TU Berlin



The goal of our RASnet project was driven by our current systems research of ICT for low-voltage grids and its open question of how to properly make use of all available data-links between the electrical grid and its operator. To ensure a robust operation of a low-voltage grid with significant proportion of renewable energy sources within the required power-quality specifications, operators need to be able to monitor the dynamic power quality state and gain control of its generation sources, both in the time domain of milliseconds.

Within the RASnet project we have integrated three MONROE development nodes into the productive infrastructure of the wireless Internet Service Provider network of Evernet, which serves about 130 households across 5 villages in a rural area of Germany based on Linux OpenWrt-WiFi-routers, including 13 photovoltaic-systems and two windparks. We have performed several measurement trials analyzing the local 3G networks and also used the MONROE nodes for alternative data uplinks to transmit our power quality measurement data from 22 OpenWrt nodes through the alternative uplinks provided by the MONROE nodes. Our self-developed power quality monitoring system within the low-voltage grid “PowQuty” allows us to monitor the state of any Smart-Grid at different resolutions hence different amounts of monitoring data needs to be transmitted to the grid provider.

In RASnet we have analyzed the transmission possibilities of unaggregated and aggregered power quality measurements using MONROE nodes:

A. without any data processing, our PowQuty measurement system is able to collect raw voltage samples from each low voltage grid measurement location at 10 kHz resolution in voltage samples,

B. by applying a statistical post-processing stage to aggregate the raw voltage measurement samples to 1 second aggregates for voltage RMS, FFT-based frequency RMS and harmonics up to H15, all data transmitted by use of the MQTT protocol

C. by a further data-processing stage to trigger power-quality events defined by the EU norm EN 50160. Our experiments with the MONROE development nodes and the resulting 3G data upload capacities enabled us to make the first step in analyzing the usage potentials of the varying network performance given by such provider networks.

All smart grid operators have the common target of an ultra-robust data and control channel to each of their smart grid network elements. Within RASnet we used the Linux vSwitch Software-Defined Networking (SDN) capabilities on the OpenWRT nodes to analyze the potentials of integrating MONROE measurement traces as input parameters for SDN controllers.