search

znes / FlEnS

Data repository for the (Fl)ensburg (En)ergy (S)cenarios
1 stars 0 forks source link

Finalize open_eGo scenarios #3

Open wolfbunke opened 6 years ago

wolfbunke commented 6 years ago
MarlonSchlemminger commented 6 years ago

Found another error I think. https://github.com/znes/FlEnS/blob/features/ego-timeseries/open_eGo/SQ/status_quo.csv#L42 Loads for AT and LU are connected to the DE_bus_el, but should be connected to their respective countries.

MarlonSchlemminger commented 6 years ago

Just realized that this is the case for all generators of AT and LU, not only for loads.

simnh commented 6 years ago

I think this should be right, as it is the same market zone? But I don't know about the transmission restrictions, if they exist it's an error

MarlonSchlemminger commented 6 years ago

There are both powerlines DE_AT and DE_LU, so I think it is an error.

MarlonSchlemminger commented 6 years ago

I have some questions concerning the load. It seems to me that there is a discrepancy between the load curves used here and the load curves in the model_draft-tables. Some examples: This is the sum of p_sets (labelled renpass) and the sum of p calculated by etrago (labelled etrago) fo the whole network. In my opinion they should be the same if the load curves for etrago and renpass are the same. discrepancy_modeldraft_renpass

Next one: Left side is the sums of all german loads (loads not connected to busses in model_draft.electrical neighbours..) in etrago, right side is the load curve from this repository and we have a discrepancy between both of them as well. discrepancy_modeldraft_renpass2

How exactly were the load curves in the csv-files calculated and what am I missing to get the connection to the model_draft tables?

eosram commented 6 years ago

Hi Marlon, thanks for reporting!! I implemented the load curve in the csv-file for the Status quo scenario in Germany and got the data from @ulfmueller. So at that time everything should have been fine. It's the summed up synthetic load curve based on slp's. If the load curves do not match any more, there must have been an update in the database tables : (. Unfortunately to fix this the load curves have to be harmonized again based on the most recent data.

MarlonSchlemminger commented 6 years ago

Alright, thanks for your answer. What is your opinion on the other topic from 10 days ago (LU and AT directly connected to the german bus)?

ulfmueller commented 6 years ago

That is weird, since I do not think that the load curves have changed. @s3pp Do you remember how our workflow was back then? Can we (or Marlon) simply reproduce it?

eosram commented 6 years ago

I just had to normalize / scale the data to the maximum value and insert it : ).

I have now updated the load profile for Germany in the Status quo scenario using the oedb relations grid.ego_pf_hv_load and grid.ego_pf_hv_load_pq_set based on the latest release powerflow version 'v0.3.0pre1'. This could potentially fix the problem. I've sent you the code @MarlonSchlemminger .

I'm not sure why there is transmission capacity defined within the market zone. In that regard the scenario is identical with the published previous version, but it is odd. Any idea @wolfbunke ? Maybe we could just unset the nominal_value from those definitions.

MarlonSchlemminger commented 6 years ago

I looked into your script and data and the normalized timeseries barely changed, but the nominal value is completely different. 87650 is in the csv-file and your script gives me 77871. The new value now gives equal load curves for renpass and the model draft tables. I only checked for Germany so far, so I can't tell about other countries.

ckaldemeyer commented 6 years ago

Alright, thanks for your answer. What is your opinion on the other topic from 10 days ago (LU and AT directly connected to the german bus)?

This was due to a common market region (EPEX) and the "original" purpose of the model for another project..