Multiple generators without rea_geom_new

[IlkaCu]

Analyzing data from table supply.ego_dp_res_powerplant and model_draft_ego_dp_supply_res_powerplant shows that 154718 generators are lacking information on rea_geom_new.
This applies to generators from all sources and all voltage levels which are spread evenly all over Germany.

@Ludee: Any ideas why this happens and how to fix this bug?

[Ludee]

I will have a look later this day. But the timing is bad, I will be on holiday for 2 weeks starting tomorrow. I can not promise to fix it today, my workload is too big atm.
But this def. falls into my purview ;)

[IlkaCu]

As I didn't hear anything from Ludee I assume he didn't manage to have a closer look. @wolfbunke Do you have any ideas how to fix this bug?
If not I would suggest as a hotfix to fill all empty cells of 'rea_geom_new' with entries from column 'geom'.

[wolfbunke]

Due to saving time to the release I agree with your fix. Let's use the column comment to add the information that the rea_geom _new is from geom.
Like:

Update table..
   set rea_geom_new = geom,
   comment = comment || 'add original geom to rea_geom_new' 
Where rea_geom_new is null;

And create a ticket for a revision of the rea allocation.

[IlkaCu]

Could you give me a hint where would be the best place to add this and the Create Index statement in the REA scripts?

[wolfbunke]

Between 'rea/ego_dp_rea_results.sql' and ego_dp_powerflow_assignment_otgid.sql maybe at the beginning of the ego_dp_powerflow_assignment_otgid.sql file in order to save a additional file ?

[nesnoj]

You gonna start the run today?
If not, I'd like to give @IlkaCu idea a quick review..

[IlkaCu]

I won't start the run before tomorrow - more likely Thursday.

[IlkaCu]

If not, I'd like to give @IlkaCu idea a quick review..

@nesnoj : Feel free to share your thoughts and comments. ;-)

[nesnoj]

It reveals that in Status Quo only 2.7% of the RES genos are affected but 39.5% (37.5GW) of entire RES capacity. Without EHV/HV level (only v_level 5..7 where geoms are crucial), this results in 13.7% (13 GW) This is quite a lot.

This one

SELECT scenario, generation_type, generation_subtype, rea_flag, count(*), round(sum(electrical_capacity)/1000) as cap_MW
  FROM supply.ego_dp_res_powerplant
  WHERE rea_geom_new is NULL AND version='v0.3.0' AND voltage_level in (5,6,7)
  GROUP BY scenario, generation_type, generation_subtype, rea_flag
  ORDER BY scenario, generation_type, generation_subtype, rea_flag;

reveals that a large share of the affected datasets could not be allocated in their original allocation methods (rea_flag) (e.g. M1-1) and also the fallback (e.g. M4_rest) did not work...right @wolfbunke @Ludee?

The allocation method bpmn:

I know this is really annoying and takes some time to fix but not to fix means we have many REA which remains at their original (energymap's) position.
For the EHV/HV this might be acceptable but might a serious problem for the distribution grids.
@boltbeard @gplssm
-> WEC, PV(ground mounted) and PV(roof mounted) remain located at the same position within on MVGD..

I'm wondering whether this is a new bug (didn't you update the REA genos recently, @wolfbunke?) or has it been there all the time? @Ludee