In this blog we continue our overview on the use of M2M communications in microgrids (see https://massm2m.wordpress.com/2012/09/11/m2m-communications-in-microgrids/).
As outlined in the previous blog, the future smartgrid is envisioned as a network of microgrids. Microgrid is a small-scale electrical network consisting of localized distributed generators, storage capacities and loads, interconnected by low-voltage (LV) cables.
Distributed generators (DGs) within microgrid typically exploit renewable energy sources (such are photo-voltaic and wind generators) whose behavior cannot be controlled, posing new challenges with respect to balancing the generated power and user loads. Basically, when controlling the DGs within microgrid, the imperative is that their output voltages and frequencies should match.
In traditional grids, the voltage and frequency outputs of generators are regulated using the so-called droop control. Droop-control is essentially a proportional control algorithm that is executed locally at each generator, driving output voltage and frequency based on the locally measured active and reactive powers. Loads and generators in traditional grids are interconnected using high-voltage cables, which have much greater reactance then resistance. Taking into account this fact, it can be shown that, when the active and reactive power that are flowing into the cable from the generator are expressed through the generator’s voltage (and assuming that the power angle is small), the active power depends on the generator’s frequency and the reactive power depends on the generator’s voltage. The overall result is that, in traditional droop control, the frequency is regulated by the local measurements of the active power, and voltage is regulated by the measurements of the reactive power.
Traditional droop control cannot be used directly in AC microgrids[*], as resistance of LV cables dominates over reactance. One way to address this issue is to modify the droop control such that control algorithm includes the information exchanged among DGs[**] using communication network.
In a recent paper , the authors apply the paradigm of distributed consensus algorithms, exploited in their earlier works (see previous blog, https://massm2m.wordpress.com/2012/09/11/m2m-communications-in-microgrids/), to conduct series of local exchanges among DGs, in order for all DGs to learn what is the global state of active and reactive power. Traditional droop control is then augmented to include this information, such that frequency controller includes a term proportional to the differences between the desired and the actual active and reactive power, and the voltage controller uses a corresponding integral term. Using small signal analysis, the authors show that improved stability of the microgrid is achieved with respect to the traditional drop control.
The authors of  consider a similar setting, but the focus of their work is the impact of communication impairments, such are packet loss and delay, on the (modified) droop control, rather than on the control algorithm itself. The key observation is that outdated information about remote measurements, when combined with locally obtained measurements, can adversely affect the performance of the droop control. In order to neutralize this effect, the authors suggest a scheme in which locally obtained signal is delayed using the same delay distribution of the remote signal. The local delay is realized using Smith predictor and it is statistically the same as the delay of the remote signal. Finally, the authors present the simulation results obtained in an example microgrid with two DGs, showing that in the proposed scheme a more stable active power output is achieved, compared to the case when there is no local delay.
Recently, our group started collaboration on the same topic with a group at Department of Energy Technology at our university, with an aim of designing of robust and simple communication algorithms in the framework of DG control, both in AC and DC microgrids. The initial results are encouraging, however, we leave their presentation for our future blogs.
 H. Liang et al, “Decentralized Inverter Control in Microgrids Based on Power Sharing Information through Wireless Communications”, to be presented at IEEE GLOBECOM 2012
 S. Ci et al, “Impact of Wireless Communication Delay on Load Sharing Among Distributed Generation Systems Through Smart Microgrids”, IEEE Wireless Communications, June 2012