What our payment and incentive scheme does is to propose an alternative which encourages people to be more efficient and form these virtual power plants. It also encourages them to provide as accurate estimates as they can about their future production capabilities.
Dr Valentin Robu
Although distributed energy resources (DERs) such as wind and solar farms have become more common in the electricity supply network over the last ten years or so, the role that these resources can play in the national energy supply has been limited. Despite the beneficial effects of becoming less reliant on conventional fossil fuel sources, the size and relative unreliability of renewable DERs has prevented their wholesale integration into the grid.
Virtual power plants (VPPs), which are formed by the aggregation of many DERs, are one potential solution that has emerged to this problem as they provide a means by which DERs can become sizeable and reliable enough to compete with traditional providers.
Researchers from the University of Southampton have designed a payment mechanism which could help so-called ‘cooperative’ VPPs (CVPPs) through the use of intelligent, multi-agent software. The grid requires estimates of energy production in which the supplier has a high degree of confidence in order to plan a supply schedule which ensures energy demands are always met.
I asked Dr Valentin Robu from the Agents, Interaction and Complexity Research Group at Southampton to tell me more about the problems with the current system and how VPPs can encourage renewable energy production in the United Kingdom…
What are the main obstacles to integrating DERs into the energy infrastructure of the UK?
The main issues – the ones we tackled in our paper, at least – are that renewable generators such as wind turbines and solar panels are often too small and unreliable to compete efficiently in the market with conventional generators. Not only are they too small, but more importantly, their supply is unreliable in the sense that it depends on wind speed, sun irradiance, cloud cover, and so on, at any given time. It is therefore crucial for the grid to get reliable estimates for better production planning.
What exactly is a virtual power plant, and how does it make connecting DERs and the grid easier?
A VPP is essentially an entity that sits between a number of small generators and the grid, and manages their production, estimation and interaction with the grid. It doesn't have to have a physical existence, other than the generators, wind turbines, etc, that are already in place. It does mean that, for example, the grid and distribution companies can make contracts with and directly query the VPP, which will deliver the power it commits to deliver from multiple members. This makes the process much more efficient, firstly because the grid doesn't have to worry about interacting with hundreds (or even thousands) of small producers, and secondly because grouping multiple generators together improves reliability. When some of them are unable to produce, others may make up the shortfall.
What are the failings of the current feed-in tariff system?
First, it is rather expensive for society. The feed-in tariff levels tend to be several times more than the cost of electricity on the open market. But, more importantly, it only rewards generators to produce; it does not reward producing at a useful time for the grid or encourage these generators to be reliable in their production.
How does the improved CVPP payment mechanism function and why is it better than other mechanisms?
Well, many of these small generators are owned by many different stakeholders, which have their own interests in mind. This means that the formation of virtual power plants in practice is not so easy to achieve.
What our payment and incentive scheme does is to propose an alternative which encourages people to be more efficient and form these virtual power plants. It also encourages them to provide as accurate estimates as they can about their future production capabilities. This information is very useful for the grid to manage demand and generation from conventional power plants.
Are software improvements enough to encourage the spread and integration of DERs, or would the whole system benefit from an overhaul?
No, absolutely not; we are not saying that. In the long term, changing to a low-carbon economy (a commitment made by all of the UK governments) and assuring energy security will require additional measures, like building additional generation and transmission infrastructure to cope with this. However, such infrastructure is very expensive to build.
What we are saying – and our method is a step in this direction – is that we can manage the resources that are already available in a much more efficient way, using new methods and advanced software techniques, or techniques inspired by fields such as Artificial Intelligence. This is essentially the vision of the Smart Grid; to achieve better control and matching of demand and supply in decentralised electricity grids.
As we are not nocturnal animals by nature, then we must benefit from sunlight, in several ways. It then follows that we can deal with sunlight - maybe we need to be more in touch with how to deal with exposure. This research shows that we do not know all the ways that we use sunlight naturally, which is probably true of other environmental factors. I will be very interested in outcomes of further research. Is there any research about sunlight and cognition?
