Change algebraic sign of feed-in tariff in input files + benchmark test by SabineHaas · Pull Request #670 · rl-institut/multi-vector-simulator

SabineHaas · 2020-11-19T09:37:07Z

Fix #610

Changes proposed in this pull request:

set feed-in tariff to negative in dict_values in C0.all()
adapted C1.check_feedin_tariff() to changes
(adapted tests to changes)
added benchmark test for feed-in tariff
make sure that feedin revenue is indeed substracted from the operational costs
❌ give a hint in readthedocs about feed-in tariff sign --> looks fine in rtd: "Price received for feeding electricity into the grid." implies that it's positive.

The following steps were realized, as well (if applies):

Use in-line comments to explain your code
Write docstrings to your code (example docstring)
❌ For new functionalities: Explain in readthedocs
Write test(s) for your new patch of code (pytests, assertion debug messages)
Update the CHANGELOG.md
Apply black (black . --exclude docs/)
Check if benchmark tests pass locally (EXECUTE_TESTS_ON=master pytest)

Please mark above checkboxes as following:

Open
Done

❌ Check not applicable to this PR

_{For more information on how to contribute check the CONTRIBUTING.md.}

SabineHaas · 2020-11-19T10:09:05Z

Plans for feed-in benchmark test:

high positive feed-in tariff (feed-in tariff < electr. price) --> feed-in takes place.
negative feed-in tariff (pay for feed-in) --> excess bus is used, no feed-in.
check revenue from feed-in tariff (probably COST_TOTAL of asset in dict_values)

This could be added e.g. to AB_grid_PV but could also be a new benchmark test - how do you usually handle this / do you have an opinion about where to place the test @smartie2076 ?

smartie2076 · 2020-11-19T14:53:16Z

Yeah, your planned changes sound good. You might have to change PFs benchmark test (#613) as well with the feedin - but it can also be that no feed in occurs in that scenario.
https://github.com/rl-institut/multi-vector-simulator/blob/dev/src/multi_vector_simulator/C1_verification.py#L86

Plans for feed-in benchmark test:

* high positive feed-in tariff (feed-in tariff < electr. price) --> feed-in takes place.

Yes, that is a good one. Please include an assertion for the costs part as well as for the timeseries.

* negative feed-in tariff (pay for feed-in) -->  excess bus is used, no feed-in.

Good

* check revenue from feed-in tariff (probably `COST_TOTAL` of asset in `dict_values`)

Yeah, but also you could check COST_OPERATIONAL_TOTAL, COST_OM, COST_DISPATCH

This could be added e.g. to AB_grid_PV but could also be a new benchmark test - how do you usually handle this / do you have an opinion about where to place the test @smartie2076

Make it a seperate benchmark test anyway. I dream of aggregating the DSO cost at some point (currently all peak demand pricing months are seperate), and then we can join it there.

Do you plan to do an investment optimization as well or only a dispatch optimization?

…atch and invest optimization

… feedin tariff + tests

SabineHaas · 2020-11-26T13:14:09Z

@smartie2076 @Bachibouzouk I'm facing an issue with the benchmark test for renewable constraint (test_benchmark_constraint.py). Both tests (constraint = 0, constraint = 0.7) fail with the changes concerning the feed-in tariff. Could you please run simulations with them and check why they fail? I couldn't find the problem.

test_benchmark_constraint.py works when feedin_tariff of provider is set to -0.4
comparing lp files of simulations with feedin_tariff=-0.4 and feedin_tariff=0.4 (done for constraint=0): look the same apart from the values -0.4 and 0.4 of course
png graph: all components connected
optimizeCap=True for all assets and no maximumCap for PV

smartie2076

Hi @SabineHaas! I actually moved some functions around because I think we should not mess with the value of the feedin tariff in energy providers. With my solution, I also have no issues with the benchmark test for the minimal renewable share. Please check #685 for my proposal (based on this branch).

smartie2076 · 2020-12-04T19:00:56Z

src/multi_vector_simulator/C0_data_processing.py

+
+
+def change_sign_of_feedin_tariff_to_minus(dict_values):
+    r"""
+    Change the sign of the feed-in tariff to minus.
+
+    Additionally, prints a logging.warning in case of the feed-in tariff is entered as
+    negative value to 'energyProviders.csv'.
+
+    Parameters
+    ----------
+    dict_values: dict
+        All input data in dict format
+
+    Returns
+    -------
+    Updated `dict_values`.
+    """
+    # check feed-in tariff of each provider
+    for provider, provider_dict in dict_values[ENERGY_PROVIDERS].items():
+        feedin_tariff = provider_dict[FEEDIN_TARIFF][VALUE]
+        # change sign of feed-in tariff and add a info msg in case it was positive and
+        # a warning msg in case it was negative as might have been meant to be a
+        # negative feed-in tariff.
+        if feedin_tariff > 0:
+            logging.info(
+                f"The {FEEDIN_TARIFF} of {provider} has been changed to a negative value, which means to earn money by feeding into the grid."
+            )
+        elif feedin_tariff < 0:
+            logging.warning(
+                f"The {FEEDIN_TARIFF} of {provider} has been changed to a positive value, which means to pay for feed-in to the grid. If you intended the value to be negative (earning) please set a positive value."
+            )
+        else:
+            pass
+        dict_values[ENERGY_PROVIDERS][provider][FEEDIN_TARIFF].update(
+            {VALUE: -feedin_tariff}
+        )


I thought that one would use this part of the code to influence how the minus is interpreted:

# define feed-in sink of the DSO define_sink( dict_values=dict_values, asset_key=dso + DSO_FEEDIN + AUTO_SINK, price=dict_values[ENERGY_PROVIDERS][dso][FEEDIN_TARIFF], inflow_direction=dict_values[ENERGY_PROVIDERS][dso][INFLOW_DIRECTION], specific_costs={VALUE: 0, UNIT: CURR + "/" + UNIT}, energy_vector=dict_values[ENERGY_PROVIDERS][dso][ENERGY_VECTOR], )

This would change it to the wanted behaviour:

# define feed-in sink of the DSO define_sink( dict_values=dict_values, asset_key=dso + DSO_FEEDIN + AUTO_SINK, price=-dict_values[ENERGY_PROVIDERS][dso][FEEDIN_TARIFF], inflow_direction=dict_values[ENERGY_PROVIDERS][dso][INFLOW_DIRECTION], specific_costs={VALUE: 0, UNIT: CURR + "/" + UNIT}, energy_vector=dict_values[ENERGY_PROVIDERS][dso][ENERGY_VECTOR], )

I think this is also better, because we would otherwise change the data that we receive from the EPA to something else.

Have you checked with this what happens with the revenue stream or anything? Otherwise we need to check it in #613.

smartie2076 · 2020-12-04T19:07:40Z

src/multi_vector_simulator/C1_verification.py

        feedin_tariff = dict_values[ENERGY_PROVIDERS][provider][FEEDIN_TARIFF]
        electricity_price = dict_values[ENERGY_PROVIDERS][provider][ENERGY_PRICE]
-        diff = feedin_tariff[VALUE] - electricity_price[VALUE]
+        diff = electricity_price[VALUE] - abs(feedin_tariff[VALUE])


If you have to pay for feed-in, ie. in your csv file you enter a negative value, and use abs() here, having to pay a lot for feedin into the grid (ie. more than if you buy electricity) would suggest an infeasible problem. This problem, however, is solvable: oemof would try to route everything into excess sinks instead.

With my above suggestion from above (changing sign when defining source only), this change could be dropped.

Suggested change

diff = electricity_price[VALUE] - abs(feedin_tariff[VALUE])

diff = feedin_tariff[VALUE] - electricity_price[VALUE]

smartie2076 · 2020-12-04T19:08:09Z

src/multi_vector_simulator/C1_verification.py

+        diff = electricity_price[VALUE] - abs(feedin_tariff[VALUE])
        if isinstance(diff, float) or isinstance(diff, int):
-            if diff > 0:
+            if diff < 0:


Dropped as suggested above

Suggested change

if diff < 0:

if diff > 0:

smartie2076 · 2020-12-04T19:08:28Z

src/multi_vector_simulator/C1_verification.py

        else:
            boolean = [
-                k > 0 for k in diff.values
+                k < 0 for k in diff.values


Dropped as suggested above:

Suggested change

k < 0 for k in diff.values

k > 0 for k in diff.values

First changelog entry

b02b420

Bachibouzouk changed the title ~~First changelog entry~~ Feed-in benchmark test Nov 19, 2020

Merge branch 'dev' into fix/feedin_tariff

2f097b7

SabineHaas changed the title ~~Feed-in benchmark test~~ Change algebraic sign of feed-in tariff in input files + benchmark test Nov 25, 2020

SabineHaas added 4 commits November 25, 2020 18:57

Add benchmark test for feed-in behaviour and feed-in revenues in disp…

dfc1076

…atch and invest optimization

Add C0.change_sign_of_feedin_tariff_to_minus() for changing sign of…

2aa2377

… feedin tariff + tests

Adapt C1.check_feedin_tariff() and make sure tests check the right thing

6d5bbd0

Update changelog

6cdc345

SabineHaas self-assigned this Nov 26, 2020

SabineHaas marked this pull request as ready for review November 26, 2020 13:05

SabineHaas requested review from Bachibouzouk and smartie2076 November 26, 2020 13:05

This was referenced Nov 30, 2020

Update to MVS pypi release greco-project/pvcompare#142

Closed

Update pvcompare to newest MVS release (0.5.3) greco-project/pvcompare#151

Merged

Update input files for upcomping MVS release greco-project/pvcompare#153

Merged

Merge branch 'dev' into fix/feedin_tariff

7eb2dd7

smartie2076 reviewed Dec 4, 2020

View reviewed changes

SabineHaas mentioned this pull request Dec 7, 2020

Fix/feedin tariff alternative #685

Merged

12 tasks

SabineHaas closed this Dec 8, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Change algebraic sign of feed-in tariff in input files + benchmark test#670

Change algebraic sign of feed-in tariff in input files + benchmark test#670
SabineHaas wants to merge 7 commits intodevfrom
fix/feedin_tariff

SabineHaas commented Nov 19, 2020 •

edited

Loading

Uh oh!

SabineHaas commented Nov 19, 2020

Uh oh!

smartie2076 commented Nov 19, 2020

Uh oh!

SabineHaas commented Nov 26, 2020

Uh oh!

smartie2076 left a comment

Uh oh!

smartie2076 Dec 4, 2020

Uh oh!

smartie2076 Dec 4, 2020

Uh oh!

smartie2076 Dec 4, 2020

Uh oh!

smartie2076 Dec 4, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	diff = electricity_price[VALUE] - abs(feedin_tariff[VALUE])
	diff = feedin_tariff[VALUE] - electricity_price[VALUE]

Conversation

SabineHaas commented Nov 19, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

SabineHaas commented Nov 19, 2020

Uh oh!

smartie2076 commented Nov 19, 2020

Uh oh!

SabineHaas commented Nov 26, 2020

Uh oh!

smartie2076 left a comment

Choose a reason for hiding this comment

Uh oh!

smartie2076 Dec 4, 2020

Choose a reason for hiding this comment

Uh oh!

smartie2076 Dec 4, 2020

Choose a reason for hiding this comment

Uh oh!

smartie2076 Dec 4, 2020

Choose a reason for hiding this comment

Uh oh!

smartie2076 Dec 4, 2020

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

SabineHaas commented Nov 19, 2020 •

edited

Loading