Power grids are critical to the world economy as most other critical infrastructure including communications depend on them. Electric utilities are transitioning towards smart power grids to enhance efficiency and reliability, with real-time measurements and control between smart devices in the field with centralized data facilities running intelligent analytics. The timing solutions for Smart Grid must address a significant list of concerns. Time synchronization is in general considered a key security vulnerability of Smart Grids, and an area of direct relevance to timing resiliency systems and the 5G System is ensuring timing integrity, e.g., providing solutions to the increasing prevalence of satellite signal loss, jamming, and the potential for spoofing [2]. Power utility companies do not want to be dependent only on satellite-based time synchronization due to several factors:

• Interference: Either caused intentionally or by environment (e.g. solar flares).
• Governmental decision: Public time synchronization service can be disabled or weakened remotely; GPS selective availability as an example.
• Maintenance: Changing the faulty GPS antenna may take some time.
• Other: Long downtime time when updating or fixing satellite systems; Galileo downtime of 100h as an example. As a comparison, 24h holdover time has been specified for power systems.

Although there are a large range of applications possible in Smart Grids in need of timing resiliency, the focus here is mainly on power sub-stations where robust time synchronization is needed for e.g., synchro phasor applications. Those are relevant for 5G System-based timing resiliency solutions and simultaneously represent the most demanding application in terms of required time synchronization performance. However, other applications are also mentioned and captured.

Precise timing is key to operating the smart power grid. One area relates to controlling the frequency (50Hz/60Hz), while another area relates to providing events with timestamped measurement values, with required accuracies down to a few hundred nanoseconds [2]. Correct timing is a key enabler for communication and orchestration of technologies for accurate and optimal wide area monitoring, protection and control in the power industry [2]. In a typical smart grid architecture, centralized monitoring systems have visibility to events occurring in a distributed hierarchy of substations and distribution systems. Having very accurate timing across the endpoints allows the monitoring systems to readily detect faults, identify the source and extent of impact, and take corrective action in a manner affecting the smallest possible portion of the grid. For such event reporting, a 1ms timing accuracy is sufficient.

Timing accuracy is also needed for power system measurements for fault detection in the current phase. Recordings of the phase may be triggered at various points when a fault is detected. Running analysis on multiple time-aligned recordings can provide a clear picture of the extent of impact. For reliable analysis, 1 ms timing accuracy is typically needed across all recordings and is today limited mostly by SNTP time sync methods typically in use.

Monitoring the power system frequency also requires timing accuracy for fault detection and resolution. In this case, accuracy between 1 μs to 10 μs is needed to provide accurate correlation of frequency.

Finally, the accuracy needed depends on the timing role of the component within the power sub-system. E.g., a PTP sync device within the sub-network generally needs to fulfil the application requirements as discussed above (e.g., down to 1μs accuracy requirement) whereas a component that is a PTP Grand Master needs to be accurate to a 250 ns requirement [3], specifically, 250 ns are assumed in (section 7.2 [3]) in order to distribute synchronization over a chain of up to 15 transparent clocks while meeting the requirement of 1 μs at the end-device.