OPTIMISATION OF NETWORK OUTAGE SCHEDULES

Project: Research

Description

Electric power systems should be operated in order to provide adequate reliability of supply of power and facilitate trading its trading. Reliability of supply is affected by random disturbances including failure of generating plant and fault outages on the network. Weather related outages cannot be prevented; in the medium to long term, however, reliability will be affected by the performance of individual power system assets which must be maintained at suitable intervals. This and construction of new assets require the scheduling of outages to permit work to be done safely which should nevertheless not significantly affect reliability of supply in the short term.

In planning outages, the transmission system operator (TSO) must ensure compliance with ‘security standards’ that articulate levels of reliability of supply for consumers and access to the electricity market for producers. If the standards cannot be met, either a planned outage must be postponed (with the risk that necessary works are not done in time) or the operation of generating plant must be constrained. The latter has the consequence of an increase in the cost of electricity and carbon emissions.

The main objective has been to facilitate the scheduling of outages such that required maintenance or construction can be done within given time windows and the operation of generation is not excessively constrained. The two most important influences on the achievement of an adequate outage plan are:

1. whether there is sufficient network capacity to allow secure operation while taking planned outages;
2. the behaviour of generation.

The former is governed by the planning criteria of the GB security standard while the latter is determined by how generating companies, which are independent of the system operator, choose to use their plant.

A particular issue in the former is the accommodation of renewable electricity generation, mainly wind power. The project team has worked with the three GB transmission owners to explore how present rules determining levels of transmission capacity might be adapted when significant amounts of wind generation are present. Wind is a highly variable resource; one impact is to increase the uncertainty in the levels of power flowing across the network. A simulation methodology has been developed to allow these variations to be explored and statistical approaches have been used to identify patterns among the various scenarios and provide rules and a visualisation that might be used by power system planners.

In respect of the other main influence, it may be noted that generation companies make decisions on use of their own plant based not only on contracts for production of energy but also on management of their own physical assets. As with transmission assets, this means the scheduling of maintenance outages at appropriate intervals.

New assets must be ordered years in advance and a limited workforce must have its work planned quite precisely. Because of this the network’s inability to accommodate an unlimited number of simultaneous outages, transmission outages must be planned over a period of years. However, from the perspective of the TSO, generators’ plans are highly uncertain and represent a risk to reliability of supply of electricity in import-limited areas but also an opportunity for network outages in export limited areas.

As a consequence of the above, the second strand of work has sought to increase a TSO’s ability to forecast generator’s outage planning. This has been done by developing a new tool intended to emulate a generation company’s own outage planning. This is based on an optimisation that can take account of future contracted levels of energy production, electricity spot market conditions, the minimum frequencies and maximum durations of generation maintenance outages and the costs of maintenance activities. By being software based, a number of scenarios can be explored and key sensitivities readily identified.
StatusFinished
Effective start/end date11/09/0631/03/10

Funding

  • EPSRC (Engineering and Physical Sciences Research Council): £83,925.00

Fingerprint

Outages
Electricity
Planning
Scheduling
Industry
Electric power systems
Wind power
Costs
Visualization