The Operability Strategy Report produced by National Grid (PDF) is an annual document that elaborates on the challenges the Transmission System Operator faces to operate the power system in the UK in the current and future energy landscape. In the 2021 report, the operability challenges are covered in five key areas: frequency, stability, voltage, restoration and thermal. Accounting for these key areas is paramount to maintaining the system operation in a reliable and efficient fashion. In this post, PSC UK’s Chief Engineer in Power Systems, Steve Nutt (SN) and Engineer Carlos Ferrandon (CF) dig into the stability section of the report, focusing on system inertia as a key component of frequency stability.
Steve Nutt (SN): Inertia is a relevant topic that Transmission System Operators (TSOs) analyze in the grid under a high penetration of renewables scenario, namely wind and solar. For example, in summer 2020, we had very low demand days due to the covid-19 pandemic[1]. Now, in the 2021 Operability Strategy Report[2], National Grid mentions scenarios of very low inertia for the future and the possible actions to tackle this challenge. From the operations planning area, and specifically in a day-ahead perspective, what are the challenges of frequency stability that you see?
Carlos Ferrandon (CF): There are indeed some challenges to plan for the day ahead horizon. The summer of 2020 provided a glimpse of a future low-carbon grid, hence the forecast of inertia in future energy scenarios by National Grid. The inclusion of the frequency stability constraints, covering against the largest in-feed in the system, implies adding constraints to the optimization problem that were not included two decades ago for conventional large power systems such as the GB case. Specifically, we are referring to the frequency nadir, which is the constraint that dictates the lowest point the system frequency will reach after a generation or interconnector outage. The so-called frequency nadir constraint is highly non-linear, and this represents challenges to the TSO to add it into a linear framework, such as the traditional UC optimization problem.
The reduced system inertia that we saw on May 23rd of 2020, as mentioned in the 2020 Operability Strategy Report, signified the TSO to dispatch power plants, i.e., plants that would not have been dispatched if the low-demand conditions and high share of wind had not been present in the system. We also saw “abnormal” generation outputs of these units since they were included in the dispatch to add inertia to the grid, hence they were operating in minimum power outputs. This action displaced low-carbon energy sources that could have been included in the normal conditions of the scheduling. All these actions were implemented to safeguard against potential large frequency deviations.
As we can see, there are indirect effects of adding the needed frequency stability constraints into a classical UC study. This event in the summer of 2020 can be extrapolated into the scenarios analyzed in the 2021 Operability Strategy report by National Grid, where they estimate that the inertia has to be kept above 96 GVAs to limit the Rate of Change of Frequency (RoCoF) to less than 0.5 Hz/s in the grid in the zero-carbon operation scenario of 2025.
SN: Regarding the low-inertia scenario that you mentioned, what is your take on the increasingly heterogeneous distribution of inertia in the GB system and how can this be included in the Unit Commitment optimization problem?
CF: Let me describe a little bit of the context here. There are a large concentration of Converter-Interfaced Generation (CIG) technologies in the north of the UK, predominantly made up of wind. In a low-demand and high penetration CIG scenario, the active power flow is usually moving from the north to the south of the UK. Under the assumption that a minimum level of inertia is needed to alleviate the effect of the RoCoF, this constraint can be added to the UC problem, acknowledging that total system inertia tackles the issue.
Going back to the scenario of the example, in the case of a considerable generation outage in the grid, under the scenario described, we would expect different dynamics of the frequency depending on where this outage is presented, coupled with a condition of a heterogeneous distribution of inertia. In other words, we could expect more sensitive areas of the grid for frequency events as the inertia is localized in specific areas of the system. For the UC problem, this would entail that the assumption of total inertia in the system may no longer respect the RoCoF constraint evenly in the grid. It makes sense to explore how a localized RoCoF constraint can handle the problem, with its respective implications.
SN: The stability section of the report mentions the inclusion of GB grid forming capability, which comes from non-synchronous generators, namely wind, solar, etc., and can mimic the properties of a conventional synchronous machine in terms of frequency response capabilities. We have seen a Grid Code Modification (GC0137) with the minimum requirements for grid forming capabilities by CIGs[4]. If this feature is deployed in the current GB power system, where would the TSO benefit the most?
CF: I think that the grid forming capability of the CIGs can benefit the system overall. The next step is to decide what participants, considering their location, can have a higher beneficial impact. It is known that the optimal location of grid forming CIGs can help to alleviate the rapid and different dynamics of the frequency in low-rotational inertia areas of the grid. Participants can also benefit by competing in the Dynamic Containment (DC) scheme that is set out in the Operability Strategy Report as well. It makes sense to think of CIGs as the next providers of this feature, acknowledging some of the conventional generators are being displaced in the current electricity landscape.
[1] National Grid Operability Strategy Report 2020 [Online PDF]
[2] National Grid Operability Strategy Report 2021 [Online PDF]
[3] https://en.wikipedia.org/wiki/Unit_commitment_problem_in_electrical_power_production
[4] GC0137: Minimum Specification Required for Provision of GB Grid Forming (GBGF) Capability (formerly Virtual Synchronous Machine/VSM Capability), National Grid [Online]