The application and assessment of active network management techniques for distribution network power flows

The UK Government regards renewable energy technology deployment to be crucial in successfully meeting reduced greenhouse gas emission targets. As such, incentives promoting the connection of clean energy sources to the UK's electricity infrastructure are influencing a significant increase of distributed generation (DG) planning applications. With an abundance of the UK's indigenous energy resources being either rural or coastal large volumes of DG are seeking connection at the lower voltage distribution networks. Connecting large quantities of intermittent generation, to distribution networks, presents significant problems to the planning and operation of these traditionally passive networks due to bi-directional power flows creating voltage fluctuations and uncertainty in power flow magnitudes. In addition, conventional planning methods results in financial barriers that are preventing DG connections due to the high cost of reinforcing the existing infrastructure. One method of avoiding, or at least deferring, this high capital costs is to adopt an Active Network Management (ANM) approach. This thesis presents and evaluates two novel ANM approaches that manage the real power output of multiple DG units, in real-time, such that distribution networks operate within thermal limits. Studies are conducted in a closed-loop simulation environment, with actual hardware, on two topologically different networks to demonstrate the flexibility of the novel application of the Optimal Power Flow (OPF) and the Constraint Satisfaction Problem (CSP) when applied to the Power Flow Management (PFM) problem. The performance of these model-based algorithms is assessed against their ability to detect thermal excursions, their solution computation time, their resilience to measurement error, their real power curtailment, their ability to conform to current DG commercial connection agreements and their ability to adapt to changes in network topology. Results reported in this thesis demonstrate the feasibility of these novel ANM approaches for PFM, and their applicability in terms of incorporating intelligence into the UK's future smart grids.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral-pg

Grantor dc:publisher.institution

University of Strathclyde

Year dc:date.issued

2012

Author dc:creator

• Dolan, Michael James

DOI

10.48730/xn7w-db81

Identifier

T13120

OAI identifier oai:identifier

oai:strathclyde:mk61rg97b