Utils Test for Cost Calculations by tolga-akan · Pull Request #455 · Multi-Energy-Systems-Optimization/mesido

tolga-akan · 2026-04-03T16:53:56Z

Crete a new general function in utilts test to check all the cost components in the problem
TODO:

Make the function as general as possible (initially for the all assets for heat networks, later we can extend it to other commodifities)

Modify var-opex part for:

pumppower cost calculation
influence of given electricity price profile

tolga-akan · 2026-04-09T11:02:05Z

@KobusVanRooyen Hi Kobus. I made a cost calculation test as general as possible. This function calculates the cost for assets that we use in heat networks. Assets like electrolyzer, pump, ... are not included yet. I did not incorporate influence of pumpower cost and electricity price profile yet. Shall I also include those? What do you think

# Conflicts: # tests/test_cold_demand.py # tests/test_elec_boiler.py # tests/test_end_scenario_sizing.py # tests/test_gas_boiler.py # tests/test_gas_electricity.py

KobusVanRooyen · 2026-04-16T12:37:29Z

-
-        hp_id = name_to_id_map["HeatPump_b97e"]

+        cost_calculation_test(heat_problem, results)


Will this test still pass if the _variable_operational_cost and parameters[f"{hp_id}.variable_operational_cost_coefficient"] might be zero for some reason?

Based on the question above-> what about adding a hard check in the test case that one expects results[f"{hp_id}__variable_operational_cost"] to be > then .... (some sensible/large number and not 0).

Some hard coded check like this will be needed in each test case where it is expected that a value should exist even if cost_calculation_test(heat_problem, results) is used. What do you think?

Yes. This is a good point. I will add additional check to all tests (if needed) to ensure regarding cost component is non zero (larger than a large enough positive value)

KobusVanRooyen · 2026-04-16T12:42:32Z

        Checks:
-        1. utils test: demand_matching_test, energy_conservation_test, heat_to_discharge_test,
-        electric_power_conservation_test, gas_pipes_head_loss_test
+        1. utils test: cost_calculation_test, demand_matching_test, energy_conservation_test,


Previously the TCO was checked, an not not anymore?

KobusVanRooyen · 2026-04-17T08:32:57Z

+        2. demand is matched
+        3. energy conservation in the network
+        4. heat to discharge
+        5. energy conservation over the heat and electricity commodity


why has item 5 been added? It is already mentioned in item 3?

actually that line was refering to the followign check i the test: np.testing.assert_array_less(0.0, results[f"{e_boiler_id}.Heat_source"])
np.testing.assert_allclose(
parameters[f"{e_boiler_id}.efficiency"] * results[f"{e_boiler_id}.Power_consumed"],
results[f"{e_boiler_id}.Heat_source"],
)

However, you are right that item 3 is already hinting the same. Now I removed item 5

KobusVanRooyen · 2026-04-17T08:33:33Z

+        2. demand is matched
+        3. energy conservation in the network
+        4. heat to discharge
+        5. energy conservation over the heat and electricity commodity


why has item 5 been added? It is already mentioned in item 3?

it should be "electric_power_conservation_test". now updated

KobusVanRooyen · 2026-04-17T08:35:05Z

        parameters = heat_problem.parameters(0)
        name_to_id_map = heat_problem.esdl_asset_name_to_id_map

+        cost_calculation_test(heat_problem, results)


checks needed before the function call to make sure the specific cost items exists (> than sensible value)

Check that ensures cost components of elec heat sourc elec are non-zero is added

KobusVanRooyen · 2026-04-17T08:38:50Z

-            )
-            total_opex += variable_operational_cost
-
-        np.testing.assert_allclose(


The obj is not tested anymore. Please add it to the new function "cost_calculation_test" in the test utils. But this check of obj value is not always required. So create a new input for the function "cost_calculation_test", something like check_objective_function and make the default value False.

So for this specific test case the input variable will be set to check_objective_function=True. Then for all the other test case check_objective_function does not have to be set, if the objective function value is not tested

KobusVanRooyen · 2026-04-17T09:24:17Z

                )
+
+
+def _get_esdl_scaled_cost(cost_esdl_info, uni_enum):


please add docstring for this function

docstring is added

KobusVanRooyen · 2026-04-17T09:24:34Z

+    return cost_scaled
+
+
+def cost_calculation_test(solution, results, atol=1e-8):


please add a docstring for this function

docstring is added

KobusVanRooyen · 2026-04-17T09:46:54Z

+        # Investment Cost
+        investment_cost = 0.0
+        if asset not in solution.energy_system_components.get("gas_pipe", []):
+            investment_cost_info = _get_esdl_scaled_cost(


Can we add testing for the the cost coefficient itself, to check that the value in the esdl (which investment_cost_info) is the same as the value stored in the MESIDO parameters (whic his parameters[f"{asset}.variable_operational_cost_coefficient"] in this case)

The benefit is that we are also checking if we have parsed and save the actual cost coef correctly. Can you add this to all the other cost coefficients in this function please

I added the suggested checks except "installation cost"and investemt cost of heat pipes. Because, there is no variable called installation_cost_coefficient. I could not figured out the reason but I see that parameters[f"{asset}.investment_cost_coefficient"] and investment_cost_info are different from each other.

KobusVanRooyen · 2026-04-17T09:48:19Z

+        np.testing.assert_allclose(
+            variable_operational_cost, results[f"{asset}__variable_operational_cost"]
+        )
+        total_opex += variable_operational_cost


please add a check for the obj function, but only checking it if check_objective_function=True

For a check on the objective function, you need to check the state of the asset (whether it is enabled or optional)

FJanssen-TNO · 2026-04-17T10:00:07Z

-
-        hp_id = name_to_id_map["HeatPump_b97e"]

+        cost_calculation_test(heat_problem, results)


The first tests should be related to the physics, heat_demand_matching energy_conservation and heat_to_discharge tests as these tests also tell us whether the majority of the code is working. If there already an issue occurs then the costs calucation might also fail, but that is then not the cause.
This is for all the tests

I agree. order of the cost test is changed.

FJanssen-TNO · 2026-04-17T10:01:47Z

+                for ii in range(1, len(solution.times())):
+                    variable_operational_cost += (
+                        var_op_costs_esdl
+                        * nominator_vector[ii]
+                        * timesteps_hr[ii - 1]
+                        / denominator
+                    )


No for loop for the times is need, like discussed before. also in the elif below.

for loops are removed and generalized by vector multiplication at the bottom

FJanssen-TNO · 2026-04-17T10:02:21Z

+        *transport_assets,
+    ]
+
+    pipe_classes = [


rename to gas_pipe_classes

renamed to gas_pipe_classes

FJanssen-TNO · 2026-04-17T10:04:47Z

+        elif asset in solution.energy_system_components.get("heat_pipe", []):
+            investment_cost = investment_cost_info * parameters[f"{asset}.length"]
+        elif asset in solution.energy_system_components.get("electricity_cable", []):
+            investment_cost = investment_cost_info * parameters[f"{asset}.length"]


combine these elif statements

FJanssen-TNO · 2026-04-17T10:06:28Z

+        )
+
+        # Installation Cost
+        if asset not in transport_assets:


Why are transport_assets excluded? Wouldn't it be better to check that it is 0.0 or None and thus also 0.0

Yes, that would be better. I will add check for also transport assets

FJanssen-TNO · 2026-04-17T10:11:40Z

+                        * timesteps_hr[ii - 1]
+                        / denominator
+                    )
+            elif asset in solution.energy_system_components.get("airco", []):


Missing an else statement. Now there are several elif statemetns, but those do not cover all the non transport assets, so without realising you are missing out on several assets. I believe all those assets should be calculated and checked. Also the heat demand etc.

I would also prefer an "else .... error exit" to cater for cases when new assets are added, which are not included in the test. This way we are made aware of an issue, else we might think the test is doing something where is might not be doing/testing anything

FJanssen-TNO · 2026-04-17T10:14:23Z

+                for ii in range(1, len(solution.times())):
+                    variable_operational_cost += (
+                        var_op_costs_esdl * nominator_vector[ii] * timesteps_hr[ii - 1]
+                    )


Can be moved down and generalized instead of being repeated several times, all variable opex is calculated with var_op_costs_esdl * nominator*dt/denominator.

I agree. now it is moved down

FJanssen-TNO · 2026-04-17T10:15:14Z

+        np.testing.assert_allclose(
+            variable_operational_cost, results[f"{asset}__variable_operational_cost"]
+        )
+        total_opex += variable_operational_cost


For a check on the objective function, you need to check the state of the asset (whether it is enabled or optional)

FJanssen-TNO · 2026-04-17T10:18:43Z

I think the costs check on demands that is covered in this test can also be removed once it is included in cost_calculation_test

FJanssen-TNO · 2026-04-17T10:23:12Z

test_max_size_and_optional_assets.py also contains many costs checks, some hard-coded values, those you should keep, but the once where it checks the calculation you should remove and add the cost_calculation_test

KobusVanRooyen · 2026-04-17T11:24:38Z

@tolga-akan review completed

tolga-akan

@KobusVanRooyen and @FJanssen-TNO I tried to address the comments as much as possible but it is not complete yet. There are still missing points that I could not respond due to lack of time.

tolga-akan added 4 commits April 3, 2026 18:51

wip

fed6324

test is generalized

24fbcd0

test is generalized. wip.

82db5a5

test is generalized.

4543af3

tolga-akan marked this pull request as ready for review April 9, 2026 09:28

tolga-akan requested review from FJanssen-TNO, KobusVanRooyen and jimrojerTNO as code owners April 9, 2026 09:28

style check

389393a

tolga-akan added 2 commits April 14, 2026 13:32

Merge branch 'main' into utils_test_for_cost

28bcb99

# Conflicts: # tests/test_cold_demand.py # tests/test_elec_boiler.py # tests/test_end_scenario_sizing.py # tests/test_gas_boiler.py # tests/test_gas_electricity.py

conflicts between main are resolved

b3c0dca

KobusVanRooyen reviewed Apr 17, 2026

View reviewed changes

FJanssen-TNO reviewed Apr 17, 2026

View reviewed changes

Merge branch 'main' into utils_test_for_cost

b3f6521

tolga-akan added 2 commits April 17, 2026 21:24

Merge branch 'main' into utils_test_for_cost

1ec7d01

Incomplete. Updates baned on PR reviews.

795ad43

tolga-akan commented Apr 17, 2026

View reviewed changes


		hp_id = name_to_id_map["HeatPump_b97e"]

		cost_calculation_test(heat_problem, results)

		return cost_scaled


		def cost_calculation_test(solution, results, atol=1e-8):

Conversation

tolga-akan commented Apr 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

tolga-akan commented Apr 9, 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!

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!

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!

KobusVanRooyen commented Apr 17, 2026

Uh oh!

tolga-akan left a comment

Choose a reason for hiding this comment

tolga-akan commented Apr 3, 2026 •

edited

Loading