Electrolyzer

Electrolyzer is a controls-oriented engineering model for hydrogen production systems. It simulates multi-stack electrolyzer operation, supports PEM and alkaline cell models, tracks degradation, and includes levelized cost of hydrogen (LCOH) analysis utilities.

What this repo provides

• Time-series simulation of one or more stacks with supervisory control logic.
• PEM and alkaline electrochemical cell models with polarization curve fitting.
• Degradation tracking (steady, fatigue, and on/off cycling) with optional penalty modes.
• Cost and LCOH analysis tools tied to simulation outputs.
• YAML-based modeling configuration with a JSON schema for validation and defaults.

Project structure

• Core simulation: electrolyzer/simulation
• Cell models: electrolyzer/simulation/cell_models
• Validation/schema: electrolyzer/tools/validation.py, electrolyzer/tools/modeling_schema.yaml
• LCOH analysis: electrolyzer/tools/analysis
• Examples: examples
• Documentation: docs

Installation

Python 3.11+ is required.

pip install .

Optional extras:

pip install ".[examples]"   # notebooks + example dependencies
pip install -e ".[develop]"  # dev + docs tooling
pip install -e ".[all]"      # everything

More detail is in docs/installing.md.

Quick start

Run a simulation from a YAML configuration and a power signal:

import numpy as np

from electrolyzer.simulation.bert import run_electrolyzer

power_signal = np.ones(3600) * 1e6  # 1 MW for 1 hour, in Watts
elec_sys, results = run_electrolyzer("examples/example_02_electrolyzer/modeling_options.yaml", power_signal)

print(results.head())

Compute LCOH using the same signal:

import numpy as np

from electrolyzer.tools.analysis.run_lcoh import run_lcoh

power_signal = np.ones(3600) * 1e6
lcoe = 0.04418  # $/kWh

lcoh_breakdown, lcoh_value = run_lcoh(
    "examples/example_04_lcoh/cost_modeling_options.yaml",
    power_signal,
    lcoe,
)

print(lcoh_value)

Modeling configuration

Models are configured with YAML files validated against a JSON schema. The schema defines defaults and accepted ranges for parameters like stack rating, cell geometry, degradation rates, and control policy settings.

• Schema: electrolyzer/tools/modeling_schema.yaml
• Example PEM configuration: examples/example_02_electrolyzer/modeling_options.yaml
• Example alkaline configuration: examples/example_06_alkaline/default_alkaline.yaml

Key configuration blocks:

• electrolyzer.supervisor: system rating and number of stacks.
• electrolyzer.controller: control strategy and decision policy flags.
• electrolyzer.stack: stack sizing, cell type, and operational settings.
• electrolyzer.degradation: degradation rates and end-of-life parameters.
• electrolyzer.cell_params: PEM or alkaline cell model parameters.
• electrolyzer.costs: LCOH input data for capex, opex, feedstock, and finance.

Control strategies

The supervisor supports multiple control modes for stack scheduling and power distribution:

• PowerSharingRotation, SequentialRotation
• EvenSplitEagerDeg, EvenSplitHesitantDeg
• SequentialEvenWearDeg, SequentialSingleWearDeg
• BaselineDeg
• DecisionControl (composed from policy flags in the YAML)

See electrolyzer/simulation/supervisor.py for logic.

Degradation modeling

Each stack tracks voltage degradation from steady operation, fatigue, and on/off cycling. You can choose whether degradation penalizes hydrogen production or increases power draw. The end-of-life voltage delta drives replacement calculations in the LCOH workflow.

Outputs

run_electrolyzer returns a supervisor object and a pandas.DataFrame of time-series results. The frame includes overall power and curtailment plus per-stack columns for degradation, cycles, uptime, hydrogen production rate, and current density.

Examples

• Basic simulation: examples/example_02_electrolyzer/example_run.py
• Polarization curve fitting: examples/example_01_polarization/example_run.py
• Controller behavior: examples/example_05_controller/example_05_controller_options.py
• Alkaline configuration: examples/example_06_alkaline/alkaline_example_run.py
• LCOH calculation: examples/example_04_lcoh/cost_example_run.py

Documentation

Docs are in docs. The landing page is docs/intro.md. If you build the Jupyter Book locally, the generated site lands in docs/_build/html.

Testing

pytest