In Korean society, energy storage systems carry the reputation of a double-edged instrument. They are indispensable infrastructure for the transition to clean energy, but the twenty-three ESS fires that occurred between 2017 and 2019 left a lasting social memory of grid-scale battery risk. Each incident generated, on average, approximately 6.7 billion won in direct losses and approximately 24 billion won in indirect losses (downtime, reconstruction, and elevated insurance costs); the Ministry of Trade, Industry and Energy and the Korea Electrical Safety Corporation (KESCO) jointly produced a comprehensive root-cause analysis1. The analysis identified five categories of cause: (1) battery cell defects, (2) inadequate protection systems, (3) unsuitable operating environments, (4) installation deficiencies, and (5) system-integration deficiencies. These five categories became the structural foundation of the KESCO ESS Fire Safety Standard 20242.
This chapter decomposes the full life cycle of ESS operations into five axes—predictive maintenance, anomaly detection, fire safety, digital twin, and recycling and second life—and surveys the standardization status and Korean infrastructure mapping along each axis. The WIA-ESS manifest classifies the operations domain through six ENUM values: PREDICTIVE_MAINTENANCE, ANOMALY_DETECTION, DIGITAL_TWIN, BATTERY_RECYCLING, SECOND_LIFE_BATTERY, and CIRCULAR_ECONOMY, and mandates disclosure of activation and implementation level for each.
Reactive maintenance repairs equipment after failure has occurred; preventive maintenance schedules inspection and replacement on a calendar to avoid failure. Predictive maintenance (PREDICTIVE_MAINTENANCE) is the next-generation approach: machine-learning analysis of operational data forecasts failure events in advance, and maintenance is performed at the cost-optimal moment. Research published by the Korea Electrotechnology Research Institute (KERI) indicates that adoption of predictive maintenance reduces maintenance cost by approximately twenty-eight percent and improves availability by approximately 4.2 percent relative to preventive maintenance.
Effective predictive maintenance requires simultaneous collection of five data channels: (1) per-cell voltage, current, and temperature (one-second intervals); (2) per-module SOC, SOH, and DOD (ten-second intervals); (3) per-system input and output and efficiency (one-minute intervals); (4) environmental conditions (ambient temperature, humidity, and airflow, at five-minute intervals); and (5) protection-system event logs (event-driven). WIA-ESS mandates that the `data_collection_spec{}` block of the manifest disclose the sampling interval and retention period for each of these five channels. The KESCO ESS Fire Safety Standard further requires a minimum twelve-month retention of all data.
A predictive maintenance pipeline typically comprises four stages: (1) preprocessing (denoising, outlier correction, normalization), (2) feature extraction (Remaining Useful Life, SOH change rate, impedance change), (3) machine-learning model training (LSTM, Transformer, Gaussian Process Regression), and (4) risk-score computation and maintenance-window determination. KIER's open-source KIER-PM 1.0 algorithm, released under the BSD 3-Clause license, is the WIA-ESS reference implementation; it achieves approximately 8.3 percent mean absolute percentage error in RUL prediction across a 6,000-cycle empirical dataset.
Anomaly detection (ANOMALY_DETECTION) detects, in real time, signals that deviate from the normal operating pattern in a statistically meaningful way. Detection typically operates across three temporal layers: (1) millisecond scale (rapid voltage or current change, short-circuit detection); (2) second-to-minute scale (accelerated temperature rise, growing cell imbalance); and (3) hour-to-day scale (accelerated SOH decline, anomalous self-discharge). WIA-ESS mandates that the `anomaly_detection_layers[]` array disclose the threshold and response time for each of the three layers.
Anomaly detection algorithms fall into three categories: (1) statistical threshold-based, (2) supervised machine-learning classification, and (3) deep-learning autoencoder. Threshold-based methods are simple but have high false-alarm rates. Supervised classification depends heavily on training data. Autoencoders enable unsupervised learning but offer limited interpretability. In practice, the three approaches are commonly combined into an ensemble. WIA-ESS classifies the algorithm with the `anomaly_algorithm` ENUM accepting rule, ml, dl, or ensemble.
Lithium-ion cell thermal runaway progresses through five stages: (1) initiation of solid electrolyte interphase (SEI) decomposition at approximately 80 to 120 degrees Celsius, typically triggered by internal short circuit or overheating; (2) electrolyte decomposition and gas evolution at approximately 130 to 160 degrees; (3) separator melting and full internal short circuit at approximately 170 to 200 degrees; (4) cathode active-material decomposition and oxygen release at approximately 200 to 230 degrees; and (5) self-accelerating reaction and ignition above approximately 230 degrees. The sequence typically completes within a few minutes; thermal runaway in one cell propagates rapidly to adjacent cells, expanding into module-level and system-level fire.
UL 9540A, Test Method for Evaluating Thermal Runaway Fire Propagation in Battery Energy Storage Systems, evaluates thermal-runaway propagation at four hierarchical levels: cell, module, unit, and installation. Systems that pass UL 9540A are certified as having limited propagation to adjacent cells and an actionable fire-suppression profile. The Korean ESS Fire Safety Standard mandates UL 9540A test reports for all new ESS deployments, with the report URL recorded in the manifest field `ul_9540a_test_report_url`.
NFPA 855, Standard for the Installation of Stationary Energy Storage Systems, defines installation-level fire safety in the United States. The standard prescribes maximum module capacity per chemistry, minimum separation distances between systems, fire-suppression system requirements, and ventilation and exhaust requirements. The Korean KESCO ESS Fire Safety Standard 2024 aligns with NFPA 855 while adapting requirements to Korean building codes and electrical installation standards.
IEC 62619, Secondary cells and batteries containing alkaline or other non-acid electrolytes—Safety requirements for secondary lithium cells and batteries for use in industrial applications, defines the safety requirements for industrial lithium-ion cells. The standard prescribes electrical, mechanical, thermal, and environmental tests; IEC 62619 certification of the underlying cell is a prerequisite for KS C 8546 compliance.
A digital twin (DIGITAL_TWIN) synchronizes a virtual model with the real-time behavior of a physical ESS asset. Operational data continuously feed the virtual model, which simulates future behavior, faults, and aging. WIA-ESS defines the digital-twin interface in the IEC 63278-1:2023 Asset Administration Shell format and exposes it through the manifest fields `digital_twin_endpoint` and `digital_twin_protocol`. In Korea, the joint ESS Digital Twin Platform operated by KIER, KERI, and KIMM hosts approximately 120 sites as of 2024 in a cloud environment.
The principal functions of a digital twin are: (1) incident simulation (thermal runaway propagation and short-circuit propagation), (2) operational optimization (multi-product value stacking), (3) cycle-aging prediction, (4) virtual commissioning, and (5) virtualization of safety inspections. Adoption of a digital twin can virtualize approximately forty percent of the quarterly KESCO safety inspection burden, substantially improving operational efficiency.
Once an SOH of 70 to 80 percent is reached, lithium-ion cells are considered to have reached end of first life and follow one of two paths: (1) second life (SECOND_LIFE_BATTERY), in which the cell is redeployed for less demanding applications (residential backup, telecom base stations, microgrids); or (2) recycling (BATTERY_RECYCLING), in which the cell is disassembled to recover cobalt, nickel, lithium, manganese, and copper. SNE Research's Battery Second-Life Market Outlook 2024 reports trading values of approximately thirty to forty-five percent of new-cell value for cells in the 70-to-80-percent SOH range; cells below 60 percent SOH are directly recycled.
EU Battery Regulation 2023/1542 imposes the following obligations on every battery placed on the EU market: (1) carbon footprint declaration from 2025; (2) minimum recycled-content thresholds from 2030 (16 percent cobalt, 6 percent nickel, 6 percent lithium); (3) mandatory digital battery passport from 2027; (4) collection targets of 51 percent in 2031 and 73 percent in 2036; and (5) recovery targets of 90 percent for cobalt and 80 percent for lithium. WIA-ESS aligns with this Regulation, and the manifest field `eu_battery_passport_id` encodes the digital passport identifier. The full digital battery passport function is covered by the separate WIA-Battery-Passport standard.
Korea's Resource Circulation Framework Act, Article 15, and the Act on the Resource Circulation of Electrical and Electronic Equipment and Vehicles, Article 13, govern the disposal and recycling of ESS batteries. ESS operators must deliver batteries to a recycler registered with the Ministry of Environment, and the recycler must separate the cells by chemistry (LFP, NCM, NCA). WIA-ESS records the disposal plan in the manifest field `end_of_life_plan`. As of 2024, approximately thirty-two recyclers are registered, five of which (SungEel HiTech, Ecopro, LG Energy Solution ESS Recycle, Samsung SDI Battery Reborn, and SK On Second-Life) are classified as large-scale operators.
The circular economy (CIRCULAR_ECONOMY) is an economic model that minimizes waste and maximizes resource reuse and recycling. In the ESS domain, circular-economy practices implement across three life-cycle stages: (1) Design for Recycling at the cell-design stage; (2) Optimized Operation during use that minimizes cycle aging; and (3) Closed-Loop Recycling at end of life. WIA-ESS mandates disclosure of compliance with these three stages in the `circular_economy_scoring{}` block.
KESCO is the national public agency that performs installation and operational safety inspections of ESS. Following the twenty-three fires of 2017 through 2019, KESCO's inspection authority was substantially strengthened, and KESCO released the ESS Fire Safety Standard 2024 in 2024, mandating quarterly safety inspections of every ESS. The Standard comprises five axes: (1) maximum module capacity per cell chemistry, (2) minimum separation distance between systems, (3) fire-suppression system requirements (sprinklers, suppression gas, central detection), (4) ventilation and exhaust requirements, and (5) emergency-stop system requirements. The manifest fields `kesco_safety_inspection_date` and `kesco_certificate_id` carry the most recent inspection date and certificate identifier respectively.
The Korean Fire Protection Association (KFPA) maintains the ESS Fire Risk Assessment Guideline 2024, which insurers use as the basis for setting ESS premiums; assets at the Anchored conformance tier of WIA-ESS receive an approximately twenty-five-percent premium discount. KFPA also maintains the ESS Fire Incident Database, which tracks statistical trends. KERI conducts UL 9540A, IEC 62619, IEC 62933, and KS C 8546 testing and certification at its Changwon and Ansan laboratories, completing approximately 240 tests in 2024.
All three Korean battery manufacturers operate ESS battery recycling and second-life businesses: LG Energy Solution's ESS Recycle (Ochang facility, established 2022), Samsung SDI's Battery Reborn (Cheonan facility, established 2023), and SK On's Second-Life (Seosan facility, established 2024). The three are registered with the Ministry of Environment and simultaneously comply with the Korean Resource Circulation Framework Act and EU Battery Regulation 2023/1542. Independent specialists SungEel HiTech (Cheonan facility) and Ecopro BM (Pohang facility) also occupy significant market share; SungEel HiTech achieves approximately 98-percent cobalt recovery and 80-percent lithium recovery through a hydrometallurgical process, while Ecopro reaches a 92-percent average recovery across five materials through an integrated hydro- and pyro-metallurgical process.
Operational conformance is segmented into three tiers. Surface implementations self-declare predictive maintenance, anomaly detection, digital twin, and recycling-plan provisions. Verified implementations attach test reports under UL 9540A, IEC 62619, and the KESCO ESS Fire Safety Standard 2024. Anchored implementations stream operational data, anomaly-detection event logs, and predictive-maintenance results into the manifest's `realized_operations{}` block and anchor it to Sigstore Rekor; quarterly KESCO inspection records are also automatically recorded.
Panel 4, operations mode, exposes visualizations for the six operational ENUMs (PREDICTIVE_MAINTENANCE, ANOMALY_DETECTION, DIGITAL_TWIN, BATTERY_RECYCLING, SECOND_LIFE_BATTERY, CIRCULAR_ECONOMY). Users enter cell chemistry, cycle count, DOD distribution, and operating environment (ambient temperature and humidity), and the simulator returns predictive-maintenance RUL estimates, anomaly-detection risk scores, digital-twin incident simulations, recycling and second-life value estimates, and UL 9540A, IEC 62619, and KESCO compliance scores.
ESS operations and safety are not purely technical concerns; they are a matter of social trust in Korea. The twenty-three fires of 2017 to 2019 left a deep imprint on the ESS industry and halted deployment growth for approximately five years. The mandatory disclosure regime of the WIA-ESS operations domain is therefore not merely a proof of technical safety but an objective commitment to rebuilding trust between industry and society.
Disclosure obligations imposed by KFPA and KESCO are particularly central to that trust. Transparent disclosure of incident data, objective post-incident root-cause analysis, and the integration of preventive measures into standards must form a virtuous cycle. The mandatory Sigstore Rekor anchoring of operational data for Anchored conformance tier provides the technical foundation for that cycle.
Standardization of recycling and second life is also a social commitment that addresses the “shadow side” of the clean energy transition. With Korean manufacturers holding approximately thirty-seven percent of the global ESS market, end-of-life cells shipped abroad must be properly processed or the original purpose of the clean energy transition is undermined. The alignment of WIA-ESS with EU Battery Regulation 2023/1542 and the direct linkage to the Korean Resource Circulation Framework Act establish Korea as a responsible supplier of clean clean energy in the global market.
An integrated review of MOTIE, Ministry of Environment, KESCO, and KFPA policies anticipates the following five-step evolution: in 2025 formal enforcement of the KESCO ESS Fire Safety Standard 2024 makes UL 9540A testing mandatory for all new ESS; in 2026 a national predictive-maintenance standardization committee publishes the Korean standard KS X PR-PM 2026; in 2027 mandatory digital-twin deployment expands to all ESS above 100 megawatts; in 2028 Korea's domestic battery digital passport scheme takes effect, aligning with EU Battery Regulation 2023/1542; and by 2030 every ESS reaches at least the Verified tier and Anchored-tier assets exceed forty percent of the fleet.
The Korean Ministry of Trade, Industry and Energy released the first comprehensive root-cause analysis of the twenty-three ESS fires of 2017 through 2019 in June 2019 and a second analysis in February 2020. The first identified three causes: inadequate protection systems, unsuitable operating environments, and installation deficiencies. The second added two additional causes: battery cell defects and system-integration deficiencies. These analyses provided the policy foundation for the ESS Fire Safety Comprehensive Plan (2019) and the ESS Fire Safety Standard 2024 (2024) and are the direct origin of the mandatory operations-domain manifest fields in WIA-ESS.
The KESCO ESS Fire Safety Standard 2024, announced in June 2024 and effective from December 2024, is the comprehensive standard that makes quarterly safety inspections mandatory for every new ESS. It rests on five axes: mandatory UL 9540A thermal runaway testing, NFPA 855-aligned installation requirements, mandatory IEC 62619 cell safety testing, quarterly KESCO inspections, and incident reporting within one hour of occurrence. WIA-ESS is designed to align with this Standard, which serves as the baseline conformance reference for every ESS installed in Korea.
All code examples, simulator panels, and manifest schemas in this chapter are maintained as a single source of truth in the GitHub public repository github.com/WIA-Official/wia-standards-public/tree/main/energy-storage. Examples in the printed edition, EPUB, and PDF are snapshots; readers should consult the repository's CHANGELOG and release notes for the latest revisions. Every manifest example carries a Sigstore DSSE signature; the corresponding Rekor transparency log index enables post-hoc verification.
弘益人間 (Hongik Ingan) — Benefit All Humanity