4634 add fuel ion equilibrium constraint by grmtrkngtn · Pull Request #4639 · ukaea/PROCESS

grmtrkngtn · 2026-10-01T15:27:17Z

This makes fuel-ion equilibrium an explicit equality constraint in PROCESS.

Previously, the required plasma fuelling rate was calculated and set internally from the fusion rate and burnup fraction. This hides one of the coupled relationships required for PROCESS to arrive at a self-consistent solution.

The change separates:

molflow_plasma_fuelling, representing the plasma fuelling rate;

molflow_plasma_fuelling_equilibrium, calculated from the plasma state.

Constraint 93 then explicitly enforces fuel-ion equilibrium by requiring:

molflow_plasma_fuelling = molflow_plasma_fuelling_equilibrium

The equilibrium requirement is calculated in the physics model, while the constraint simply compares the two quantities.

Motivation

This is part of the work required to introduce a PROCESS solution mode for uncertainty-quantification evaluations.

For UQ, we want to evaluate PROCESS at sampled inputs without performing a full design optimisation. PROCESS must therefore be able to solve the coupled equations required for a physically self-consistent plant state independently of the optimisation problem.

Making fuel-ion equilibrium an explicit closure equation is one step towards separating:

solving the PROCESS model for a self-consistent state, from
optimising the design.

This will allow a future solution mode to solve the required internal equilibrium equations for each UQ sample while keeping the sampled/design quantities fixed.

Docs

I have re-written the solver docs page to explain the role this plays in PROCESS, and how we view what a PROCESS solution is.

codecov-commenter · 2026-10-01T15:34:44Z

Codecov Report

❌ Patch coverage is 60.00000% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 49.75%. Comparing base (2029202) to head (f7dfcb8).
⚠️ Report is 13 commits behind head on main.

Files with missing lines	Patch %	Lines
process/models/physics/physics.py	0.00%	3 Missing ⚠️
process/core/io/plot/summary.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4639      +/-   ##
==========================================
- Coverage   49.92%   49.75%   -0.17%     
==========================================
  Files         151      151              
  Lines       30225    30309      +84     
==========================================
- Hits        15089    15080       -9     
- Misses      15136    15229      +93

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

timothy-nunn · 2026-10-02T07:38:28Z

@grmtrkngtn could you possibly explain (and justify) why the regression tests have changed results here? Its not immediately obvious why this has happened

grmtrkngtn · 2026-10-05T07:59:43Z

@grmtrkngtn could you possibly explain (and justify) why the regression tests have changed results here? Its not immediately obvious why this has happened

I noticed that too Tim, although they didn't seem to be significant changes - I can't explain it as these changes are not active in any of the regression tests!

timothy-nunn

With these current changes it strikes me that the new constraint should always be active and that a sensible molflow_plasma_fuelling should be specified in each input file (or a sensible default applied).

There is also an inconsistency between models whether molflow_plasma_fuelling or molflow_plasma_fuelling_required is used.

If we use molflow_plasma_fuelling in the models, then I think the new iteration variable and constraint will always need to be active (and that should be documented).

If we use molflow_plasma_fuelling_required in the models, then the new iteration variable is not useful (unless I am missing something).

timothy-nunn · 2026-10-05T08:30:35Z

@grmtrkngtn this will be where the changes have come from. Because the iteration variable and constraint is not on molflow_plasma_fuelling will always be the default value, whereas before it was molflow_plasma_fuelling_required which was calculated.

I've now set a default value that is roughly representative of a large tokamak / DEMO LAR case.

It is worth noting that solutions can be quite sensitive to this value. That is expected, since the assumed fuelling rate affects the plasma particle balance. In optimisation mode users can activate ixc = 178 together with icc = 93 and allow PROCESS to solve for the fuelling rate. Otherwise, an appropriate value needs to be specified in the input file.

timothy-nunn · 2026-10-05T08:30:55Z

            self.data.physics.figmer,
            _fusrat,
-            self.data.physics.molflow_plasma_fuelling_required,
+            self.data.physics.molflow_plasma_fuelling_equilibrium,


How come in vacuum we are using molflow_plasma_fuelling but here we use molflow_plasma_fuelling_equilibrium?

phyaux() calculates the equilibrium fuelling rate, the fuelling rate required to sustain fusion.

molflow_plasma_fuelling is the actual fuelling throughput used by downstream engineering models such as the vacuum system.

Constraint 93 enforces equality between them at a solved point, but they should remain distinct during the solve

… directly

…fter fast particle calculation

This reverts commit 8dc8e88.

Co-authored-by: Timothy <75321887+timothy-nunn@users.noreply.github.com>

Co-authored-by: Christopher Ashe <91618944+chris-ashe@users.noreply.github.com>

chris-ashe · 2026-10-12T09:18:40Z

@grmtrkngtn Am seeing commits from main on the tracked changes. Please rebase and then I will review

…date associated outputs, labels and tests.

chris-ashe · 2026-10-12T13:20:50Z

+
+At the heart of a PROCESS solution is a plasma operating in a self-consistent equilibrium state. Its state cannot be chosen arbitrarily: plasma temperature, density, heating, fusion reactions, radiation, transport losses and fuelling are all coupled.
+
+The principal plasma state is described by the electron temperature and density,


Is the solve here specifically for the volume averaged temperature and density?

chris-ashe · 2026-10-12T13:21:23Z

+The principal plasma state is described by the electron temperature and density,
+
+$$
+T_e,\qquad n_e.


If the above is true, then:

Suggested change

T_e,\qquad n_e.

\langle T_e \rangle ,\qquad \langle n_e \rangle.

chris-ashe · 2026-10-12T13:24:43Z

+
+This gives an important distinction:

 $$


Not sure what is happening here with the formatting but it is broken here in the Github viewer

chris-ashe · 2026-10-12T13:26:40Z


-Several possible figures of merit are available, all of which are in the 
-source file `evaluators.f90`. 
+The switch `i_figure_merit` selects the objective. A positive value indicates minimisation, while a negative value indicates maximisation.


I think it maybe be good here and throughout to give some reference to the controlling enum objects for the switches

chris-ashe · 2026-10-12T13:27:37Z

    )


+@ConstraintManager.register_constraint(93, "", "=")


Suggested change

@ConstraintManager.register_constraint(93, "", "=")

@ConstraintManager.register_constraint(93, "particles/s", "=")

chris-ashe · 2026-10-12T13:28:01Z

+    178: IterationVariable(
+        "molflow_plasma_fuelling",
+        "physics",
+        1.0e20,


Suggested change

1.0e20,

1.0e18,

chris-ashe · 2026-10-12T13:28:12Z

    "f_len_sol_power_decay_inboard_outboard": InputVariable(
        "physics", float, range=(0.01, 2.0)
    ),
+    "molflow_plasma_fuelling": InputVariable("physics", float, range=(1.0e20, 1.0e23)),


Suggested change

"molflow_plasma_fuelling": InputVariable("physics", float, range=(1.0e20, 1.0e23)),

"molflow_plasma_fuelling": InputVariable("physics", float, range=(1.0e18, 1.0e23)),

Implementation of beta fixed point solve, based on Jon M work.

6c34f15

grmtrkngtn self-assigned this Oct 1, 2026

grmtrkngtn requested a review from a team as a code owner October 1, 2026 15:27

grmtrkngtn added the Convergence Solver or convergence-related problems label Oct 1, 2026

grmtrkngtn linked an issue Oct 1, 2026 that may be closed by this pull request

Add fuel-ion equilibrium constraint #4634

Open

timothy-nunn assigned timothy-nunn and unassigned grmtrkngtn Oct 2, 2026

timothy-nunn requested a review from a team October 2, 2026 07:36

timothy-nunn reviewed Oct 5, 2026

View reviewed changes

grmtrkngtn and others added 18 commits October 5, 2026 15:47

Revert caller.py to main

707589c

Derive plasma beta explicitly from current plasma state

8e6cb33

Remove constraint 1 as it is not required

8dc8e88

Remove beta_total_vol_avg as ixc 5 as it is now calculated in physics…

f01251e

… directly

Tidy up removal of constraint 1

ee85479

Remove icc = 1 and ixc = 5

0b1a95b

Remove icc = 1 and ixc = 5

6478096

Remove icc = 1 and ixc = 5

c9362c1

Remove icc = 1 and ixc = 5

59315f5

Remove icc = 1 and ixc = 5

144fe7d

Tidy up beta calculations into plasmabeta.run, and move beta update a…

122540b

…fter fast particle calculation

Revert "Remove constraint 1 as it is not required"

7cbaa0f

This reverts commit 8dc8e88.

Add back in descriptions for icc = 1

3368eb6

Remove reference to itv 5

01d3e39

Remove references to plasma beta consistency equations

b90996c

Update plasma_beta docs to reflect new status of icc = 1

a6bc5b0

Tidy up references to icc = 1 and ixc = 5

d449469

Update process/data_structure/physics_variables.py

ed7e2af

Co-authored-by: Timothy <75321887+timothy-nunn@users.noreply.github.com>

removed electron density as ixc to correctly form system of equations

da86937

grmtrkngtn mentioned this pull request Oct 8, 2026

Derive plasma beta explicitly from the current plasma state #4640

Open

grmtrkngtn and others added 6 commits October 8, 2026 13:54

copy latest large tokamak eval from regression test over

b8b906a

remove icc = 1 and ixc 5 and beta_total_vol_avg

be3e518

remove icc = 1 and ixc 5 and beta_total_vol_avg

61bf24a

remove icc = 1 and ixc 5 and beta_total_vol_avg

79917aa

Update documentation/source/physics-models/plasma_beta/plasma_beta.md

55b1f4d

Co-authored-by: Christopher Ashe <91618944+chris-ashe@users.noreply.github.com>

Moved fast alpha beta calculation into beta class

4508b31

grmtrkngtn requested a review from a team October 9, 2026 15:06

jonmaddock and others added 18 commits October 12, 2026 11:48

Add fuelling rate constraint

246286a

Add fuelling rate as optimisation parameter

60e8839

Remove calculation from constraint 93 to phyaux. Rename variables, up…

32ab25b

…date associated outputs, labels and tests.

Add ixc 178, initial value, and icc = 92 to DEMO LAR

311efd2

Add ixc 178, initial value, and icc = 92 to ST

16b94d9

Add default value for molflow_plasma_fuelling

c0a74c9

Add icc = 93 and value for molflow_plasma_fuelling

4d2a736

Update n_euqality_constraints for inclusion of icc = 93

18e3d34

Add icc = 93, ixc = 178, up n_equality_contraints

15aa20f

Remove icc = 1 and ixc = 5

d75ccc5

add in constraint 93

c656cad

Add in constraint 93 and solve with electron density

2d13bd3

Add in constraint 93 and solve with electron density

14c2c3b

Update for icc = 93

1915e8c

Update for icc = 93, remove beta consistency

dc1690b

Update for icc = 93, remove beta consistency

f7bed33

Update to describe governing equations of PROCESS solution

d16ae9f

Revamp of solver guide with icc = 93

350abb5

grmtrkngtn force-pushed the 4634-add-fuel-ion-equilibrium-constraint branch from 9702aff to 350abb5 Compare October 12, 2026 10:52

chris-ashe requested changes Oct 12, 2026

View reviewed changes


		At the heart of a PROCESS solution is a plasma operating in a self-consistent equilibrium state. Its state cannot be chosen arbitrarily: plasma temperature, density, heating, fusion reactions, radiation, transport losses and fuelling are all coupled.

		The principal plasma state is described by the electron temperature and density,

	T_e,\qquad n_e.
	\langle T_e \rangle ,\qquad \langle n_e \rangle.

	@ConstraintManager.register_constraint(93, "", "=")
	@ConstraintManager.register_constraint(93, "particles/s", "=")

	"molflow_plasma_fuelling": InputVariable("physics", float, range=(1.0e20, 1.0e23)),
	"molflow_plasma_fuelling": InputVariable("physics", float, range=(1.0e18, 1.0e23)),

Conversation

grmtrkngtn commented Oct 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Docs

Uh oh!

codecov-commenter commented Oct 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

timothy-nunn commented Oct 2, 2026

Uh oh!

grmtrkngtn commented Oct 5, 2026

Uh oh!

timothy-nunn 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!

chris-ashe commented Oct 12, 2026

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!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

grmtrkngtn commented Oct 1, 2026 •

edited

Loading

codecov-commenter commented Oct 1, 2026 •

edited

Loading