Separate biofuel supply; add biofuel boiler #419

brynpickering · 2024-07-25T21:24:14Z

Fixes #417
Adds partial fix to #39

Checklist

Any checks which are not relevant to the PR can be pre-checked by the PR creator. All others should be checked by the reviewer. You can add extra checklist items here if required by the PR.

for more information, see https://pre-commit.ci

timtroendle

Looks good to me!

My only concern is the removal of the existing mechanism. This leads to the addition of at least one more carrier in all models that use biofuel. Do we know about the runtime and numerical scaling implications?

timtroendle · 2024-07-30T15:59:18Z

docs/model/customisation.md

@@ -15,13 +15,13 @@ You have the following three options:
 With the Calliope model in your hands, you will be able to change any model parameter, any technology specifics, and the model definition to your liking.
 This kind of customisation can be useful to get to know the model and its parameters.
 To create reliable results, we advise making manual changes only to the model definition (`example-model.yaml`) as this makes it possible to trace those changes later.
-A typical customisation here would be to change the solver from `gurobi` to an open-source solver, e.g. `cbc` (see [Calliope's documentation](https://calliope.readthedocs.io/en/v{{ calliope_version }}/user/config_defaults.html#run-configuration)).
+A typical customisation here would be to change the solver from `gurobi` to an open-source solver, e.g. `cbc` (see [Calliope's documentation](<https://calliope.readthedocs.io/en/v{{> calliope_version }}/user/config_defaults.html#run-configuration)).


Can you briefly explain this change?

hmm, that looks like it snuck in for some reason. It certainly doesn't look like it makes any sense to me.

timtroendle · 2024-07-30T16:01:10Z

docs/model/customisation.md

+
+        **biofuel_boiler**: Biofuel-consuming boiler.
+
+        **biofuel_tech_heat_to_demand**: Dummy technology to convert biofuel boiler output to a carrier that can be used to meet heat demand.


This is a unclear to me. Why is this a dummy technology?

Because it doesn't actually represent a "real" technology. It's just a modelling artefact to be able to have storage buffers per heat tech (this will be a non-issue when using calliope v0.7)

timtroendle · 2024-07-30T16:05:01Z

templates/models/techs/conversion/electricity-from-biofuel.yaml.jinja

+        essentials:
+            name: Electricity from anaerobically digested biofuel
+            parent: conversion
+            carrier_in: biofuel


This seems to remove the old approach, meaning that it won't be possible anymore to avoid modelling biofuel explicitly. I am fine with as long as it doesn't impact performance much. Do you know?

I don't know about memory use, but runtime for building and solving the optimisation problem increases by ~3% (43.8s -> 45.1s) in a test I made (UK+IRL, 1yr @ 2hr resolution, using %timeit magic).

actually, ran it again and it was the other way around. So, within margins of error, there is no change 😅

timtroendle · 2024-07-30T16:15:45Z

templates/models/techs/supply/biofuel.yaml.jinja

+        locs: [{{ id }}]
+        techs: [biofuel_supply]
+        carrier_prod_max:
+            biofuel: {{ location.biofuel_potential_mwh_per_year * scaling_factors.power }} # {{ (1 / scaling_factors.power) | unit("MWh") }}


This line may be negative for numerical scaling as it introduces a fairly large number. Previously this was solved by applying the efficiency and assuming a 6 month use period, which reduced the number by ~factor 4.

Differences in ranges:

Before

[2024-07-30 17:51:46] DEBUG Matrix range [4e-05, 1e+02] [2024-07-30 17:51:46] DEBUG Objective range [1e+00, 1e+00] [2024-07-30 17:51:46] DEBUG Bounds range [1e-05, 2e+01] [2024-07-30 17:51:46] DEBUG RHS range [8e-10, 2e+02]

After

[2024-07-30 18:04:13] DEBUG Matrix range [4e-05, 1e+02] [2024-07-30 18:04:13] DEBUG Objective range [1e+00, 1e+00] [2024-07-30 18:04:13] DEBUG Bounds range [1e-05, 2e+01] [2024-07-30 18:04:13] DEBUG RHS range [8e-10, 1e+03]

So, yes, the RHS range increases by an order of magnitude.

Not sure there is an equivalent way to handle that with biofuel as a separate carrier. I guess you just don't need carrier_prod_max if you have an hourly resource since that resource already effectively sets an annual value once it's all been added up.

Yeah, so without the annual limit, it becomes:

[2024-07-30 18:25:14] DEBUG Matrix range [4e-05, 1e+02] [2024-07-30 18:25:14] DEBUG Objective range [1e+00, 1e+00] [2024-07-30 18:25:14] DEBUG Bounds range [1e-05, 2e+01] [2024-07-30 18:25:14] DEBUG RHS range [8e-10, 5e+02]

But that would mean that biofuel must be burned whenever it's becoming available as a resource, right? That wouldn't be ideal either.

No, the resource per timestep just gives a quantity available. So long as force_resource isn't set to True it won't be forced to make use of it. It can also choose to make use of it but to store it instead of having it converted to electricity etc.

brynpickering · 2024-07-30T17:15:22Z

@timtroendle if we keep the existing approach as an option, how would you see it being included as a model file? It could be kept as its own module alongside biofuel.yaml, but then it needs a name that clearly distinguishes it (supply/electricity-from-biofuel.yaml, as a distinction from conversion/electricity-from-biofuel.yaml? Or supply/bio-electricity.yaml or similar?)

timtroendle · 2024-08-01T16:22:43Z

I am not sure whether we should keep it, but if we do a name like supply/electricity-only-biofuel.smk may work?

timtroendle · 2024-08-01T16:24:07Z

I mean the numerical scaling is an issue, but then again, it didn't seem to affect your runs, right? So maybe we can get around it without keeping the original approach.

… limit

for more information, see https://pre-commit.ci

brynpickering · 2024-08-02T09:39:34Z

@timtroendle I've reorganised the techs slightly so that we have an electrified-biofuel.yaml module that can be used in place of biofuel.yaml and electricity-from-biofuel.yaml. They are drop in replacements for each other, so the objective function value is the same using either, but the latter would be used when introducing other uses for biofuel (e.g. heating).

Separate biofuel supply; add biofuel boiler

df9c9f8

brynpickering requested a review from timtroendle July 25, 2024 21:24

[pre-commit.ci] auto fixes from pre-commit.com hooks

2fb31df

for more information, see https://pre-commit.ci

timtroendle approved these changes Jul 30, 2024

View reviewed changes

timtroendle reviewed Jul 30, 2024

View reviewed changes

brynpickering and others added 2 commits August 2, 2024 10:35

Introduce electricity-only biofuel tech; remove annual biofuel supply…

7e32a08

… limit

[pre-commit.ci] auto fixes from pre-commit.com hooks

0cb6209

for more information, see https://pre-commit.ci

brynpickering mentioned this pull request Aug 3, 2024

Add synthetic fuel production & methane boiler #424

Merged

5 tasks

timtroendle approved these changes Aug 5, 2024

View reviewed changes

timtroendle merged commit 331313c into develop Aug 5, 2024
4 checks passed

timtroendle deleted the separate-biofuel branch August 5, 2024 15:13

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Separate biofuel supply; add biofuel boiler #419

Separate biofuel supply; add biofuel boiler #419

brynpickering commented Jul 25, 2024 •

edited by timtroendle

Loading

timtroendle left a comment

timtroendle Jul 30, 2024

brynpickering Jul 30, 2024

timtroendle Jul 30, 2024

brynpickering Jul 30, 2024

timtroendle Jul 30, 2024

brynpickering Jul 30, 2024 •

edited

Loading

brynpickering Jul 30, 2024

timtroendle Jul 30, 2024

brynpickering Jul 30, 2024

brynpickering Jul 30, 2024

brynpickering Jul 30, 2024

timtroendle Aug 1, 2024

brynpickering Aug 2, 2024

brynpickering commented Jul 30, 2024

timtroendle commented Aug 1, 2024

timtroendle commented Aug 1, 2024

brynpickering commented Aug 2, 2024


		biofuel_boiler: Biofuel-consuming boiler.

		biofuel_tech_heat_to_demand: Dummy technology to convert biofuel boiler output to a carrier that can be used to meet heat demand.

Separate biofuel supply; add biofuel boiler #419

Separate biofuel supply; add biofuel boiler #419

Conversation

brynpickering commented Jul 25, 2024 • edited by timtroendle Loading

Checklist

timtroendle left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

brynpickering Jul 30, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Before

After

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

brynpickering commented Jul 30, 2024

timtroendle commented Aug 1, 2024

timtroendle commented Aug 1, 2024

brynpickering commented Aug 2, 2024

brynpickering commented Jul 25, 2024 •

edited by timtroendle

Loading

brynpickering Jul 30, 2024 •

edited

Loading