Global acceptance and adoption of distributed energy resources, such as solar, has over the years led to transformation of the electrical behavior of Low Voltage grids. Traditionally, the generation of power required to feed the loads of the entire electrical grid was centralized in large power plants placed at the High Voltage level where the low voltage grids were merely seen as loads. Nonetheless, this operational philosophy is changing due to the appearance of Decentralized or Distributed Generation based on Renewable Energy Sources such as Photo Voltaic panels or small-scale wind turbines. Additionally, a new type of loads such as single-phase connected heat pumps, electric boilers or electrical vehicles produce voltage and power profiles never seen before at the low voltage level. However, these new loads and Distributed Generation can cause technical challenges, such as sudden voltage dips and swells, unexpected and uncontrollable voltage variations, congestion or bidirectional power flows. For better prediction of such power variations, deployment of Smart Meters is being implemented at the low voltage level, which is close to 100% in some countries.
In essence, due to limited bandwidth and high delays in access to Smart Meter measurements, it is not possible in most cases to access measurements from the complete set of smart meters in a low-voltage grid area for distribution grid monitoring. Distribution system state estimation can be performed based on measurements of voltage and active and reactive power from a subset of selected smart meters. Increasing the number of selected smart meters will, on the one hand, increase the accuracy of distribution system state estimation, while on the other hand, it will degrade timeliness of the monitoring data. To harmonize this, Aalborg University researchers in Denmark: Dr. Mohammed Kemal, Dr. Ruben Sanchez, Dr. Rasmus Olsen, Dr. Florin Iov and Professor Hans-Peter Schwefel, proposed to use part of the idle time of the legacy periodic smart meter data collection for access to measurements from the subset of selected smart meters for distribution system state estimation. Their work is currently published in the research journal, Electrical Power and Energy Systems.
Basically, the researchers’ goal was to extend existing Smart Meter data access schemes by dedicated requests to voltage measurements of a subset of selected smart meters. To realize this, they analyzed the trade-off that results from increasing the number of selected smart meters for dedicated measurement access: the more- Smart Meters are used, the smaller the error of the state estimator while at the same time the timeliness of the grid state becomes worse. In particular, this trade-off was quantitatively characterized for an example of a realistic Low-Voltage grid with low bandwidth access to Smart Meters when applying the known method of weighted least squares (WLS) state estimation.
In their approach, the system context and the general approach for embedding access to selected Smart Meter data for state estimation in legacy systems were first introduced. Next, the researchers analyzed the timeliness of the accessed data as resulting from such access to selected SMs. The team then introduced the approach for distribution system state estimation, the assessment scenario and the results for the state estimation accuracy. Lastly, the results of the quantitative analysis characterize the trade-off between timeliness and state estimation accuracy were presented.
In summary, the study introduced the new concept of utilizing the spare idle periods for near real-time monitoring systems, in addition to a methodology for assessing the tradeoff between timeliness and state estimation accuracy. The methodology was applied to an example grid with 20 SMs for different realistic AMI communication topologies. In a statement to Advances in Engineering, the authors explained their work presented a novel and valuable approach that could be applied in future works researching on the same line.
Mohammed Kemal, Ruben Sanchez, Rasmus Olsen, Florin Iov, Hans-Peter Schwefel. On the trade-off between timeliness and accuracy for low voltage distribution system grid monitoring utilizing smart meter data. Electrical Power and Energy Systems: volume 121 (2020) 106090.