Network Nodes for Smart Thermal Grids
Abstract
Analogous to the smart grid in electrical context, we have developed a smart thermal grid providing features to supply grid network partners either with heat or to consume superfluous thermal energy. A local communication and supply network connects several customers with each other. Depending on offer or demand, the transport medium can be transmitted bidirectional between interconnected network node stations. The network for information exchange and trading operates parallel to the thermal supply network.
Our developed network nodes consist of a hydraulic module for transportation, measurement and control technology as well as a microcomputer for advanced tasks and communication. Furthermore, the microcomputer provides a swarm intelligence controller. Our controller is responsible for data communication, processing network events and realizing a strategy for optimized network operation. Network nodes are able to optimize the network in a decentralized manner. Network nodes are devices that manage all transport-, measuring and control tasks. This paper gives a brief insight into our development efforts to the hardware and software platform of our microcomputer as well as the routing strategy based on ant colony optimization.
Keywords
Download Options
Introduction
In general, heat energy is one of the most important energy types especially because of a high cost impact. Thermal energy is required e.g. as process heat as well as in heating processes. Hence, technological processes are causing huge quantities of thermal energy consumption and there is a lot of wasted and not reused energy. Against this, we have countless consumers that could reuse excess energy. Therefore, we searched for innovative approaches providing consumers with unused and potentially wasted energy from producers. The resulting smart thermal grid provides a solution to serve either cooling or heating demands by exchanging wasted energy amounts.
Our basic intention is to reduce the amount of required primary energy and the related carbon dioxide emission by reducing wasted energy. Therefore, we need an efficient heat transmission and consumption using different temperature levels within a common grid and smart control. A sensor network gathers required grid state information e.g. temperature and pressure values, actor parameters, flow rates as well as consumer demands. For our smart thermal grid, we have developed a special distribution node. Those nodes manage energy- and data-flows within our grid and control the hydraulic and thermal as well as the communication level.
The approach of our smart thermal grid is a multidirectional energy transmission between consumers and node stations in form of bivalent and multivalent nodes using high and low temperature levels. Therefore, we have developed hydraulic devices for transportation of a thermal carrier medium in several directions with different temperatures. Distribution nodes include hardware for measuring and a microcontroller, as illustrated in Fig. 1. In cases of communication failure, our microcontrollers automatically switch to internal failsafe mode and keep the network running.
Our network controller (NC) is part of each distribution node and we have a connection between them to explore the entire network structure automatically. Each NC provides their network and grid state information as well as consumer demands and offers. Additionally they have an autonomous search algorithm to implement a decentralized control approach.
Conclusion
This paper provided a brief insight into our development efforts for creating smart thermal grids with focus on our developed network controller. Therefore, we discussed the technical architecture of our designed hardware platform, its software stack and some important implementation details. Furthermore, we presented our search algorithm based on ant colony optimization. We have implemented a simulation environment to combine existing grid nodes in a real test station with a virtual smart thermal grid for further researches and successfully tested our search strategy. Currently, we get optimized results at any time and we could solve the important aspect to deal with highly dynamic networks. Further work could deal with tests and efficiency evaluation of alternative search algorithms in different network structures.