1. Reliability Analysis for Power Networks
A power transmission or distribution network has a sophisticated
protection system which detects short-circuit faults and automatically
removes the faulted equipment form service. In many cases alternative
supply paths are available and so customers do not experience any interruption
of supply. However, for some faults, especially in lower voltage
distribution networks, there will be a period (ranging from seconds to
days) during which the power is cut off.
This project will design and develop a software program
to analyze the available protection system and the network configuration
in conjunction with statistical data on the likelihood of fault occurrences,
in order to estimate the overall reliability of supply to any customer.
This will give an expected number of hours lost for each consumer per annum,
and would be useful to Engineers designing new networks or improving existing
networks.
This project would appeal to a student interested in
reliability analysis, and familiar with computer programming.
2. Predictive Modeling of Power Transformers
This project would design dynamic models of the behavior
of large transformers in power transmission and distribution networks.
These models would then be implemented in computer software and would eventually
enable on-line condition monitoring of transformers (to predict in-service
failures). They would also provide approximate thermal modeling,
which could be introduced into power system analysis software. The
safe working limit of a transformer is mainly dependent on its operating
temperature, and by allowing for thermal inertia it would be possible to
operate the transformer closer to its working limit.
This project would suit a student interested in power
and energy systems, and in computer programming.
3. Local Search Algorithms for Minimising the Cost of Energy Losses in Distribution Networks
Distribution networks are usually designed as ‘meshed’
systems (containing loops), but are operated with a number of switches
open to create a radial system (with no closed loops). The BIPS group has
existing algorithms for optimising the selection of open points in these
networks. At present the algorithms are reliable and appear to give good
solutions. However, a considerable amount of computer time is required
for the search processes. It is also not know whether these algorithms
produce an optimal solution.
Some new local search algorithms have recently been introduced.
This project will investigate the application of these algorithms to the
above problem, and will investigate the computational performance obtained.
The project would appeal to someone interested in computer algorithms and
programming. Initial knowledge of power system modeling would not
be needed.
4. Optimal Energy Contract Portfolio Analysis
Power utility companies are now de-regulated and must
compete in an open market. Since large companies buy and sell energy
in large blocks under a variety of different contract prices and conditions,
there is a need for sophisticated portfolio analysis to help companies
to maximise profit while minimising risk. This is an entirely new
project area for power systems, although there is considerable literature
available pertaining to financial markets. The project would be suitable
for a student interested in opening up a new field of investigation and
who is interested in mathematical modeling and in computer algorithms.