Abstract
The extensive alterations within the structure of energy producers and consumers as well as the binding fulfilment of various energy policy objectives require changes in today’s energy systems and new methods for network planning and operational control. Since the greatest challenge for implementation of such changes is represented by the electrical grid in comparison with heat grids or gas networks, there have been various projects at the Chair of Energieverbundtechnik to enable a corresponding modelling of the electrical grids using a cellular approach in order to create future, efficient hybrid networks. Since an existing grid of the 5 kV medium voltage level has already been fully modelled in the NEPLAN [1] software program, a small partial network was taken for testing the applicability of different network simplification and reduction methods using an algorithm programmed in MATLAB [2]. The purpose is to improve the previous cell model regarding its accuracy of power flows. In order to facilitate an improvement in accuracy, the electrical lines have to be modelled correctly, which means each line needs to be represented by a 2x2 matrix according to four-pole-theory. A description of the not yet reduced overall network is achieved by setting up the nodal equations for each node within the network. If this system of equations is represented in matrix notation, the nodal admittance matrix is obtained. These matrix elements represent the coefficients of the nodal voltages within the nodal equations. The coefficients are formed by the corresponding elements in the two-port matrix. Both matrices contain information about the topology of the network. In order to derive a cell model, presently used network reduction methods were reviewed for their applicability. The overall network, which needs to be divided into cells before a reduction method can be tested, is then partitioned in a system of nodes in order to enable a systematic approach for network reduction. In the partitioning process the nodes are assigned to an external, an internal and a border system. The border system includes those nodes that connect the internal system with die external system. According to this scheme a nodal admittance matrix is set up for each cell. An appropriate solution method is used to gradually reduce the network of each cell, starting with those cells that have the lowest number of tie lines connected to other cells. The necessary parameters for a new, improved cell model are calculated in MATLAB [2]. In that equivalent model the power flows between the cells are obtained and the tie lines connecting the cells are replaced by fictitious lines, taking into account all the lines allocated within the cell. With these parameters a new equivalent network is modelled in NEPLAN [1] correspondent to the cellular approach. A load flow simulation in NEPLAN [1] allows the results to be compared with those of the original network and those of the previous cell model using an E-RLC-module. This comparison makes it possible to evaluate whether any kinds of improvements have been achieved.
Translated title of the contribution | Applicability of network reduction methodologies for a cellular approach |
---|---|
Original language | German |
Qualification | Dipl.-Ing. |
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 19 Oct 2018 |
Publication status | Published - 2018 |
Bibliographical note
embargoed until nullKeywords
- network reduction
- four-pole-theory
- two port
- load flow calculation
- Ward-method
- REI-method
- Kron-reduction