Managing Volatility in Distribution Networks with Active Network Management

The traffic light concept of the German Association of Energy and Water (BDEW) to coordinate and market power in the distribution system level defines different mechanisms to ensure network stability in dependence of the criticality of the system state. But what is a critical network condition and how is it possible to detect a critical network condition timely, effective, and possibly also at distinct locations? Especially when facing the insufficient number of measuring devices at medium and low voltage levels. A possible solution for this problem are modern techniques for network state estimation, which determine a realistic picture of the actual network conditions with the aid of a limited number of real time measurements. This paper shows how network state estimation can be applied to detect critical network states and how the right conclusions for grid stability can be derived with respect to the traffic light concept of the BDEW. CHALLENGES TO IMPLEMENT THE BDEW TRAFFIC LIGHT CONCEPT The traffic light concept is designed to coordinate grid and market on distribution level and defines several approaches to ensure network stability depending on the severity of the network state. The question is what is a critical network state and how can such a critical network state be predicted early enough in an efficient manner with sufficient localization. Current situation in German distribution grid Looking more closely on the degree of transparency in the German distribution grid, we experience a well monitored – means measured 110 kV level while the 20 kV level and below in most cases lacks any information. Measuring equipment is mainly limited to the large transformer substations while the medium voltage grid is not automated at all. However, the growing number of distributed energy resources is leading to unclear power flow directions and volatility. The traffic light concept is an approach to manage those problems. Possible solution with network state estimation Possible counter measures are new concepts to estimate the network state despite the limited number of real-time measurements [1]. These algorithms use active and reactive power measurements as well as all available current and voltage magnitude measurements in combination with the static network model, load profiles and generation forecast to calculate the most probable network state given the input data. The algorithms incorporate repeated power flow executions. The distributed energy resources (wind, photovoltaic and bio mass ...) should be aggregated per distribution substation and the respective power supply should be either calculated based on measured reference systems or modelled based on generation forecast data. With this approach a detailed view of the network state can be obtained independent of the availability of real-time measurements at each single station. In addition the power flow based algorithms provide the information about availability of flexibilities on market level (yellow traffic light). As at the same time those counter measures can be simulated and evaluated, it is possible to initiate further measures as early as possible (red traffic light). Thus the network state estimation provides a sound base to identify critical network states and to calculate and evaluate counter measures. CALCULATION OF THE EXPECTED NETWORK STATE The following sections show how modern algorithms for network state estimation can be used to detect in a systematic way whether the network state is reaching a critical state and consequently to calculate counter measures. The network state estimation is typically an integrated component of the power control system for distribution grids. This application calculates network state information in real-time and is re-calculating whenever the network topology changes or significant measurement changes occur. It calculates voltage magnitude and angle at each network node. The following information is presented to the operator as comprehensive results: · Alarms to indicate critical network states that need immediate attention. · Limit violations to indicate areas of problems · Total active and reactive power as well as power factor as key figures representing the energy exchange with the transmission grid · Voltage reserve to indicate available flexibilities Network state estimation algorithm The network state estimation algorithm in [1] works with measurement areas and load groups. Measurement areas are network subsections bordered by measurements or open switches. Load groups are loads with similar weighting factors within one measurement area. The estimation problem is solved using Weighted Least Squares (WLS) approach defined as a minimization of the CIRED Workshop Helsinki 14-15 June 2016