The Economic Provision of Frequency Response in a Low Carbon Future

  • Heather Wyman-Pain

Student thesis: Doctoral ThesisPhD


The introduction of intermittent, renewable generators in recent years has had numerous effects on the operation of the electricity system and the wider environment. One of the greatest operational challenges is maintaining system balance between generation and demand, both in the long term and short term. System frequency is utilised as a measure of the short term mismatch, with frequency response returning the frequency to a nominal value, traditionally by increasing or decreasing the generation output level. This was done by large scale, flexible, thermal generators in the past, typically fossil fuel power stations, which formed the standard frequency response capability requirements. However the cost of providing the same level of frequency response for other generator types is ignored by the technical requirements for participation in the electricity markets. The future of frequency response in a low carbon system must take into account the variety of different generators which are capable of providing system frequency support and the costs associated with their methods to do so.
This thesis addresses this specific issue by considering the costs and capabilities of different technology in operation, the changes of system frequency and the creation of a new frequency response market. An assessment of electricity production cost in relation to generation output levels, using the levelised cost of energy method, is presented with different generator types compared. This demonstrated coal and CCGTs are the most economical choice for operating at a lower output to provide frequency response, only 30% the cost of utilising nuclear to provide the same level of response.
As the penetration of intermittent generators increases the number and nature of frequency events is anticipated to change, making this an essential consideration in the maintenance of system frequency. Trends in real system frequency data indicated an increase in frequency events during weekday evenings and a minor rise in the mornings, coinciding when demand is sharply changing suggesting response to demand changes
is slower than ideal. Additionally high frequency events were over double the length of low frequency events, suggesting high frequency response is slower than desired, but the system frequency remained within the required limits. Generation mixes were found to be less impactful than anticipated, with only minor relationships between rate of change of frequency and synchronous generation levels. These associations inform the
needs of a new frequency response market.
Current frequency response markets are overly complicated and must change to continue to remain economic in a low carbon future. A reduction in the number of ancillary service markets and ensuring all different generators are able to operate in a transparent and fair manner to aid in system frequency maintenance is essential. The technical and payment specifications for different markets were utilised to inform the creation
ii of a new frequency response market which simulated comparable costs in the future system with those of the current system.
Date of Award1 Nov 2021
Original languageEnglish
Awarding Institution
  • University of Bath
SupervisorFurong Li (Supervisor) & Kang Ma (Supervisor)

Cite this