Adaptive Management of Energy Packets

The Electric Power Grid is an excellent example of a Cyber-Physical Systems (CPS) which includes sources of fossile and renewable energy, together with electro-mechanical conversion systems, electric energy transport systems, physical systems which consume energy, and distributed computer systems and networks that serve to monitor and control the system as a whole. With the increasing introduction of ICT into the Power Grid, we are seeing that all the classical issues of distributed systems and networks are coming into the heart and core of the issues related to energy. When renewable energy is used energy storage also becomes very useful. Storage units can include electric car batteries and uninterruptible power supplies, dams, and compressed air units. The presence of intermittent sources of energy, and storage units lead to an analogy between computer networks which store and forward and download data, and electrical networks that furnish energy both from direct production and storage. Thus this paper discusses Energy Packet Networks (EPNs), a concept for systems which store and forward quantised energy units from generators and renewable sources, to storage and consumers. An EPN can also be a framework for energy harvesting networks at the micro-scale, and bio-nano scale energy networks. We present a mathematical model for EPNs and analyse an example the where stored energy and a fast ramp-up source are used to meet random energy surges assuming a storage unit and a random flow of renewable energy. We compute the probability that all requests are satisfied, or that some cannot be satisfied, and the spill over rate of renewable energy, and then evaluate the effect of an imperfect communication and computer-control system where message losses and delays can degrade the energy system performance.