A Novel Approach To The Optimal Planning of Active Distribution Networks

Many factors, such as environmental, technological, economic and regulatory changes are resulting in the evolution of the size and shape of electricity distribution networks in the UK. The growth of distributed generation (often from intermittent sources), may result in the need for future networks to operate with higher degrees of flexibility than conventional passive distribution networks. The Supergen FlexNet research consortium is committed to investigating how to achieve a vision of flexible electricity networks. The work this paper summarises forms part of the Size and Shape of Future Networks work stream, which investigates the factors which will determine the characteristics of future networks such as generation mix, distributed generation penetration levels, demand, controllability, and therefore determine the degree of flexibility required to meet the energy challenges of the future. Increased levels of flexibility will allow higher penetrations of generation to connect to the distribution network without affecting network security; and can potentially increase the utilisation of network assets, deferring the need for reinforcement of network infrastructure. The deployment of such flexible control systems will mark a major change in the way distribution networks operate and, in turn, how they are designed. The paper examines the shortcomings of traditional approaches for distribution network planning and design when implemented for future flexible networks. Currently in their infancy, novel flexible active control schemes are deployed to facilitate generation connections on a case-by-case, bespoke basis, and do not feature in the long or medium term design task for distribution networks. As these technologies become more mature, established and widespread on distribution networks, traditional distribution network design techniques based upon deterministic maximum/minimum network conditions will result in undesirable conservatively designed networks. There is also a need to address the result of uncertainties in the future network design process, which will be accentuated by the advent of flexibly controlled networks. This paper will introduce key technologies which will contribute to the flexibility of future networks and their characteristics; illustrate the traditional techniques used for the planning and design of distribution networks; the shortcomings in these methods when applied to fully flexible networks; and propose new approaches to the network design task which takes uncertainties into account. The use of Stochastic Programming (SP) as a supporting analytical tool for distribution network planning and design under uncertainty is introduced and the formulation of the SP distribution network planning problem is addressed. Stochastic programming techniques facilitate the solving of optimization problems in the presence of uncertain data. The paper will discuss the accuracy of probabilistic data of stochastic parameters and its potential effect on the SP solution, and the role of risk in the decision making process.