The growing adoption of microgrids, renewables, and digital transformation have finally put synchronization on the front burner.
The greening of the power grid is about more than moving to renewable power generation. The distributed and intermittent characteristics of solar and wind, especially residential rooftop solar, pose different engineering challenges to the power grid. Known in the industry as distributed energy resources (DERs), renewables will ultimately introduce digital technologies for automation and oversight into many parts of the grid that were previously unmonitored. These smart grid technologies represent many significant changes, including how grid assets across the grid need to be precisely synchronized.
DERs disrupting the energy supply chain
With the proliferation of DERs, the energy supply chain is transforming. The traditional centralized architecture – from large power plants flowing to substations to customers – is changed beyond recognition. While the large power plants continue to supply the bulk of the energy, there are many DERs – from wind and solar farms to storage systems to electric vehicles (EVs) – distributed everywhere in the grid with rapidly growing generation capacity. With so many new active actors, energy can now flow in many different directions.
To integrate these new active actors into grid operations, utilities need control and automation applications. They also need line and load condition data collected at all locations of DERs and substations. More than ever, they need to deploy digital monitoring, control, and automation applications using IEDs (intelligent electronic devices). This, in turn, requires coordination of IED activities, measurement of data, and scheduling of DERs, all orchestrated between DER sites, substations, and a control center. A common time reference across the whole grid thus becomes critical to build a visualization of the grid and to coordinate temporal behavior of grid assets, particularly for time-sensitive applications.
For example, a synchrophasor phase measurement unit (PMU) — an application that measures power flow with high sample rates — requires time accuracy on the order of 1 microsecond (ms) so that utilities can time-align line measurement data across the grid to analyze grid events. Applications like digital fault recorders (DFR) need accuracy in the order of 1 millisecond (ms) to correlate with lightning strike data to determine the origin of the fault. Other grid applications, such as IEC61850 Generic Object-Oriented Substation Events (GOOSE) and Sampled Values (SVs), also require accurate time synchronization on the order of 1ms and 1ms, respectively.
Getting in sync
Distribution system operators, who in the past have only needed minimal communications coverage in most of their medium and low voltage service territories, now face the challenge of distributing time synchronization to their grid assets everywhere.
Utilities are not new to synchronization. They have long relied on time-division multiplexing (TDM) technology to distribute frequency synchronization throughout their communications network because TDM network equipment needs frequency synchronization to transmit and receive digital grid data properly.
With TDM equipment going out of support, utilities have been migrating to IP multi-protocol label switching (IP/MPLS) networks to support IP/Ethernet-based and legacy TDM-based applications. The need for frequency synchronization persists. The prevalent way to distribute frequency synchronization is synchronous Ethernet. For network domains that do not support synchronous Ethernet, Timing over Packet (ToP) technologies are also a viable option. Common ToP technologies include adaptive clock recovery (ACR) and differential clock recovery (DCR). However, these technologies fall short addressing the needs of the smart grid. The key shortcoming is their lack of capability to distribute time synchronization throughout the grid.
Time synchronization received prominent attention in the aftermath of North America’s 2003 East Coast blackout. As investigators reviewed events logs, they found conflicting time stamps on thousands of records across multiple grids due to inconsistencies that made it extraordinarily difficult to reconstruct the incident and identify the root cause.
The North American Electric Reliability Corporation (NERC) responded to this situation, recommending that internal clocks for disturbance measuring equipment be synchronized to within 2ms. While the sector overall has been slow to move on that guidance, the demands (and opportunities) of smart grids are renewing interest and bringing the issue back to the forefront. New communications networking solutions are coming onto the scene to ensure reliable and accurate time distributions across the grid.
IP/MPLS networks for time synchronization
The advent of digital substations powered by IEC 61850 brings time synchronization into the spotlight. To support the timing needs of digital substations, utilities have depended on global navigation satellite systems (GNSS), e.g., GPSas the source at each substation. It is not practical, however, to equip every substation with antenna and receiver infrastructure to receive GNSS signals, and signals are susceptible to natural, accidental, and intentional interferences. Also, as IEDs proliferate within smart grids, it isn’t possible to connect every IED to the GNSS receiver with an Inter-range Instrumentation Group (IRIG) interface over copper cable.
Since any degradation in timing synchronization will impact grid applications and operations, it is important that utilities can distribute accurate time from their control centers to all substations over their wide area network (WAN) — as the backup clock for substations with GNSS, and as the primary clock for substations, without.
Many utilities have already evolved their WANs to IP/MPLS networks to bring connectivity to substations. This makes the WAN an attractive option to distribute synchronization for both frequency and time. In light of this, the International Electrotechnical Commission (IEC) defined a new profile for power utility automation based on IEEE1588, IEC 61850-9-3, to fulfill these needs of synchronizing grid assets. The Institute of Electrical and Electronics Engineers (IEEE) also defined a power profile in IEEE C37.238. And in cases where the WAN only supports the ITU-T (International Telecom Union – Telecom sector) telecom profiles, time synchronization can still be brought inside substations with profile interworking performed by the substation router.
An IP/MPLS WAN combined with IEEE1588 makes it possible to meet the NERC targets and digital substation synchronization needs. IEEE1588v2 defines a Precision Time Protocol to synchronize clocks in a packet-switched network. It accurately distributes timing information from a central source to all substations connected to the network, with hardware-assist capability to automatically compensate for accuracy loss incurred by network impairments. It also provides multiple redundancy protection schemes for high resiliency at the synchronization and network layers. In this way, it can act either as the primary timing source or as a backup to GPS timing in case of signal loss.
Synchronization is not just integral to communications networks; it is foundational to digital grid operations. Grid applications, including synchrophasor and GOOSE, require their IEDs to have reliable access to accurate time. The NERC has been recommending better synchronization for close to two decades. The IEC has also facilitated the introduction of synchronization with their standard works. The growing adoption of microgrids, renewables, and digital transformation have finally put synchronization on the front burner. With the arrival of IEEE 1588 implementation in today’s IP/MPLS networks, the timing couldn’t be better.