Electrolyser Cell Fault Detection via Neural Voltage Degradation Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting faults in chlor-alkali electrolyser cells are inadequate, as they only detect abnormal degradations after they have progressed to a point of no return, leading to potential safety hazards and inefficiencies in energy consumption.
Innovation Solution
A neural network architecture is employed to predict synthetic cell voltages based on cell-specific parameters and normal degradation, allowing for early detection of faults by comparing measured voltages to synthetic ones, with the ability to output alerts or shutdown signals when voltage differences reach a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fault detection methods are used in electrolyser cells, then the detection process is simple, but faults are only detected after they have progressed to a point of no return, leading to safety hazards and energy inefficiency
Solution Approach 1:
The system performs preliminary action by predicting future cell voltages based on historical data and normal degradation patterns before actual faults occur. The neural network model continuously forecasts expected voltage values and compares them with measured values, enabling early detection of deviations that indicate developing faults. This preliminary prediction approach allows the system to identify potential failures before they reach a critical point, transforming reactive fault detection into proactive prevention.
Solution Approach 2:
The patent introduces an intermediary mechanism - a neural network model that acts as a mediator between historical voltage data and current fault detection. This intermediary model learns normal degradation patterns and generates predicted voltage values that serve as a reference for detecting actual faults. The model translates complex historical data into meaningful predictions, enabling reliable fault detection without requiring direct observation of the fault itself until it becomes critical.
2Reliability
If early fault detection is implemented using neural networks, then safety issues are prevented and cell lifespan is extended, but the system complexity and computational requirements increase
Solution Approach 1:
The system implements self-service by using its own historical voltage data to train and continuously improve its predictive capabilities. The neural network model learns from the electrolyser's actual operational patterns and degradation trajectory, adapting to cell-specific characteristics without requiring external intervention or manual calibration. This self-learning approach enables the system to automatically adjust to normal aging patterns, distinguishing them from abnormal fault conditions, thereby maintaining high reliability while reducing the need for external system complexity.
Solution Approach 2:
The patent employs feedback mechanisms where the difference between predicted and actual voltage values is continuously monitored. When deviations exceed predefined thresholds, the system generates alerts or shutdown signals, creating a closed-loop control system. This feedback approach enables real-time adjustments and continuous improvement of detection accuracy, maintaining high safety standards while managing system complexity through adaptive thresholding and incremental learning from operational data.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, systems, and assemblies for detecting faults in an electrolyser having a plurality of electrolysis cells are described. The method comprises obtaining voltage measurements of the electrolysis cells during operation of the electrolyser, generating synthetic cell voltages for the electrolysis cells using a neural network architecture that takes into account normal cell degradation based on cell-specific parameters, comparing the voltage measurements to the synthetic cell voltages for corresponding ones of the electrolysis cells to obtain voltage differences, and detecting a fault in the electrolyser when at least one of the voltage differences reaches a threshold.