This week’s blog comes from Yifu Ding, a student on our MSc in Sustainable Energy Futures course. She has written us a piece on her work looking into frequency response, a major problem as more and more renewables become part of our energy system.
In light of the decarbonisation trend in the UK grid, the portfolio of power generation has shifted to a mixture with less fossil fuel plants and more renewable energy generators. However this is a problem when it comes to a quite esoteric part of our energy system, frequency response, which is the focus of my research project.
For a power system to maintain the system frequency around the nominal value (50 Hz in the UK) generation must closely match demand 24 hours a day, 365 days a year. In a conventional power plants like coal, gas and even nuclear, electricity is generated by a turbine, basically a large spinning chunk of metal. Our electricity system cleverly uses the inertia of these turbines to help balance the power being supplied at high or low electricity demands by speeding them up or slowing them down
In an event of the generation outage or surge in demand, inertial energy is released (or captured) to prevent the frequency from increasing or decreasing. Currently this is the first and quickest method of frequency response (FR) and maintains the electricity grid’s stability. It is only the first step and longer term stabilisation of the system is done by FR services procured by the system operator.
My work is trying to predict our future FR requirements and investigate possible solutions to minimise the balancing cost in a low carbon future.
To do this research I am working with Dr Roberto Moreira from the Department of Electrical and Electronic Engineering, Dr Salvador Acha from the Department of Chemical Engineering and Dagoberto Cedillos an alumnus of the MSc in Sustainable Energy Futures, now at Open Energi.
Increasing frequency response requirements
In GB power system, approximately 70% of the system inertia is provided by large conventional power stations. Unfortunately, non-synchronous renewables generation units such as solar PV and wind turbines do not contribute. If a large number of thermal plants are going to be closed to meet carbon reduction targets, the system inertia will dramatically reduce, and the requirement for alternative FR will increase accordingly.
In my thesis, I have built a simplified power system model for Great Britain, which is able to estimate the FR requirement according to the generation mix and system inertia on hourly basis. In this case, the FR requirements can be estimated in different scenarios.
Fast frequency response
I reviewed papers and reports to find possible solutions to increasing FR requirements and balancing costs. I plan to use my model to test their effectiveness in terms of reducing costs. One of the solutions is delivering faster-acting response, which can reduce the overall volume of responses needed. We can think about a power system with the stable frequency as a large tank with the stable level of water. The inlet and outlet represent the generation and demand. Then there is a sudden imbalance occurring between inlet and outlet. If we wait for a long time to respond, the water level in the tank will become quite low, and we need a large additional inlet current to catch up with.
In July 2016, National Grids launched and tendered a sub-second FR service called enhanced frequency response (EFR), compared with the primary frequency response (PFR) which is delivered in 10 seconds. It is a remarkable improvement and I will use my model to estimate how much PFR that the new service can offset.
Balancing market designs
Another area I investigated was how to address the FR problem through the market design. Apart from introducing the sub-second FR, National Grids also plan to change the balancing market to attract more FR providers, therefore the volume of FR in the future can be guaranteed.
In the System Needs and Product Strategy report [PDF] that National Grid has published recently, it has been proposed that current balancing services should be simplified considering multiple markets and unclear services criterion at present. Obsolete products will be removed and then requirements will be standardised along a single timescale to reduce the complexity and overlapping. In the future, National Grid also envisages grouping similar products into one single product. In this case, there will be one kind of product with different parameters such as delivery speed, duration and ramp rate instead of different products with one parameter.
All of these strategies aim to provide a level playing field for all technologies. That means they all treat equally based on their real performances, and the market should figure out the optimum solution on its own. I believe my research will provide valuable insight into the problem, outlining the costs and effectiveness of various solutions of the frequency response problem.
I am originally from Beijing, China. Having completed degrees in Electrical Engineering and its Automation from the North China Electrical Power University and Electronics and Electrical Engineering from the University of Edinburgh, I chose to continue my studies with the MSc in Sustainable Energy Futures at Imperial College London.
My keen interest is firstly the power system and electronics. In 2015, I participated a competition held by ABB and was invited to a tour in the main laboratory. There I was presented with their products and solutions for a low-carbon grid in China. China has diverse and rich renewable resources, but most of the electricity still comes from coal power plants, which contributed to 57,2 % (943 GW) in 2016. I decided to explore the energy topic especially the large-scale system. I believe my country has the great potential and will improve a lot in this aspect in next decade.