Combining the Internet of Things, the increase of available data, and machine learning with the trends of renewable and micro-generation promises to significantly change the energy market of the future. It will be greener, efficient, and most of all automated and semi-distributed.
From Demand to Supply-driven Energy Market
Most of us grew up in a world where electricity was generated on demand. Large power plants on one side, a transmission network in the middle, and the consumption – households and industry – on the other side. Demand drove how many power plants are built and how much they generate at any given time. This is about to change for two reasons. Firstly, renewable energy and secondly the Internet of Things (IoT).
Renewables bring challenges with new generation locations, micro-generation, and unpredictability. Traditionally, power generators had plants close to consumption to reduce loss and transmission cost. Germany is a good example how this is being challenged. Most of Germany’s energy-intensive industry is in the west and south, but wind generation is mostly in the north at the coast. That electricity has to be moved south which requires new transmission capacities.
From Centralised to Decentralised Energy Market
Micro-generation like households with solar panels or farmers with small wind generation add strains on the energy market, e.g. on pricing, and muddies the waters further on forecasting capacities and profitability of traditional generation assets. This ties into the unpredictability of renewable assets. Germany demonstrated that on peak days when the sun shines and the wind blows energy can become so readily available that it drives prices into the negative. That is bad news for anyone trading and worse for anyone owning and running generation assets. The lack of large-scale storage means that forecasting pricing and investment strategies become significantly more important. The days of somewhat predictable price ranges are gone. And so the mix of Excel-sheets with simple price-curves, experience and gut-feeling do not serve well anymore.
From Heuristic to Self-Organising Energy Market
Collecting and analysing more data better is the first step. But the question is not only how to manage this intermediate state of the energy market but where it will find a new stable state. Besides renewables, IoT will play a major role. Consider a near future with a significant number of electric vehicles, ubiquitous micro-generation and major renewable generation sites, as well as highly connected devices from fridges, boilers, down to lettuce – yes lettuce. Ignoring the lettuce for a moment, electric vehicles provide an opportunity to capture and buffer some of the spikes and troughs of renewables. Other devices like boilers and fridges can similarly contribute. They have combined significant consumption and can defer or advance some of the consumption flattening out daily spikes.
Naturally, for this to work data-driven intelligent agents and forecasting services are essential. The markets have to become more decentralised and automated. Manual trading in a few exchanges and B2B is far too inefficient and intransparent. Households should participate equally as should large consumers and producers. No one wants to do this manually, but consumers might be open to merely set some parameters, e.g. on a mobile app if they prefer to prioritise renewables or are more price sensitive. Many consumers can choose already via intermediary contracts in a more rigid way. But technically nothing should stop them in the future to have more dynamic control and cut down on intermediaries. For example, why can’t an electric vehicle be used as buffer storage when the owner is on vacation? Or why shouldn’t a consumer be able to switch from green to ‘cheap’ energy whenever she feels like it?
The evolution of such agents could lead to highly integrated intelligent energy management. Simple RFID chips and sensors become continuously cheaper and maybe soon printable for cents. This may lead to a revolution in the supply chain that may affect energy trading too. Imagine lettuces being harvested and immediately tagged with such an inexpensive sensor chip on the field. Supermarkets would love it. It provides an audit trail on where a product has come from and how it was treated, e.g. how old it is and if the correct cooling chain was observed. Now imagine you purchase your groceries and put them in your fully connected car, which in turn can read what has been purchased from the chips and send the information to your intelligent agent. Your agent can then decide if you need more cooling or heating at home and if it should plan to sell your solar panel energy or store it in your electric vehicle. Machine learning can even work out special events – say your annual spring barbeque – and react accordingly to your needs optimising energy use and cost without a single interaction.
None of this scenario is science fiction. Most technologies are available, and it is conceivable that in the near future we have an intelligent IoT integrated energy market emerging.