A heterarchic hybrid coordination strategy for congestion management and market optimizaton using the DREAM framework

The increase of the proportion of renewable energy resources poses a considerable challenge to electricity grid operators in normal, critical and emergency operational circumstances of electricity grids. Combining cost-effective commercial operation of customer demand and supply and operation within (severe) distribution constraints is difficult. The DREAM project (started 1-9-2013) proposes a software agent based solution to tackle this problem, based on the analysis of diverse use cases ranging from imbalance and ancillary services market operation in the commercial and to self-healing applications in the distribution grid operating domain. Besides common hierarchical concepts to operate MV and LV grids, in the DREAM control and coordination architecture, the software agents also allow operation in a heterarchical setting, offering the possibility to determine the most appropriate leading agent, depending on the particular operational circumstances of the grid. The software components facilitating these novel control and coordination concepts are grouped according to functionality into packages in a layered structure. The package structure is depicted in Figure 1. On the basis of the LV and MV power distribution network and ICT communication networks four software packages with basic functions are defined:•In the monitoring & control package, functionality for demand response and DG-RES integration in a distributed computing setting is designed with interfaces to existing standards in the electricity domain like IEC-61850 and CIM but also more general purpose M2M protocols in the IoT (Internet of Things) to allow a higher granularity.•The persistence package features the design of a distributed data storage architecture for storage of the data with flexible requirements on storage duration, security and resilience using a distributed database approach with noSQL (not only SQL).•Different topologies for coordination algorithms, for instance PowerMatcher may be predefined for normal, critical and emergency operational strategies.•Using the set-up package the connectivity between grid components can be defined statically or emergent, depending on a grid event, using discovery protocols. Figure 1 DREAM package structureThe next layer comprises three packages to implement the use case applications:•The forecasting package, allowing consumption and production forecasts based on persistent data collected. These may include previous realizations and also external information like meteorological forecasts and prices.•The flexibility utilization package, containing tools for optimizing energy and power profiles of production or consumption devices via LP and combinatorial approaches.•The agent protocol implementation package. This package allows implementing different types of agent-based coordination in a multiprocessing, distributed computing environment. The approach allows electricity grid applications ranging from commercial optimization using VPPs via congestion management at the MV/LV level to self-healing applications with grid elementary cells coordinating with one another on a peer-to-peer or ad-hoc federation basis. The architecture has been defined using UML from a large collection of use cases specifications, class diagrams and sequence diagrams. A first version of the software architecture is available. Currently, industrial validation of the architecture, extension and implementation in Java to achieve interoperability between industrial platforms are underway. Several different real life and living lab field tests are planned together with consortium partners from industry to validate the framework.