This pooling function of virtual power plant (VPP) enables the creation of marketable portfolios of separate energy and flexibility offers that actually are too small to be marketable when taken individually. The pooling function thus qualifies little amounts of generation and storage or consumption flexibility for market participation and allows them to participate in the regional, national and European markets. Furthermore, the aggregator or virtual power plant functions as an interface between network oriented needs and flexibility management on the generation and demand side.
The increasing dynamics and complexity of energy supply structures are resulting in new requirements for the architecture, scalability and flexibility of technical solutions. A special challenge arises regarding the multitude of interfaces with distributed communication systems. On the other hand side, the optimized management of decentralized generation, storage and consumption units gains importance. Taking into account these general assumption as well as the energy economy’s general requirements, the E-Energy model region RegModHarz has developed a multi-layered architecture that supports the management of a diversity of decentralized energy units as well as energy management. Throughout the implementation phase of the model project, this multi-layered architecture has been translated into a fully functioning virtual power plant that was successfully operated in the field tests. The project consortium conceptualized an IT platform that proved to be very robust and efficient.
In the model region RegModHarz, the pool coordinator (aggregator) and the energy unit manager together in personal union operate the project’s virtual power plant. The pool coordinator’s responsibilities include the marketing of energy and the market transactions. The project design emphasized the marketing of aggregated DER capacity on wholesale markets. Furthermore, the virtual power plant included a central server application, a database and a graphical operator interface (control room). During the field test, 25 energy units were integrated into the virtual power plant via ICT infrastructure. The bundling processes in RegModHarz strongly focused on decentralised and fluctuating generation.
Virtual Power Plant and core communication partners
The software architecture is based on a classic three-pillar-architecture – Model-View-Controller.
- The presentation layer (View) functions as a control room for the virtual power plant. It enables the interactions with the unit energy management system (e.g. the Siemens Power Bridges deployed at the decentralised generation units) and the pool coordinator and assumes the graphical processing of the data it receives.
- The application layer (Controller) contains the VPP-Backend that is responsible for the business logics and data processing of the Virtual Power Plant.
- The project realised the data storage layer (Model) by implementing a central, relational database.
The control room was designed to be the core piece of the RegModHarz virtual power plant. In the control room, various information, such as price signals from supraregional markets, weather and generation prognoses and information from generation units connected to the VPP system, are being received and processed. The control system enables the control room to decide on the optimised exploitation strategy for the connected flexibilities of generation, storage and consumption units. Bi-directional communication systems enables to administrate the incoming and outgoing data as well as the schedules for the VPP-integrated decentralized units.
The connection and communication with the decentralised units is effected by means of a preexisting ICT infrastructure. In fact, the system can make use of DSL connection, mobile radio technologies or especially dedicated lines or wires and a maximally cost efficient connection to the decentralized units can be achieved for the flexible RegModHarz VPP system.
The so called trade management includes the direct interface to the market place, on which the VPP can offer its product bundles, and the customers. It serves as a tool for contract and billing management as well. It registers and documents, whether a customer concludes a contract with an aggregator and meets the flexibility requirements imposed on him by the contract.
One of the most important functions taken over by the virtual power plant system in RegModHarz was the optimisation strategy as the central activity taken over by the virtual power plant. The control room received prognoses (generation, load, prices) from external service provided during the project’s lifetime in order to optimise its schedules and products. The field trial proved that in case the prognoses were based on a complex fundamental data analysis, their quality improved considerably in comparison to a statistical time series analysis. The individual prognoses were eventually transmitted to the central schedule planning tool to plan the further processes. The starting point for e.g. price prognoses was chosen to be the prices of the day-ahead-spot market at the energy exchange EPEX. The prices for the Intra-Day Spot Market accordingly served as guidance for the intraday price prognoses.
In order to provide for flexibilities and enable the necessary control of decentralised units for the virtual power plant, the participating units were equipped with intelligent energy managers and were connected to the control room via various ICT technologies. These energy managers, in most cases the prototype product Siemens PowerBridge was applied, send the possible schedules regarding their respective unit to the control room, where all sent schedules are being optimized against each other based on the parameters of the prognoses analysis. Schedules in this context refer to data regarding the planned operation of the units and parameters regarding flexibilities that can be leveraged.
All incoming data from prognoses and possible schedules of connected units converge in the centralised data hub of the virtual power plant: they are saved, evaluated and prepared. Based on the achieved trade results, the control room calculates the actual schedule that are being send back to the connected units.
The operation of the virtual power plant concept in the Harz model region has been evaluated within a two cycle field test.
Historical EPEX market data from the year 2008 made up the foundation to calculate revenues within both field tests. The development accompanying field test began with the rollout of a first beta version of the server software and the control room in January 2011 and lasted until July 2012. Step by step, the different types of units were connected (wind, engine-based cogeneration plants, photovoltaics etc.).The integration of DER occurred with 35 units (four of them based on a simulation). The administrated nominal capacity was calculated to be at around 120 MW generation and 40 MW load. The VPP’s energy management system placed 29 GWh of electricity on the EPEX day ahead and intraday market between April 2012 and July 2012, which equals a revenue of 2,7 million Euros.
The second field trial is constituted by a part of the field test described above and was carried out within 14 days (July 21st 2012 – July 5th 2012). Within this timeframe, only real, renewable, fluctuating and controllable energy units were connected to the virtual power plant. As described before, actual schedules were calculated for these units, which then implemented the schedules they received from the control room of the VPP, while the VPP traded the bundled amounts of scheduled load and generation on a simulated EPEX market. During this field test about 80 MW of general nominal capacity were administrated. The field test took its course without technical disturbances and an overall amount of 2,2 GWH electricity could be placed on the market, which equals a revenue of 250.000 Euros.