Anaerobic Digester Re-Scaling #1535
Conversation
| @@ -940,6 +940,9 @@ def model(self): | |||
| m.fs.unit.liquid_phase.properties_in[0].TSS | |||
| m.fs.unit.liquid_phase.properties_out[0].TSS | |||
|
|
|||
| iscale.calculate_scaling_factors(m) | |||
| iscale.set_scaling_factor(m.fs.unit.liquid_phase.heat, 1e-6) | |||
There was a problem hiding this comment.
WHy did you need this change?
There was a problem hiding this comment.
This change is not needed. It may have been necessary at some point while I was working on this PR, but I'll undo it
| iscale.set_scaling_factor(self.KH_co2, 1e2) | ||
| iscale.set_scaling_factor(self.KH_ch4, 1e2) | ||
| iscale.set_scaling_factor(self.KH_h2, 1e2) | ||
| iscale.set_scaling_factor(self.hydraulic_retention_time, 1e-6) | ||
| iscale.set_scaling_factor(self.volume_AD, 1e-2) | ||
| iscale.set_scaling_factor(self.volume_vapor, 1e-2) | ||
| iscale.set_scaling_factor(self.liquid_phase.rate_reaction_generation, 1e4) | ||
| iscale.set_scaling_factor(self.liquid_phase.mass_transfer_term, 1e2) | ||
| iscale.set_scaling_factor(self.liquid_phase.rate_reaction_extent, 1e4) | ||
| iscale.set_scaling_factor(self.liquid_phase.enthalpy_transfer, 1e0) | ||
| iscale.set_scaling_factor(self.liquid_phase.volume, 1e-2) | ||
| iscale.set_scaling_factor(self.electricity_consumption, 1e0) |
There was a problem hiding this comment.
Shouldn't these be handled in the scaler class?
There was a problem hiding this comment.
Andrew mentioned that we should not mix and match the new and old scaling tools. In other words, the scaler class should not make use of iscale. This is the reason why I've moved all of the iscale.set_scaling_factor to the calculate_scaling_factors function (in this PR and in others), which won't be called when exclusively using the new scaling tools.
If your question is why don't I set these default scaling factors in the ADScaler class, it's because I figured these would make more sense as user-defined scaling factors. I know I did do some testing with having default scaling factors for these variables in the new scaler class, but I forget how thoroughly I played around with this. Currently, the only variable scaled by default in the new scaler class is volume.
| @@ -56,71 +56,71 @@ def test_solve(self, system_frame): | |||
| 0.24219, rel=1e-3 | |||
| ) | |||
There was a problem hiding this comment.
Would it make sense to check the condition number at the flowsheet level between old scaling approach and new?
There was a problem hiding this comment.
Then, every time another PR impacts scaling at this flowsheet level, we can track whether condition number improves or not as we go.
Summary/Motivation:
Adds the new IDAES scaling routine (scaler objects) to the anaerobic digester unit model. This should be merged after #1519 and #1530 are merged in.
Changes proposed in this PR:
Legal Acknowledgement
By contributing to this software project, I agree to the following terms and conditions for my contribution: