Feature/bert basic #81

elenya-grant · 2025-02-26T18:18:43Z

Feature-add: electrolyzer model partial sync-up

Primary changes

replaced hard-coded values with attributes in pem cell
added functionality for lifetime performance assessment
added functionality to specify EOL and whether power consumption or hydrogen production are effected by degradation
added functionality for hydrogen production to decrease with degradation

Related issue, if one exists

Impacted areas of the software

Modified files

electrolyzer/simulation/cell.py changed to electrolyzer/simulation/cell_models/pem.py
tests/test_cell.py changed to tests/test_PEM_cell.py
electrolyzer/simulation/bert.py: added run_LCA function which estimates lifetime performance
electrolyzer/simulation/base.py: new file, added base class
electrolyzer/simulation/cell_models/cell.py: new file (empty right now) to function as cell model base class
electrolyzer/simulation/cell_models/alkaline.py: clean-ups and parameter renaming. Also fixed U_rev calc (used to be redundant)
electrolyzer/simulation/cell_models/pem.py: added p_anode, p_cathode, i_0_c, i_0_a, alpha_a, alpha_c, e_m, R_ohmic_elec, f_1, and f_2 as attributes (replacing hard-coded values). Removed dryer_loss input from calc_mass_flow_rate()
electrolyzer/simulation/cluster.py: new file (empty right now) to contain cluster class
electrolyzer/simulation/stack.py: added end-of-life definition attributes (eol_eff_percent_loss) and variable that can be used to either increase power consumption or reduce hydrogen production as stack is degraded (hydrogen_degradation_penalty). Added two functions, run_power_deg_penalty() and run_h2_deg_penalty() which are now called in run() to apply degradation impacts to the power consumption or the hydrogen production. Also added new functions for calculating and estimating degradation impacts on lifetime performance which are:
- calc_end_of_life_voltage()
- estimate_time_until_replacement()
- estimate_stack_life()
- estimate_life_performance_from_year
electrolyzer/tools/modeling_schema.yaml: updated for new inputs to stack and cell models

Tests changed:

tests/glue_code/test_run_electrolyzer.py: value changed because removed dryer loss
tests/glue_code/test_run_lcoh.py: values changed because removed dryer loss
tests/test_stack.py: added new input parameters to input dictionary
tests/test_PEM_cell.py: added new input parameters to input dictionary

New files:

docs/alkaline_ref.md: references for alkaline electrolyzer model
electrolyzer/tools/validators.py: validators

Additional supporting information

added more cell parameters to input file instead of hard-coded (PR Updated PEM cell modeling schema #75)
added hydrogen production losses occurring from degradation
added lifetime performance estimation in stack.py

Test results, if applicable

Tests pass

related issue

…option

…mat_features

genevievestarke · 2025-08-29T14:37:30Z

electrolyzer/simulation/bert.py

+        # stack_LCA = pd.concat([stack_LCA,temp_df])
+    return {
+        "Refurb Schedule": sys_refurb,
+        "AEP [kWh/year]": sys_aep,


Is AEP here the input power to the electrolyzer system?

I see now that it is, maybe we can call it power_used? I'm open to discussion on this!

genevievestarke

Looks great to me, just a few comments for clarity!

genevievestarke · 2025-08-29T14:38:08Z

electrolyzer/simulation/bert.py

+        refturb_schedule, ahp_kg, aep_kWh = stack.estimate_life_performance_from_year(
+            plant_life_years
+        )
+        sys_refurb = pd.concat(


This should be sys_AEP?

good catch! thank you!

genevievestarke · 2025-08-29T14:45:50Z

electrolyzer/simulation/stack.py

        self.DTSS = self.calc_state_space()
+        self.d_eol = self.calc_end_of_life_voltage()
+
+    def run_power_deg_penalty(self, P_in):


Is this function applying a penalty on power?

yes - I added a doctoring!

genevievestarke · 2025-08-29T14:49:35Z

electrolyzer/simulation/cell_models/alkaline.py

    turndown_ratio: float
    max_current_density: float

+    f_1: float


Can we define f_1 and f_2 here?

genevievestarke · 2025-08-29T14:51:06Z

electrolyzer/simulation/cell_models/alkaline.py

The three cell model files (alkaline.py, cell.py, and pem.py) have a lot of commented out values. Can we clean those up so they're less confusing?

I worked on cleaning up alkaline.py mostly. I have left some remaining comments because I think they would be useful for a) future development/bugfixes and b) documentation (once we get some doc pages more properly up n running). But - I did try to remove a lot of non-informative commented out code.

johnjasa

Thanks for your patience on this review, Elenya! I really like the amount of effort and organization you put into it. I think with a few relatively small changes, mostly for clarity and docstrings, this is good to come in. I've noted some places where I think you add more context, though I didn't individually mark every method that could use an updated docstring.

I also pushed up some extremely minor typographical updates directly to the branch.

johnjasa · 2025-09-02T23:45:30Z

electrolyzer/simulation/base.py

+from electrolyzer.tools.type_dec import FromDictMixin
+
+
+class BaseClass(FromDictMixin):


I see this class is added in this PR but not used anywhere yet. Is that intended?

yes. It will be updated in a following PR to function as a general base-class for the classes.

johnjasa · 2025-09-02T23:46:45Z

electrolyzer/simulation/cell_models/alkaline.py

+        # self.A_electrode = self.electrode["A_electrode"]
+        # self.e_a = self.electrode["e_a"]
+        # self.e_c = self.electrode["e_c"]


I think this is Gen's prior comment, but if these aren't used then feel free to remove them throughout

johnjasa · 2025-09-02T23:47:17Z

electrolyzer/simulation/cell_models/alkaline.py

+        # Ra = (
+        #     rho_nickle_eff
+        #     * (self.e_a / self.A_electrode)
+        #     * (1 + (temp_coeff * (T_C - tref)))
+        # )


johnjasa · 2025-09-02T23:47:40Z

electrolyzer/simulation/cell_models/alkaline.py

+    # def calc_open_circuit_voltage(self, T_C):
+    #     """
+    #     I [A]: current
+    #     T_C [C]: temperature
+    #     return :: E_rev0 [V/cell]: open-circuit voltage
+
+    #     TODO: Are we correcting for temperature twice? U_rev0 should be just 1.229 and
+    #     never change (possibly?)
+
+    #     Reference: [Gambou, Guilbert,et al 2022]: Eqn 14
+    #     """
+    #     # General Nerst Equation
+    #     # Eqn 14 of [Gambou, Guilbert,et al 2022]
+    #     T_K = convert_temperature([T_C], "C", "K")[0]
+    #     E_rev0 = (
+    #         1.5184
+    #         - (1.5421 * (10 ** (-3)) * T_K)
+    #         + (9.523 * (10 ** (-5)) * T_K * np.log(T_K))
+    #         + (9.84 * (10 ** (-8)) * (T_K**2))
+    #     )
+    #     # OR should this just be 1.229?
+    #     # E_rev_fake = 1.229


johnjasa · 2025-09-02T23:50:00Z

electrolyzer/simulation/stack.py

 )


+F, _, _ = physical_constants["Faraday constant"]  # Faraday's constant [C/mol]


This is fun! I haven't used scipy's constants much, nice idea.

johnjasa · 2025-09-02T23:53:33Z

electrolyzer/simulation/stack.py

+        """
+        eol_eff_percent_loss [%]: efficiency drop that indicates end-of-life
+        (between 1 and 100)
+        """


Update docstring to match what the method is doing

johnjasa · 2025-09-03T00:00:51Z

electrolyzer/simulation/stack.py

+
+
 @define
 class Stack(FromDictMixin):


My overarching comment and suggestion for this Stack class is that there's a lot of pretty complex logic contained within its methods. I'd recommend giving a pretty detailed docstring for the class or methods that details what's going on. For instance, it took me a bit to trace through all the logic that happens when stacks are on or off and how time is tracked (maybe I'm a bit slow), but it'd be helpful for others to have a little guide for what's happening here.

I tried to add some doc strings to some of the operation-based methods, although I am not 100% sure on how some of these functions work or are connected to the Supervisor class (I didn't make the controllers or these functions). Future development would have some of these "operation" type functions happening at the "cluster" level instead of the stack. This would also include updating the functionality for the system to work with feedback control (as it does now) or with open-loop operation (not an option at the moment).

I wrote or helped with the operation functions so I can update those. Probably better to get this pull request in and put those in in a different pull request?

johnjasa · 2025-09-03T00:12:21Z

electrolyzer/simulation/stack.py

+        stack_life = (1 / frac_of_life_used) * (self.uptime / 3600)  # [hrs]
+        return stack_life
+
+    def estimate_life_performance_from_year(self, plant_life_years: int):


For most of these methods, a short docstring explaining what they're doing and what the logic is would be quite helpful

johnjasa · 2025-09-03T00:12:57Z

electrolyzer/tools/modeling_schema.yaml

+                    hydrogen_degradation_penalty:
+                        type: boolean
+                        default: True
+                        description: Toggle whether degradation is applied to hydrogen or power


If it's True, is the degradation applied to hydrogen? Maybe make that more clear in the description

johnjasa · 2025-09-03T00:14:26Z

electrolyzer/tools/modeling_schema.yaml

+                                    # A_electrode:
+                                    #     type: number
+                                    #     default: 300
+                                    #     description: electrode total area
+                                    #     units: cm^2
+                                    # e_a:
+                                    #     type: number
+                                    #     default: 0.2
+                                    #     description: anode thickness
+                                    #     units: cm
+                                    # e_c:
+                                    #     type: number
+                                    #     default: 0.2
+                                    #     description: cathode thickness
+                                    #     units: cm
+                                    e_e:


Remove if unused?

johnjasa

Thanks for the changes! This is a great step towards a better set of electrolyzer models in here.

genevievestarke

Overall, I think it looks great and we should merge in! I'd like to talk about how we apply the power degradation penalty at some point, but that doesn't have to be today.

genevievestarke · 2025-09-19T19:24:49Z

electrolyzer/simulation/stack.py

+            - **V_cell** (float): cell voltage in Volts
+        """
+
+        I_stack = self.electrolyzer_model(


I'm still a bit confused how this applies a penalty on the power consumed. Maybe we can chat about it! Not a blocker, just to better understand the code.

* added updated degradation calcs * New degradation rates * updated test for new deg calcs * updat pre-commit yaml * remove whitespace * update test_regression * fix test_regression * fix test_regression * troubleshooting tests * test_regression total h2 4 decimals * Update test_run_electrolyzer.py * Added example #6, conversion efficiency * changed formatting for tests * reformatted for cleaned up deg rates * delete duplicate example * small formatting changes * WIP adjust modeling schema for alkaline cells * Bump isort version for pre-commit bugfix (#60) * alkaline example formatting * update test validation * update test_validation.py * updating test_run_electrolyzer with new modeling schema * fit_params difference * save for power_to_current debug * troubleshooting before test_run_lcoh.py * test_optimization new schema * tidying up a bit * controller example debugging * changed modeling schema interface in test_optimization * removing unnecessary comment blocks and unused files * updating examples for new schema * Issue #61: Update Cathod Pressure to be 30bar - added P_STD from scipy physical constants, which is 1 bar in Pascals - mods to calc_open_circuit_voltage() include updated p_cathode, split up partial pressures for hydrogen/oxygen according to Dalton's law. Issue #63: Nernst equation was incorrect. This issue only became became apparent when adjusting cathode pressure. - changed E_cell calculation to use the correct partial pressures and replaced R+T with R*T. - Updated test_cell.py, test_stack.py, test_run_electrolyzer.py to account for the new model change. - Incomplete is test_run_lcoh.py. Need assistance to adjust lcoh_breakdown. * Had to update LCOH test because of changes to cell voltage. Life Totals, RESULTS, and lcoh_results changed with latest cell modification. * add generic citation file * delete unneeded comments * resolve merge conflicts * Updated dryer loss to 0.2%. Test stack lowered the kWh/kg. Test run_electrolyzer produced more h2. * Test lcoh results in a lower lcoh. * Updated membrane thickness in calc_ohmic_overpot. 20 microns is closer to today's commercially available * Updated test stack and fit params. Needed to update example_01 by adding dt to stack_dict * Updated run_electrolyzer test. Hydrogen output increased * Updated run_lcoh test. Results show a lower lcoh * updating reg test value after changes from merge * changed PEM_Cell class to PEMCell * updating relative imports to explicit * updating Alkaline_Cell class to AlkalineCell * adding typing for stack.cell * fixing linting * removing commented code * WIP: Reformatting repo into BERT structure * updated all py file import paths * updated jupyter notebook import paths * updated pre-commit config * Prep for release (#82) * removed setup and added to pyproject * Updated CI workflow * WIP: Added CI and docs build * Added jupyterbook docs * Updated readme * change CI to install all * Update pyproject.toml * Feature/bert basic (#81) * reformatted * updated PEM schema and removed LCOH tests temporarily * updated alkaline modeling schema and alkaline Urev0 calc * update lcoh glue_code test regression values * added hydrogen loss from degradation and input toggle to choose that option * added LCA functions to stack.py * added eol_eff_percent_loss as degradation input for stack * added run_LCA functionto run_electrolyzer script * added more tracking variables to stack * updated all py file import paths * updated jupyter notebook import paths * added validators and minor fix in stack for cell type * added skeleton for cluster and cell classes * added generic base class * added example skeleton for simulation config * added file for power-curve function templates * updated inputs for Stack in test_stack * added import to fix tests to stack.py * removed empty PEM cell test * renamed test_cell to test_PEM_cell * updated validator format * removed files unrelated to primary feature-adds * reformatted bert.py * added back-in placeholder files * removed alkaline sources in alkaline model into md file * Typographical changes * minor cleanups to alkaline cell and stack * removed commented out code and added docstrings * fixed aep datadrame in bert.py * added more docstrings to stack.py * added comments to pem.py for faradaic coefficients --------- Co-authored-by: bayc <[email protected]> Co-authored-by: John Jasa <[email protected]> --------- Co-authored-by: ZackTully <[email protected]> Co-authored-by: Zachary Tully <[email protected]> Co-authored-by: nriccobo <[email protected]> Co-authored-by: elenya-grant <[email protected]> Co-authored-by: nRiccobo <[email protected]> Co-authored-by: bayc <[email protected]> Co-authored-by: Chris Bay <[email protected]> Co-authored-by: John Jasa <[email protected]>

* added updated degradation calcs * New degradation rates * updated test for new deg calcs * updat pre-commit yaml * remove whitespace * update test_regression * fix test_regression * fix test_regression * troubleshooting tests * test_regression total h2 4 decimals * Update test_run_electrolyzer.py * Added example #6, conversion efficiency * changed formatting for tests * reformatted for cleaned up deg rates * delete duplicate example * small formatting changes * WIP adjust modeling schema for alkaline cells * Bump isort version for pre-commit bugfix (#60) * alkaline example formatting * update test validation * update test_validation.py * updating test_run_electrolyzer with new modeling schema * fit_params difference * save for power_to_current debug * troubleshooting before test_run_lcoh.py * test_optimization new schema * tidying up a bit * controller example debugging * changed modeling schema interface in test_optimization * removing unnecessary comment blocks and unused files * updating examples for new schema * Issue #61: Update Cathod Pressure to be 30bar - added P_STD from scipy physical constants, which is 1 bar in Pascals - mods to calc_open_circuit_voltage() include updated p_cathode, split up partial pressures for hydrogen/oxygen according to Dalton's law. Issue #63: Nernst equation was incorrect. This issue only became became apparent when adjusting cathode pressure. - changed E_cell calculation to use the correct partial pressures and replaced R+T with R*T. - Updated test_cell.py, test_stack.py, test_run_electrolyzer.py to account for the new model change. - Incomplete is test_run_lcoh.py. Need assistance to adjust lcoh_breakdown. * Had to update LCOH test because of changes to cell voltage. Life Totals, RESULTS, and lcoh_results changed with latest cell modification. * add generic citation file * delete unneeded comments * resolve merge conflicts * Updated dryer loss to 0.2%. Test stack lowered the kWh/kg. Test run_electrolyzer produced more h2. * Test lcoh results in a lower lcoh. * Updated membrane thickness in calc_ohmic_overpot. 20 microns is closer to today's commercially available * Updated test stack and fit params. Needed to update example_01 by adding dt to stack_dict * Updated run_electrolyzer test. Hydrogen output increased * Updated run_lcoh test. Results show a lower lcoh * updating reg test value after changes from merge * changed PEM_Cell class to PEMCell * updating relative imports to explicit * updating Alkaline_Cell class to AlkalineCell * adding typing for stack.cell * fixing linting * removing commented code * WIP: Reformatting repo into BERT structure * updated all py file import paths * updated jupyter notebook import paths * updated pre-commit config * Prep for release (#82) * removed setup and added to pyproject * Updated CI workflow * WIP: Added CI and docs build * Added jupyterbook docs * Updated readme * change CI to install all * Update pyproject.toml * Feature/bert basic (#81) * reformatted * updated PEM schema and removed LCOH tests temporarily * updated alkaline modeling schema and alkaline Urev0 calc * update lcoh glue_code test regression values * added hydrogen loss from degradation and input toggle to choose that option * added LCA functions to stack.py * added eol_eff_percent_loss as degradation input for stack * added run_LCA functionto run_electrolyzer script * added more tracking variables to stack * updated all py file import paths * updated jupyter notebook import paths * added validators and minor fix in stack for cell type * added skeleton for cluster and cell classes * added generic base class * added example skeleton for simulation config * added file for power-curve function templates * updated inputs for Stack in test_stack * added import to fix tests to stack.py * removed empty PEM cell test * renamed test_cell to test_PEM_cell * updated validator format * removed files unrelated to primary feature-adds * reformatted bert.py * added back-in placeholder files * removed alkaline sources in alkaline model into md file * Typographical changes * minor cleanups to alkaline cell and stack * removed commented out code and added docstrings * fixed aep datadrame in bert.py * added more docstrings to stack.py * added comments to pem.py for faradaic coefficients --------- Co-authored-by: bayc <[email protected]> Co-authored-by: John Jasa <[email protected]> * Adding readme --------- Co-authored-by: ZackTully <[email protected]> Co-authored-by: Zachary Tully <[email protected]> Co-authored-by: nriccobo <[email protected]> Co-authored-by: elenya-grant <[email protected]> Co-authored-by: nRiccobo <[email protected]> Co-authored-by: Cameron Irmas <[email protected]> Co-authored-by: bayc <[email protected]> Co-authored-by: Chris Bay <[email protected]>

elenya-grant and others added 28 commits January 26, 2024 16:32

reformatted

b7a2ed8

updated PEM schema and removed LCOH tests temporarily

a7e3778

updated alkaline modeling schema and alkaline Urev0 calc

43808e1

update lcoh glue_code test regression values

aae1054

added hydrogen loss from degradation and input toggle to choose that …

ff5a447

…option

added LCA functions to stack.py

6e9935c

added eol_eff_percent_loss as degradation input for stack

bacfbc6

added run_LCA functionto run_electrolyzer script

c758104

added more tracking variables to stack

663700a

updated all py file import paths

2705327

updated jupyter notebook import paths

27606ee

added validators and minor fix in stack for cell type

c3b8665

added skeleton for cluster and cell classes

53f8c8a

added generic base class

81e60a8

added example skeleton for simulation config

12d25a0

added file for power-curve function templates

86b219c

Merge remote-tracking branch 'origin/BERT/schema_update' into old_for…

c9c8813

…mat_features

updated inputs for Stack in test_stack

43e4bf7

fixed merge conflicts in stack.py

2e07166

added import to fix tests to stack.py

a65c78d

removed empty PEM cell test

3b687ad

renamed test_cell to test_PEM_cell

3582018

merged develop into feature/bert_basic

366a2bb

updated validator format

2522fc0

removed files unrelated to primary feature-adds

47e0920

reformatted bert.py

c30b051

added back-in placeholder files

6216c44

removed alkaline sources in alkaline model into md file

4aa7019

johnjasa mentioned this pull request Jul 29, 2025

Fix calc_open_circuit_voltage in PEM_cell.py #83

Open

elenya-grant requested a review from johnjasa August 7, 2025 20:59

elenya-grant requested review from bayc and genevievestarke August 7, 2025 20:59

elenya-grant marked this pull request as ready for review August 7, 2025 20:59

Merge remote-tracking branch 'upstream/develop' into feature/bert_basic

f7c9884

genevievestarke reviewed Aug 29, 2025

View reviewed changes

genevievestarke requested changes Aug 29, 2025

View reviewed changes

Typographical changes

b6a5c57

johnjasa requested changes Sep 3, 2025

View reviewed changes

elenya-grant added 3 commits September 3, 2025 11:45

minor cleanups to alkaline cell and stack

348e559

removed commented out code and added docstrings

a39b015

merged stack

98efbd1

elenya-grant requested review from genevievestarke and johnjasa September 3, 2025 18:22

elenya-grant added 3 commits September 3, 2025 12:39

fixed aep datadrame in bert.py

1a4afbf

added more docstrings to stack.py

85f7847

added comments to pem.py for faradaic coefficients

6231d3a

johnjasa approved these changes Sep 4, 2025

View reviewed changes

genevievestarke approved these changes Sep 19, 2025

View reviewed changes

johnjasa merged commit 0d88b3a into NREL:develop Sep 24, 2025
3 checks passed

		from electrolyzer.tools.type_dec import FromDictMixin


		class BaseClass(FromDictMixin):

		)


		F, _, _ = physical_constants["Faraday constant"] # Faraday's constant [C/mol]

Feature/bert basic #81

Feature/bert basic #81

Uh oh!

Conversation

elenya-grant commented Feb 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

genevievestarke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

johnjasa left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

johnjasa left a comment

Choose a reason for hiding this comment

Uh oh!

genevievestarke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

elenya-grant commented Feb 26, 2025 •

edited

Loading