Cyber-Physical Systems on the Megawatt Scale: The impact of battery control on grid frequency stability

TL;DR

The paper identifies a robust $1 \min$ pattern in grid frequency that appears worldwide and links it to the EMS-driven operation of battery-based storage in UPS devices. It combines cross-grid frequency analyses, wave-form characterization, and inertia-aware power mapping via the aggregated swing equation to show that the pattern scales with decreasing inertia and translates into periodic active-power fluctuations ${P_T(h)}$ when normalized by $E_{rot}$. The authors provide direct evidence from campus UPS installations and a Mallorca case study, demonstrating a DC-side origin that propagates through distribution to transmission levels, and they propose mitigation through EMS control changes such as offset randomization. The work highlights a previously underappreciated cyber-physical coupling risk that could threaten frequency stability in future low-inertia grids and offers practical control-based remedies to safeguard grid operation and security.

Abstract

Electric power systems are undergoing fundamental change. The shift to inverter-based generation challenges frequency stability, while growing digitalisation heightens vulnerability to errors and attacks. Here we identify an emerging risk at the intersection of cyber-physical coupling and control system design. We show that grid frequency time series worldwide exhibit a persistent one-minute oscillatory pattern, whose origin has remained largely unexplained. We trace this pattern back to the energy management systems of battery electric storage systems and demonstrate that the pattern amplitude has increased substantially in the Nordic and British grids. We argue that this effect is a potential burden for stability in future grids with low inertia and an increasing penetration with batteries and smart devices, though it can be mitigated by a revision of battery control algorithms.

Figures (23)

• Figure 1: Fourier Spectra of the grid frequency $f(t)$ for nine grids on different continents that exhibit a $1 \, min$ pattern. Measurements for the CE, GB and Nordic grids are taken on the transmission grid level, for the other grids on the distribution grid level. Details on the data sources and processing are described in the Methods section.
• Figure 2: Characterization of the $1 \, min$ pattern. a Average Waveform over a day in local time, that is Central European Time (CET) and Central European Summer Time (CEST) for the CE grid in 2018. The pattern's amplitude changes during the day, but the overall shape remains consistent: The frequency consistently decreases during the first half of the minute. b Average Waveform over the year for the CE grid. The pattern is highly stable; only a slight seasonal profile in the amplitudes is visible. c Weekly profile of the amplitudes for the CE grid in 2018. The pattern is significantly weaker during the night. On weekends, the morning increase is later than on weekdays. Time is given in local time, and the shaded area corresponds to the standard deviation. d Long-term trends for the weekly averaged amplitudes for the three major European grids. We observe a seasonal pattern: the amplitudes are stronger in summer than in winter, except for the Christmas/New Year period. For the GB and Nordic grids, the strength is also clearly increasing over the years. During the COVID-19 lockdowns in spring 2020, we observed stronger amplitudes for all grids. Again, the shaded area corresponds to the standard deviation.
• Figure 3: Statistical Analysis of the power amplitudes $P_{T}(h)$, thus removing the impact of inertia, for the GB grid. a: The power amplitudes show a profound daily pattern, similar to the amplitudes in the frequency. The hour is given in local time, and the shaded area corresponds to the standard deviation. b: Comparison of the weekly averaged and standardized power and frequency amplitudes time series. We conclude that the seasonal pattern and the long-term trend in the frequency amplitudes can be attributed to changes in the grid inertia.
• Figure 4: Fourier Spectrum of the Active Power Time Series for selected PQ-Meters in a building on the Jülich research campus. a: Electrical Diagram of Building 1 specifying the location of the PQ Meters at the Transformers. The building contains offices and computing infrastructures. b: For Transformer 1 we measure a $1 \, min$ pattern in the active power demand. Attached to this transformer are offices, a heat pump, and a UPS. For Transformer 6, we do not observe any significant patterns in the same measurement period. However, the noise level is also significantly higher, and potential patterns could be hidden.
• Figure 5: Recurrent Current Peaks in a UPS cause periodic $1 \, min$power variations across all voltage levels up to the connection to $110 \, kV$ grid. a Electrical diagram of the UPS at Building 3 and its connection to the $110 \, kV$ distribution grid, indicating the locations of all measurement devices used to trace the pattern. b Recurrent narrow current peaks with $1 \, min$ periodicity are observed at the DC side of the battery. c Although no clear $1 \, min$ pattern is evident in the three-phase AC currents immediately before the grid side inverter, Fourier analysis reveals a $1 \, min$ component. To exclude the influence of downstream loads, the AC output current after the building side inverter is subtracted such that only the internal consumption is considered. d Fourier spectra show that the $1 \, min$ current pattern gives rise to a corresponding pattern in the active power demand, which can be traced across two transformers and a bus bar up to the connection with the $110 \, kV$ distribution grid.