Sungrow’s modular inverters feature an AI-enabled fault diagnosis function
The company claims that it detects up to 20 common fault types with more than 96% diagnosis accuracy within seconds
In a joint collaboration with TÜV Rheinland, the faulty 1+X 2.0 system was replaced and repaired within 1 hour of deliberate fault generation
Global PV inverter and energy storage solutions provider Sungrow has published the results of a simulation that demonstrates the advantages of its 1+X 2.0 series modular inverter, by diagnosing faults and the subsequent replacement of the faulty unit in under 1 hour.
According to the company’s website, each inverter comprises an 800 kW IGBT-based power module (DC-AC) in a sleek modular enclosure. Once these inverters are placed side-by-side in parallel connection (electrical), the cumulative capacity (AC) feeds the DC-coupled battery banks (charging) via an intermediate bidirectional power conversion system (PCS) and vice-versa.
The company emphasizes modularity as the key feature of these inverters. Unlike the centralized configuration of the central inverter, which requires a full shutdown during any fault, this configuration enables quick isolation and replacement of the faulty unit with a healthy one without shutting the system down. The company states that these inverters feature an AI-enabled fault-diagnosis capability that combines machine learning, neural networks, and expert systems to automatically identify faults. This is followed by the fault type identification (20 common fault types) and the generation of repair suggestions within seconds. The company claims its diagnostic accuracy exceeds 96%, covering major fault types like IGBT faults, reactor overheat, and PDP faults, among others. It adds that this ‘smart’ diagnostic function is getting smarter by fetching and feeding operational diagnostic data day-to-day, making it more precise and reliable.
As mentioned above, these faulty inverters, each weighing 105 kg, can be dismantled and replaced by 2 on-site O&M personnel using auxiliary tools. With automatic fault diagnosis and subsequent replacement, the system reduces partial system breakdown to less than 1 hour and optimizes O&M costs, while avoiding full system shutdown.
The company cites a case study, jointly conducted with TÜV Rheinland at the Tuanfeng plant in Hubei Province, China, to validate the modular inverter’s real-world performance. After inducing a deliberate IGBT fault in the inverter, which was successfully detected, the subsequent affected component was identified within 3 minutes, and the detailed fault description and relevant repair suggestions were displayed at the system human-machine interface (HMI). This was followed by the removal of the front cover, disconnection of the copper bar terminals, and final replacement with a healthy one in under 58 minutes, the company claims.
