e2nIEE / simbench

Electrical Power System Benchmark Dataset directly usable with pandapower (use the project website for other tools)

Home Page:https://simbench.de/en/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Bug: Storage Profiles

SteffenMeinecke opened this issue · comments

This issue is discussed along with the example grid "1-LV-semiurb4--2-no_sw".

These are the hypotheses on which I am based:

  • The storages of this example grid include the PV plants and the batteries
  • The storage profiles, e.g. "Storage_PV5_H0-A", is generated by maximizing the self-consumption (in documentation section 3.5.5, 1st subsection). To maximize the self-consumption, the corresponding load profile must be taken into account. Therefore, to clarify the correlation, the storage profile names include the corresponding load profile name.
  • The storage profiles do not include the corresponding load profiles
  • storages and loads are with consumer based singing system

This behavior is expected for the active power of the storages:

  • the storage is used only to shift the PV infeed which exceeds the load consume to points of time in which the PV infeed is less than the load consume => the storage (PV and battery) always acts as generator and never as load. To let the profile be positive, the active power in the storage table is negative.
  • If the infeed of a storage exceeds the corresponding load consume, the PV power must exceed the load consume and SOC must already be 100% => abs(storage power) > abs(load power)
  • If SOC > 0%, the storages do infeed as much as possible => abs(storage power) = abs(load power)
  • If SOC = 0%, the PV power is less than the load consume => abs(storage power) < abs(load power)
  • Neglecting the energy difference of SOC at the beginning of the full year time series and at the ending, the energy infeed of the storages must be the same as the PV plants would have without batteries

Revealing the bug:

Refering to the last behavior expectation, the sum of profile "Storage_PV5_H0-A" (1st storage profile of the example grid) is 4548.6 but the sum of profile "PV5" is 2703.8.
These two graphs show the time series calculation for the first 4 days revealing that the storage power seems to be the corresponding load with a factor of around -2. If that is the case and the assumptions and expactation made above are ok, the storage profile values are to high and thus do not fit.
1-LV-semiurb4--2-no_sw 1  Speicher   Last Tag1-4
1-LV-semiurb4--2-no_sw Leistungssummen Tag1-4