method is then compared with independent maintenance strategies for each asset class under the same total budget and evaluation period, and its applicability is verified in terms of maintenance cost, expected loss reduction benefit, and investment efficiency.

Conventional time-based inspection of power transformers does not reflect differences in the condition and risk of individual units when inspection intervals and items are set. This paper proposes a risk-based inspection planning framework that determines the inspection intervals and items of 154 kV power transformers using statistical risk curves and a mapping between failure modes and inspection items derived from failure mode and effects analysis (FMEA). The asset-level probability of failure is obtained from a two-parameter Weibull distribution and apportioned to the winding, bushing, and on-load tap changer (OLTC) according to their failure contributions, and the consequence of failure is evaluated in cost terms for the reliability, financial, safety, and environmental aspects. The risk curve is updated with inspection results; the next inspection time is the year in which the updated risk reaches the next reference risk, and the inspection items follow a grade based on the annual risk increment. Transformers selected in the same year are prioritized by transformer-level risk within the available budget. A scenario-based case study of five transformers analyzes how the inspection times and items change under different condition-change settings and compares the number of inspections, total inspection cost, and present value with time-based inspection. The inspection assignments and unassigned unit-years of the two approaches are also compared under the same budget constraint.

Data sparsity, particularly the partial loss of diagnostic data caused by sensor failures, transmission errors, or missed inspections, frequently occurs in practical power transformer operations and significantly degrades the accuracy and reliability of health index (HI) assessments. In this study, a machine learning-based HI evaluation framework is developed using key diagnostic input parameters systematically derived from failure mode and effect analysis (FMEA) and established transformer diagnostic practices. To compensate for missing data, an unsupervised autoencoder (AE)-based imputation method is introduced and benchmarked against conventional statistical supplementation techniques, namely mean and mode imputation. The experimental results, obtained using real inspection-based transformer diagnostic records, demonstrate that the AE-based approach effectively preserves inter-variable correlations and latent data structures by learning nonlinear feature relationships. As a result, the proposed method maintains robust and consistent HI classification performance under varying missing-data conditions. Furthermore, validation using confirmed transformer failure cases shows that the AE method more accurately reconstructs missing dissolved gas analysis indicators and improves the identification of high-risk equipment compared with statistical imputation. Overall, the proposed approach provides decision-consistent HI evaluations even when diagnostic data are incomplete, thereby reducing uncertainty in maintenance planning and minimizing the need for additional follow-up inspections solely to compensate for missing information.

This paper presents a frequency-dependent cost analysis model for optimizing the economic performance of Low-Frequency AC (LFAC) transmission systems for offshore wind integration. Unlike previous studies that assumed fixed frequencies (e.g., 16.7 Hz or 20 Hz), the proposed model evaluates how variations in frequency affect the cost of key components, including transformers, offshore platforms, and submarine cables. By modeling the cost of the total system as a function of frequency, transmission capacity, and distance, the system’s optimal operating frequency is identified under various conditions. In addition, LFAC, High-Voltage AC (HVAC), and High-Voltage DC (HVDC) systems are compared based on the concept of economic critical distance, which defines the boundary between the most economically favorable transmission technologies depending on distance and capacity conditions. The results show that HVAC is the most cost-effective solution for short transmission distances, LFAC is optimal for medium-range transmission, and HVDC is favorable for long-distance applications. Notably, LFAC outperforms HVDC when cable multiplication is avoided, confirming its competitiveness in midrange scenarios. These findings demonstrate that considering frequency as a design parameter significantly improves the cost efficiency of LFAC systems and should be reflected in the early-stage planning of offshore transmission system design.

The growing scale of offshore wind farms and increasing transmission distances has driven the demand for optimized offshore substation (OSS) configurations. This study proposes a comprehensive techno-economic framework to minimize the total lifecycle cost (LCC) of an OSS by determining the optimal number of OSSs and transformers considering wind farm capacity and transmission distance. The methodology incorporates three cost models: capital expenditure (CAPEX), operational expenditure (OPEX), and expected energy not supplied (EENS). CAPEX considers transformer costs, topside structural mass effects, and nonlinear installation costs. OPEX accounts for substation maintenance and vessel operating expenses, and EENS is calculated using transformer failure probability models and redundancy configurations. The optimization is performed through scenario-based simulations and a net present value (NPV)-based comparative analysis to determine the cost-effective configurations. The quantitative analysis demonstrates that for small- to medium-scale wind farms (500–1000 MW), configurations using 1–2 substations and 3–4 transformers achieve minimal LCC regardless of the transmission distance. In contrast, large-scale wind farms (≥1500 MW) require additional substations to mitigate transmission losses and disruption risks, particularly over long distances. These results demonstrate that OSS design should holistically balance initial investment costs, operational reliability, and supply security, providing practical insights for cost-effective planning of next-generation offshore wind projects.

Renewable energy sources continue to increase due to the energy transition; thus, the generation output of conventional power sources is decreasing. The installation of renewable energy sources can lead to the concentration of these sources depending on geographical characteristics, which may cause bottlenecks between the renewable energy generation sites and load centers, and such bottlenecks can result in power shortages in load centers and may cause issues that limit the integration of renewable energy. Thus, this paper proposes the application of an LFAC multiport asynchronous grid connection system to solve these problems, where the frequency conversion device uses a variable frequency transformer (VFT). In addition, the installation location of the proposed LFAC multiport asynchronous grid connection system is selected using a grid partitioning technique, and the optimal power flow is performed to minimize the generation costs. The grid partitioning was conducted using the IEEE 39 bus system, and the feasibility of the proposed LFAC multiport asynchronous grid connection system was verified through simulation tests of the generation and load demands in the grid. In addition, the VFT control and optimal power flow control performances were confirmed through MATLAB/Simulink.

Power companies employ comprehensive evaluations to manage facilities, such as transformers, lines, and generators, assessing their status and calculating the probability of failure. Within this asset management framework, a key metric that indicates the condition of facilities is the health index. To develop an optimal investment strategy for facility maintenance and management, a method for accurately calculating the health index is essential. Recent studies have explored the application of machine learning to improve accuracy. However, these studies ignore the risk of misclassification, which can result in significant financial losses when predicting the health index of power transformers. Therefore, this study proposes a health index calculation algorithm that incorporates cost-sensitive learning to assess the impact of each misclassification scenario, construct a cost matrix, and minimize the cost through learning. Input parameters are selected on the basis of the power transformer’s structure, failure mode effect analysis, failure diagnosis techniques, and historical failure cases. During the learning process, the relationship between actual and predicted classes is analyzed, and the risk associated with each misclassification scenario is quantified to describe the cost matrix. The effectiveness of the proposed health index calculation algorithm is validated using in-service power transformer inspection data, demonstrating a reduction in the risk of specific misclassification scenarios.