Integrating Renewable Energy Sources with HiPerDNO
Project Description
Summary
Future electricity distribution networks with mass deployment of network equipment sensors and instrumentation, millions of smart meters, small-scale embedded generation, and responsive load will generate vast amounts of data requiring near to real-time analysis. So-called cloud and grid computing will enable scalable data mining, feature extraction, and near to real-time state estimation. These and other HPC tools and techniques have been recently developed to cost-effectively solve large scale computational challenges in areas such as genomics, biomedicine, particle physics and other major scientific and engineering fields that require similarly scalable communications, computation and data analysis. Based on such recent success it is the aim of this research project is to develop a new generation of distribution network management systems that exploit novel near to real-time HPC solutions with inherent security and intelligent communications for smart distribution network operation and management. Cost effective scalable HPC solutions will be developed and initially demonstrated for realistic distribution network data traffic and management scenarios via off-line field trials involving several distribution network owners and operators.
Duration
1stFebruary 2010 - 31stJanuary 2013
Consortium
This 6.5M (4.4M) Euro project involves 11 partners from 6 European countries. Each partner has been selected as a consequence of their ability to contribute to the overall project. Brunel University is co-ordinating this project.
Co-ordinator
Dr G.A.Taylor, Brunel Institute of Power Systems, Brunel University
Joint Call ICT-Energy
Novel ICT solutions for Smart Electricity Distribution Networks
Project Overview
Distribution Network Operation and Management Systems
Existing electricity distribution management systems (DMS) have been designed using operational and algorithmic procedures that are highly centralised. As more of the distribution network becomes active, moving rapidly from the current minority to the future majority [1], accurately estimating the state of the system becomes essential and therefore algorithms and procedures must be highly scalable to achieve the required near to real-time state estimation. Presently there are severely limited data processing, storage and communications resources within existing DMS. Data communications bandwidth is one example of potentially severe bottlenecks in state-of-the-art DMS. Furthermore, existing electricity distribution networks lack significant monitoring or metering points and hence do not support the massively distributed connection of active network assets (sensors and instrumentation, responsive load) or active electricity customers (smart meters, small-scale embedded generation). Many European countries are planning or beginning major asset replacement and renewal programmes with regard to electricity distribution networks. UK Power Networks in the UK is committed to major asset replacement and renewal over the next five years [1]. Such a major asset replacement initiative represents an ideal opportunity for the massively distributed deployment of state of the art network metering, monitoring and instrumentation. However, major research and development of high performance ICT infrastructure will also be required to fully exploit the investment return on such new assets and to defer further asset deployment. Future potential financial benefits and cost savings to large-scale distribution network owners and operators can be measured in hundreds of millions of Euros [1].
ICT and High Performance Computing (HPC) Technology
Existing cost-effective HPC technology has typically been applied to off-line engineering and scientific applications [2]. However, emerging near to real-time HPC technologies offer great potential for application in the operation of critical infrastructures such as electricity distribution networks [2]. Operational and algorithmic procedures hosted on such HPC platforms will be highly scalable, robust, and secure. In addition, they will offer near to real-time scalable performance for applications such as data mining and feature extraction that can support and enhance state estimation. Standard architectures for applying near to real-time HPC to a new generation of DMS will be proposed and off-line trial-tested in collaboration with distribution network operators (DNOs) and owners. This new generation of DMS will represent a significant step change for electricity distribution networks, and is potentially disruptive to traditional operational practice. Therefore, significant knowledge exchange with Distribution Network Operators (DNOs) will be achieved through seminars and workshops. In addition a rigorous cost-benefit analysis of this new generation of DMS will be conducted to support and enable industry adoption.
Key References
- Ofgem UK, 'Electricity Distribution Price Control Review 2010-2015', www.ofgem.gov.uk
- M.R. Irving, G.A. Taylor and P. Hobson, 'Plug in to Grid-Computing', IEEE Power and Energy Magazine. Vol. 2 No. 2, Pgs. 40-44, 2004.