A heterarchic hybrid coordination approach for congestion management using the DREAM framework

Software agent-based strategies using micro-economic theory like PowerMatcher[1] have been utilized to coordinate demand and supply matching for electricity. Virtual power plants (VPPs) using these strategies have been tested in living lab environments on a scale of up to hundreds of households. So far, the coordination configuration of a VPP is fixed in these settings. The DREAM [2] framework architecture uses heterarchies to make parts of a VPP flexible in coordination strategy depending on the current operational grid status. In this way, a sub-VPP, serving one coordination objective, can decouple from and couple to an overarching VPP with another coordination objective dynamically. In this paper a grid congestion simulation with an overarching VPP coordinating demand and supply for electricity market optimization [3] and a sub-VPP reacting to a heat-pump congestion event in winter and a PV overproduction event in summer is described. The simulation was run in a static simulator [4]. The LV-segment consisted of ‘flameless’ residential areas with well-insulated homes with primarily heat pumps for heating and some renovated homes with local gas-fired co-generators of heat and electricity. Households additionally had solar cells, batteries and EV charging units. The goal of the additional coordination sub-VPP was to solve grid stability issues like congestion due to heat pump loads in winter and overproduction by PV in summer in this physical part locally, while the rest of the cluster remained unaffected and still optimizing for the commercial goal. The results were analyzed in terms of infringement of comfort parameters and performance in adapting the flexible load and generation. It appeared, substantial load shedding and load shifting of devices is possible to show the synergy in solving the grid stability issues evenly sharing the discomfort to the individual heating devices. By changing their charging strategy, the new algorithm also showed heat storage and electricity storage devices providing additional support. INTRODUCTION The simplest definition of flexibility can be given as a bandwidth around a required momentary power value of a connection or a device. The flexibility can be discriminated in a part to be used in normal operation, that, for example, might be expressed in a PowerMatcher bid-curve [1]. This part may be utilized by a commercial party in normal grid operation to balance the portfolio. The timescale for it to be mobilized is in the order of tens of seconds to minutes. Another part can be utilized as a contingency reserve by a grid operator to prevent the operation of the power system to go into a critical situation. The second part typically has to be used in the seconds scale. In micro-economic terms, primary processes, using electrical energy, have different utilities for flexibility on these different time scales. This economic utility can be translated in a price that can be used for coordination.