In theory, we have had a smart grid in Germany for several years now, in which generation, storage and consumption are coordinated with each other in real time through central monitoring and control. In practice, it doesn't work quite as well. But the term 'smart' is a bit of a mixed bag. What used to be 'smart' is often no longer so today - take the smart meter, for example, which from today's perspective is just a simple (calibratable) digital meter and, compared to the intelligent features and communication options of a smart phone, is now completely outdated and has no quantifiable user benefits.

Against this backdrop, VHPready e. V. began a discussion a few months ago on how the open industry standard of the same name for controlling and connecting decentralized energy systems can be adapted to future requirements of innovation-driven digitalization. The current version 4.0 of the VHPready standard was developed in recent years for the marketing of balancing energy and is currently being used in various virtual power plants for precisely this one area of application. VHPready 4.0 is a classic telecontrol specification with some special application-related features. The standard enables decentralized energy systems of different types to communicate with the central control system of a control power provider within the process level of our nationwide electricity grid. A VHPready 4.0 telecontrol connection between the control center and the energy system consists of a monitoring and control direction: Status messages and event messages are transmitted in one direction, while control commands and status queries are transmitted in the opposite direction. In practice, a TCP/IP-based client-server connection via IEC 60870-5-104 with the typical time specifications for control energy is used for this.

Switching decisions are made in the control center and transmitted to the energy system as a single control command or control command sequence (schedule). Execution is monitored by status queries from the control system.

Regulated markets

On the high-voltage side, the German electricity grid is divided into four control zones, for each of which a transmission system operator (TSO) has been assigned sole responsibility by the state. In order to fulfill this task, the four TSOs 50Hertz, Amprion, TenneT and TransnetBW have drawn up binding control energy specifications that have been incorporated into VHPready 4.0. In addition to a system-related pre-qualification check, these include various IT requirements such as the closed user group, the VPN and the communication media break as well as the requirement that no other tasks, such as condition monitoring, predictive maintenance and remote maintenance, may be carried out via the control energy communication connection. If a combined heat and power plant or heat pump already has an internet connection via which the respective manufacturer remotely monitors the system for any faults, the TSO regulations stipulate that this communication connection may not be used for control energy operation under any circumstances. This additional function is then only possible via a completely separate communication technology. The associated acceptance problems and additional costs plus the pre-qualification procedure of the TSOs are probably the main reasons why many suitable decentralized systems are not integrated into a smart grid at all and why VHPready 4.0 has only been able to spread on the market to a manageable extent so far.

The VHPready 5.0 standard provides for the on-site use of AI algorithms for the efficient operation of integrated systems. For example, a small combined heat and power plant (CHP) can then intelligently control local storage systems as an overall system.

© SSV Software Systems

As innovations in microelectronics and IT, which are coming onto the market in ever shorter cycles, are only spreading at a very slow pace in the regulated energy markets, VHPready e. V. sees a permanent need for action to expand the application possibilities of networked energy systems. The VHPready association board now wants to meet this challenge with an innovation offensive. A task force will be formed at an innovation workshop at the end of October and version 5.0 of the communication standard for networking decentralized energy systems will then be developed. Two important technological building blocks for further developments are artificial intelligence (AI) in the form of machine learning and protocol-independent information models, so that energy systems can also be networked with each other via VHPready 5.0 in future and can communicate with each other without the help of a higher-level control room.

The future is decentralized and open

From a technical perspective, it can be assumed that centralized remote control of decentralized energy systems using outdated telecontrol protocols will become less important in the future. In addition to the spread of the IoT, the need for open, intelligent and autonomous solutions in particular will contribute to this. Ultimately, the cyber-physical systems of the Industry 4.0 world will no longer be controlled remotely from a central control room instance. They also require local artificial intelligence that enables them to make autonomous decisions. In this respect, it will be relatively easy to transfer such concepts to networked decentralized energy systems in order to enable new applications and business models. Pilot applications can already be found in the field of energy management and energy efficiency applications, although these are predominantly based on rule-based systems and closed architectures.

In VHPready 4.0, the data points of the individual energy systems are currently embedded directly in the protocol. However, open device interfaces require data-agnostic protocols. In addition, XML-based information models with semantic descriptions, for example, are required in order to easily interconnect network systems without the need for individual device engineering. By using modern algorithms from the AI environment, such as supervised machine learning and neural networks, it is possible to automatically create a schedule for a combined heat and power plant and battery storage system in order to enable the best possible overall operation with local decision-making. The schedule used for this is a mathematical model in which local and external influencing variables (e.g. individual system status, local energy demand based on historical data, weather reports and weather forecasts or operating parameters of external energy grids) and user behavior patterns are taken into account.

Author:
Klaus-Dieter Walter is a member of the management board at SSV Software Systems.