Due to the recurrent change of real power systems, e.g. the increase of decentral energy resources, as well as new tools, new generations of suitable benchmark data sets are necessary. In contrast to the broad variety of existing grid models, SimBench will be an extensive data pool of combinable grid models for extra high to low voltage levels, including time series for load, generation and storages. This data set will ensure that a large number of addressed applications can be suitably tested and compared.

If extreme grid structures are needed by use cases, we will provide “extreme” grids or “extreme” parameter combinations, if possible.

We collected about 40 use cases dealing with grid analysis, planning und operation.

Our SimBench-process is iterative with feedback loops for improvements and corrections. Not only the research institutes of our consortium but also different important German DSOs provide their expertise in the grid model development. In addition we will provide usage advices to prevent from false usage as best as possible.

The four research partners have rich experiences in the fields of grid analysis, planning and operation and make use of it in SimBench project. Moreover we are part of the German research network energy, where we actively participate in (online) meetings.

Yes.

We will focus on static modelling and intend to set up a full year time series with a temporal resolution of quarter hours.

We do not aim at modelling the entirety of existing grids via generating representative grids, but focus on relevant use cases. The more network classes are created, the better not only the multiplicity of real networks can be mapped. We can also better generate different network challenges and conditions that may be necessary for the relevant use cases. On the other hand a high number of classes would reduce the comparability of different solutions. This is because investigations based on different data and assumptions cannot be properly compared.
In summary, we will create as many grid classes as needed and as few as possible. We expected these grid numbers: EHV:1, HV:2, MV:4, LV:6

We will provide some time series for comparisons and to describe fundamental behavior of storage devices. In this way the users of our data base will be able to benchmark their better solutions against our reference.

SimBench grids focus is benchmark grids for testing purposes.

At the current development stage we develop an independent data structure based on csv files. We will provide this data structure later on as well as models for pandapower, DIgSILENT PowerFactory and Integral.

We aim at providing a freely accessible data base with the right to use it for testing your grid solutions. The determination of the exact license will be soon.