Understanding the Complexity of Green Hydrogen Production — Mirroring Digital Twin
https://simacro.com/wp-content/uploads/2025/11/0_3d_Rendering_Battery_1280x720.mp4
Green hydrogen production based on renewable energy is far more than a simple energy conversion process. The intermittency and variability of solar and wind power, the degradation of electrolyzers, the PSA and hydrogen carrier conversion processes, and the interconnections between production, storage, and transport systems create a level of operational complexity far greater than that of conventional chemical plants.
If this complexity cannot be holistically managed, efficiency drops, equipment life shortens, and economic viability declines. The first step toward overcoming these challenges begins with the Mirroring Digital Twin.
Mirroring Reality through Soft Sensing and Synthetic Data Augmentation
At the heart of green hydrogen production lies the electrolyzer, whose efficiency and durability are highly sensitive to operating conditions. In the mirroring stage, real-time data such as voltage, current density, temperature, pressure, and gas flow rate are continuously collected and synchronized with a virtual electrolyzer model.
The resulting real-time VI (Voltage-Current) chart enables:
- GAP analysis between the actual operation point and the monthly average VI curve,
- Separation and estimation of Activation, Ohmic, and Concentration losses using Soft Sensing,
- Visualization of macroscopic degradation behaviors across electrode, membrane, and electrolyte domains — serving as an ageing-reflective operational indicator in real time.
Over time, the gradual shift or slope change of the VI curve reveals decreasing electrode activity, rising membrane resistance, and declining electrolyte conductivity — quantifiable signs of long-term degradation.
Soft sensing estimates internal variables that cannot be directly measured (such as overpotential, membrane resistance, and ion conductivity), providing early insights into efficiency loss and degradation progression.
Virtual Sensing Layer and Degradation Indicators
The Mirroring Digital Twin establishes a Virtual Sensing Layer, where real-time sensor data are fused with soft-sensing models and synthetic data generation to extend plant observability beyond measurable variables. This fusion creates a comprehensive digital reflection of the electrolyzer’s internal behavior — bridging the gap between what is measured and what is inferred.
Within this layer, the twin continuously evaluates key degradation indicators, derived from both measured and synthetic data, such as:
- Voltage drift rate (ΔV/Δt) — representing gradual efficiency decline,
- Ohmic resistance trend — indicating membrane or electrode resistance growth,
- Electrolyte conductivity (κ) variation — tracking long-term ionic degradation.
By quantifying these parameters, the virtual sensing layer provides early insights into the health and ageing of the electrolyzer system, enabling proactive maintenance and improved operational reliability.
Membrane Degradation and Crossover Safety Monitoring
Gas crossover between the hydrogen and oxygen compartments is a key indicator of membrane ageing — but the behaviors of hydrogen and oxygen crossover differ fundamentally in both dynamics and detectability.
While oxygen crossover is the critical safety concern, its transport through the membrane exhibits a threshold-like behavior: below a certain degradation level, O₂ diffusion remains negligible, and only after a critical point does rapid permeation occur. This makes direct oxygen measurement an unreliable early diagnostic.
In contrast, hydrogen crossover progresses gradually and continuously as the membrane deteriorates. By mirroring the measured H₂ concentration on the anode side, the Mirroring Digital Twin can track how gas permeability evolves over time, effectively mapping the trajectory of membrane degradation.
Through this process, the digital twin updates membrane correction parameters (permeability, diffusion rate, or tortuosity) to reflect real-time degradation, allowing the system to soft-sense oxygen crossover behavior before it becomes physically measurable. The gradual H₂ crossover trend thus forms the basis for a Membrane Degradation Index (MDI), while the potential onset of O₂ crossover serves as its associated safety risk indicator.
In essence, hydrogen crossover acts as the mirror of degradation, and oxygen crossover as the mirror of risk — together enabling predictive, model-informed membrane health monitoring.
Example: Electrolyzer Degradation Mechanisms & Monitoring Metrics
| Degradation Mechanism | Primary Effect | Monitoring Indicator (Example) |
|---|---|---|
| ① Electrode catalyst loss (corrosion, detachment) | Increased activation overpotential | Activation Overpotential, i₀, Charge Transfer Coefficient |
| ② Electrode structural change (pore clogging, reduced active area) | Higher charge-transfer resistance | VI curve slope variation (Soft Sensing estimation) |
| ③ Membrane damage (crack, contamination, wetting) | Lower ionic conductivity, gas crossover | Anode H₂ ↑, Cathode O₂ ↑, Membrane resistance ↑, Tortuosity change |
| ④ Fouling / Scaling (salt deposition, oxide film) | Increased resistance, heat transfer loss | Δ(ΔV/ΔI) slope, ΔT(in–out), ΔP → Fouling Resistance |
| ⑤ Electrolyte concentration shift | Conductivity loss, corrosion acceleration | Electrolyte conductivity (k), pH variation |
These indicators are continuously calculated in the Mirroring Digital Twin environment, combining sensor data and model-based estimations. This enables quantitative, real-time monitoring of degradation across electrodes, membranes, and electrolytes — forming the foundation for advanced diagnostics and decision support in the AI-Assisted DT stageWith headquarters in Boston and Seoul, SIMACRO has completed over 90 commercial modeling projects across 40 companies since 2018. Collaborating with global technology leaders such as AspenTech, Emerson, and OLI, SIMACRO is committed to advancing digital innovation in the process industry.
About SIMACRODesignerWonseok Lee is Executive VP of Operations at SIMACRO. As an expert in process design, modeling, and simulation, he performed business and solution consulting in the process industry at AspenTech for 20 years.
About Wonseok LeeDesigner