The ac power distribution cabinets in the CEPRI laboratory.
Smart grid technologies in China are now developing rapidly. Power flow on medium-voltage (MV) and low-voltage (LV) networks is bidirectional as a result of the integration of wind, solar, electric vehicles and other distributed energy resources (DERs). Therefore, traditional distribution network planning faces substantial challenges and must incorporate new techniques. Considerable uncertainty exists in load forecasting, planning and operation of distribution networks following the integration of DERs.
Research predicts that DERs will capture about 30% of the energy market in China by 2030. If this prediction materializes, it will create a significant change in the energy market. To support this huge transition, it is important to develop an understanding of how DERs work with existing power networks.
To satisfy this major transition before investment decisions are made, it is necessary to consider the siting and capacity of DERs together with network reliability to evaluate the impact of the DERs under different network conditions. Benchmarks for this integration have been established in America and Europe, but these are not applicable to the existing distribution networks in China, where the China Electric Power Research Institute (CEPRI) has addressed the lack of basic research.
Topology of the Benchmark System
The new benchmark is based on several physical networks in China, including some that already have distribution networks with DER integration. A typical distribution network comprises a 110/10-kV substation equipped with two 40-MVA transformers that supply seven 10-kV distribution circuits, five overhead feeders and two underground circuits with a total capacity of 60 MVA.
One of the overhead line feeders represents a circuit supplying a low load density rural area with the remaining four overhead line feeders supplying urban areas with a medium load density. The two 10-kV underground cable feeders represent the circuits that supply the urban/city center, which has a high load density.
The cost of each asset is determined on a two-part cost basis: fixed and variable. The fixed cost includes the capital cost of the assets, labor costs, fixed operating costs and ownership costs, while the variable cost includes the annual power losses based on the annual load demand.
As a first step in developing a benchmark, the characteristics of each overhead line and underground cable must be identified for the length of the network between each node. The circuit parameters in terms of the unit resistance and reactance values are required for each network section between nodes, together with the assigned maximum current rating for the asset. For the benchmark system, it is then necessary to evaluate the economical loading for each size of conductor based on capital and loss costs.
Collectively, these circuit parameters provide the key data to determine the load-transfer capability of each section of the network in terms of the economic rating and maximum thermal rating (in practice, the thermal rating is considerably higher than the economic rating):
- Asset parameters. To comply with the N-1 system security standard, the peak load on a two-quantity 40-MVA, 110/10-kV substation should not exceed 150% of the single transformer capacity, that is, 60 MVA. The network power factor (pf) is a variable because of residential loads having a pf=0.85 and commercial/industrial loads having a pf=0.95.
- Network load coincidence factor. On an actual distribution network, the load behavior is dominated by coincidence because the peak loads on the various sections of the network do not occur simultaneously. To take this into account when evaluating the network peak load, a coincidence factor of 0.75 is used.
- MV benchmark voltage levels. The nominal base voltage of the MV benchmark network needs to accommodate various network voltages that can vary from 6 kV to 12.47 kV. In practice, the HV/MV transformers installed are equipped with automatic on-load tap-changers that offer a tapping range of ±5% in 2.5% incremental steps on the primary side and ±10% in 1.25% incremental steps on the secondary side.
Model Network Application
Daily profiles of the loads and power factors for the residential, commercial and industrial customers connected to each node on the MV benchmark distribution network were based on simulations using the commercially available PSD-BPATM program developed by the power system department of CEPRI based on the BPA program. For the model, standard network conductors and cable sizes were used, with cross-sectional areas of 185-sq mm (0.29-inch) overhead line conductors and 300-sq mm (0.47-inch) underground cables.