Probabilistic Grid Planning with Consideration of Dispersed Generation and Electric Vehicles

Abstract

One objective of grid planning for low voltage grids is to determine cable diameters of feeders. This usually involves simulating worst case scenarios to estimate peak loads, although the probability of such scenarios to actually occur is not known. Furthermore, today's grid planning needs to account for new technologies like dispersed generation and electric vehicles. A worst case approach, for example, would consider generation without any simultaneous load. However, in practical experience there is always a minimum load, which mitigates the effect of voltage rise due to generation. This publication presents a new approach to probabilistic grid planning under the consideration of risks of certain scenarios to occur. This methodology can help to plan grids more efficiently.

Time series based distribution grid planning approach with decentralised voltage regulation

“…With the network voltage profile being dependent on the nodal reactive power in-feed, the communication steps 1-5 are executed until the network status reaches a steady state. The results are written to a database afterwards (6). The adjustment of is done in parallel for all DGU units in the network.…”

“…nodes with the controllable units set to PV, the load flow is recalculated to determine the necessary reactive power for every controllable node to reach the requested , (4). This amount of reactive power is send to the corresponding node agent (5), which is redistributing the requested reactive power to the connected controllable units (6). The requested amount for each DGU is depending on its nominal power.…”

“…Today, these control schemes can be considered in the conventional network planning process, where only extreme scenarios (feed-back-case, load-case) are taken into account, via adjusting the constant factors of DGU. This downside remains for the probabilistic network planning approach where the probability of occurrence of scenarios can be evaluated [6] but complex controls and negotiations are difficult to model.…”