Towards an Integrated Grid16 October 2009 Mitch Beedie
Increasing efficiency and ensuring supply from renewable sources are key priorities in promoting their uptake. Mitch Beedie looks at how virtual power stations could be the answer.
A common criticism of renewable power seems about to be answered. Power collected from sunshine and wind is variable and will allegedly make for a chaotic grid that will need fossil or nuclear fuel power stations as backup. However, the solution could come from integrating renewables into the grid rather than just connecting them up haphazardly.
'Virtual power plants' or VPPs aim to tie together hundreds or thousands of micro-generation power stations into a single, secure web-connected system. They have the ambitious aim of making renewable energy resources more reliable and cost effective, without fundamental infrastructure changes and without building a new plant.
VPPs, it is claimed, will also increase efficiency by supplying power from sources located physically close to demand to reduce transmission losses. And, unlike large coal powered stations, they will react quickly to follow loads at short notice. How to create such VPPs and maximise energy storage solutions to fully realise the potential of renewable energy will be a key focus when industry leaders meet at the upcoming Arena International Energy Storage Solutions conference in London on 8-9 December.
On the supply side, VPPs connect small, distributed micro-generators such as solar (including domestic photovoltaic) and wind energy, biomass microCHP and small hydro. When there is no sun, wind power takes the load. When there is no wind, biomass, stored heat and stored energy in pumped water systems fill the gap.
On the demand side, consumers can for example be warned of excessive consumption to switch off unnecessary loads like unused lights. And in the middle, a control station predicts the power available and aggregates supplies to produce a reliable output.
There have already been small scale trials, including Australia's scientific research organisation CSIRO in Newcastle, New South Wales, and the University of Kassel in Germany. Perhaps most important, though, is the EU's Fenix (flexible electricity networks to integrate the expected 'energy evolution') field trials. Among other things, the four-year project aims to standardise the different data and transmission protocols of the generation, transmission and distribution companies.
A cluster of generators
The Fenix consortium has 18 partners. The project is based around transmission and distribution utilities, with support from distributed power providers, equipment and ICT manufacturers, research centres and universities, and regulation and standardisation organisations. A generic architecture is being tested by two complementary demonstrations covering the northern and southern markets.
The northern demonstration uses a cluster of small-scale generators linked to a common low voltage distribution network owned by Woking Borough Council (WBC). WBC uses domestic and community scale CHP, a fuel cell and PV panels to supply its civic centre, conference centre and other municipal facilities at the Pool in the Park. The transmission network comes from the UK's National Grid, which ensures system security and procures ancillary services to balance supply and demand. EDF Energy operates various power sources and minimises imbalance.
The southern demonstration, in Spain, started in September 2009. It covers larger distributed resources such as wind farms and industrial cogeneration for medium-voltage networks. It is led by Iberdrola, with the system again simulating rather than controlling real energy demand. The 100MW being simulated could supply half the peak load of the Álava province's 300,000 population. If successful, it will control real resources and could then be rolled out countrywide.
Fenix is demonstrating that distributed resources can supply the grid. Each connects to a local intelligent 'Fenix box', which handles the transmission and distribution control and information interfaces as well as their associated protocols. It controls the local loads and generators, in addition to showing the available generation and the current demand.
The data is sent to an e-terra distributed management system hosted by Areva T&D that guides how best to dispatch the generation.
The system is updated in real time with all the supervisory control and data acquisition (SCADA) values of the chosen region. It handles layered communication and control for normal and abnormal operation.
The e-terracontrol system is central and ideally suited to Fenix. It is an open platform SCADA system that supports many industry standard protocols. It has distributed architecture support with automatic configuration and restart, remote viewing and operation.
Full redundancy gives no single point of failure and helps keep utility availability above the 99.95% industry standard. Optional Kerberos-based authentication, authorisation and audit features help ensure security. The system can accept 50,000 points, 128 serial communication circuits and 200 network-based TCP/IP connections.
The demonstrations show that VPPs can perform voltage control, participate in the 'day ahead energy market', provide the necessary tertiary reserve, ensure security of supply and handle network contingencies.
Load shifting and valley filling can postpone the need to build costly new plants and also help reduce the need for huge energy storage. Adding a heat buffer like a hot water tank to a microCHP can give some flexibility in electricity production, for example. By decoupling demand from heat production, electricity can be generated during the most beneficial periods.
System security is of course a major issue. Transmission companies are normally responsible but VPPs will need the distribution companies to implement active network management to ensure a secure system. This moves from traditional central control now typically controlling hundreds of generators to distributed control of hundreds of thousands of generators and controllable loads.
From prototype to practicalities
Work is still needed on the physical wide-area network (WAN) communication layer and the communication protocol (such as JMS or SOAP), grammar and semantics. Messages themselves will have to be standardised to communicate performance requirements such as energy quantity and estimated duration, highest possible downtime and estimated costs. This will all be part of the Fenix recommendation on amending international power standards and that brings the possibility of a totally integrated grid.
The price will be critical, however, and particularly the price of the Fenix box. Engineering firm ECRO SRL has developed a Linux-based prototype, which experts say could cost just €100 in large quantities. For most generators, the benefits will greatly outweigh costs.
Whether adding €100 to millions of new PV installations will increase repayment times unacceptably, although, is yet to be seen.
To reduce overall system costs, the output from domestic PV systems might better be modelled by the VPPs rather than actually controlled. There are still many 'low-cost' energy improvements that could be made in houses and businesses that would bring immediate energy savings. As with all energy improvements, the decision should come down to repayment times.