Please enable JavaScript.

Coggle requires JavaScript to display documents.

Implementing PTP in Substations Link1 Link2 (Precision Time Protocol (2.5…

- - - - 1.2.3.1 can be used in substation automation systems to overcome incompatibilities and shortcomings of existing time distribution systems.
      - 1.2.3.2 “Power Profile” - The operation of PTP using the is explained and examples of how PTP can be used in new and existing substations are presented.
- - - - shows separate timing and communication networks in a substation automation system.
        
        shows the use of IRIG-B for time synchronisation and Ethernet for data transfer, however RS485 may be used in older substations in place of Ethernet.
      - IRIG-B time code
        
        conveying time & date information & synchronisation pulses
        
        be transmitted as raw pulses over copper cables (coaxial or twisted pair) and fibre-optic cables, OR as an amplitude modulated (AM) 1 kHz carrier over coaxial cable.
        
        IEEE Std C37.118.1-2011
        
        provide year, time zone offset from UTC, daylight saving (Summer) time, and time quality that are essential for substation automation
        
        has a number of options
        
        cannot use one IRIG-B Signal
        
        various vendors of substation equipment can be mutually exclusive
        
        ex: P/R is configured to match the master clock: UTC/locat time modulate/unmodulated signals
        
        how the time code is formatted & transmitted
        
        challenges
        
        the burden (loading) on the time distribution network, transmission line termination, immunity to noise, galvanic isolation and wiring maintenance
        
        Master clock - 15 mA to 150 mA
        P/R - 5 mA to 10 mA
    - - an accurate synchronisation reference,
        but does not include “time of day” information
      - sufficient for SV process bus
      - but not enough for event time stamping or cryptographic message authentication (to prevent replay attacks)
      - IEC 60044-8, IEC 61850-9-2 process bus
        
        The draft IEC 61869-9 standard for merging unit communication retains 1-PPS over fibre-optic cable as an option for time synchronisation
      - requires a dedicated distribution network, which can use metallic (coaxial or twisted pair) or fibre-optic (multi-mode or single-mode) cables.
      - The rise and fall time (tf) between the 10% and 90% levels must be less than 200 ns, and the high time (th) must be between 10 µs and 500 ms (measured at the 50% level)
    - - Distribution
        
        electrical means
        
        can used multi-drop connections
        
        but this could result in potential rise between panels
        
        fiber optics
        
        ensures galvanic isolation and eliminates inductive or capacitive interference
        
        dedicated distribution repeaters are required to split the signal for each protection relay
        
        The 9-2LE guideline for IEC 61850-9-2 requires time synchronisation to be performed using fibre-optic cable
        
        This in turn requires the use of a pulse distributor or a clock with multiple outputs if there is more than one merging unit.
      - The propagation delay
        
        5 ns/meter through copper/fiber-optic cable
        
        Compensation
        
        may require compensation by the connected devices
        
        The 9-2LE guideline sets a limit of 2 µs “error” before compensation is required.
        
        Compensation is a manual process that requires specific cable lengths distribution repeater delays to be known for each connected device.
    - - Time synchronisation systems in substations have historically used a separate distribution system with its own cabling (coaxial, twisted pair or fibre-optic)
      - Data communications between protection relays and the SCADA system have no influence on the accuracy of time synchronisation. Separate systems increase the cost of construction through extra cable, terminal blocks and documentation, and this can be significant for large transmission substations.
  - - - NTP servers (clocks) and clients (protection relays) can achieve accuracies of 1-4 ms, :warning:but take care of network design to minimize packet delay variation
      - NTP vs. IGIP-B
        
        general purpose time synchronisation is time is
        always transmitted with respect to UTC
        
        IEC 61850 and IEEE Std 1815 (DNP3) require UTC
      - :smiley: supports the simultaneous use of multiple master clocks by a client for more accurate and reliable operation
      - :cry: NTP does not achieve the microsecond-level accuracy required for synchrophasors and sampled value process buses
    - - IEEE Std 1588-2008, PTPv2, 1588v2
      - :check: very accurate time synchronisation by using special Ethernet hardware that records the exact time a PTP synchronisation message is received at the Ethernet card.
      - :check: This information can compensate for the uncertainty introduced by real time operating systems and other processing delays in both the synchronisation master and the devices that are to be synchronised
      - :check: the same port can be used for IEC 61850, DNP3, IEC 60870-5-104, Modbus/IP and other substation automation protocols.
      - PTP supports multiple master-capable clocks, but these vote amongst themselves to choose a single clock to be the “grandmaster”
    - - widely used for substation automation systems
      - can be used to synchronise the internal clocks of devices throughout a substation
      - not requiring additional cabling, but does require support for suitable protocols by the various protection
        relays, power quality meters and other such devices
      - NTP and PTP can compensate for propagation delay through
        bidirectional communication
      - NTP is a more established standard and is widely used, but PTP offers greater performance through the use of special networking hardware
      - PTP requires specific types of Ethernet switch, but NTP does not
      - Redundancy - Both networked protocols support multiple master clocks
- - - - the ultimate source of time for synchronisation using PTP,
        and usually has a GPS (or other satellite system) receiver built in
    - - the source of time that other clocks on the network synchronise to
    - - the end-user of PTP, which may be a protection relay with native support forPTP or a translation device (such as Tekron’s PTP Translator) that generates a legacy timesynchronisation signal such as IRIG-B or 1-PPS
    - - an Ethernet switch that measures the time taken for a PTP synchronisation message to transit the device and provides this information to clocks receiving the PTP event message
    - - has multiple PTP ports and may serve as a source of time, i.e. be a slave clock to an upstream source and a master clock to downstream devices
  - - - These contain the time value from the master clock in the form of the number of seconds and nanoseconds since midnight on 1 January 1970. The Sync messages (using two-step clocks) are transmitted unaltered by the transparent clocks. ta represents the time at the grandmaster clock. Announce messages are treated the same way.
    - - Peer delay (Request /Response /Follow up) are exchanged between neighbours to estimate the propagation delay of each path between devices. The Peer Delay mechanism uses two or three separate message types to measure the propagation delay (depending on one-step or two-step operation).
    - - These contain the precise time stamp of when the previous Sync message was sent, along with Correction information.
      - Follow Up messages contain information updated by transparent clocks in the network. Follow Up messages will differ throughout the network, reflecting the different network delays to each node.
        
        Each transparent clock records the propagation delay of links between itself and it’s immediate peers. As a Sync message passes through a transparent clock, the clock calculates a local correction value by adding the propagation delay of the link the message arrived on and the residence time of the message within the clock. This local correction value is then added to correction field of the corresponding Follow Up message. When the messages arrive at the slave clock it adds it’s recorded link propagation delay to the correction value which then represents the total time taken for the sync message to travel from the Master to the slave, the path delay.
        
        Because the total path delay value is contributed to by each component in the path the sync message takes, the peer-to-peer mechanism used in the Power Profile is very responsive to changes in network topology.
        
        It is important to note that while the Follow Up messages may look identical, they will be different at each point in the network. Transparent clocks alter the contents of the message while retaining the original source address of the grandmaster.
      - The Correction is the sum of the transparent clock residence times and propagation delays been the grandmaster and that point in the network,and is represented as nanoseconds and fractions of nanoseconds.
    - - These are information messages transmitted by the grandmaster that provide details of time accuracy of the reference (e.g. GPS receiver) and other PTP protocol information.
  - - - The PTP standard allows for a number of options, and as with IRIG-B, some options are mutually exclusive. PTPv2 introduced the concept of “profiles” that restrict the available options and may mandate certain features for specific applications.
    - - The power industry has a profile, IEEE Std C37.238-2011, that provides a set of optimised parameters and minimal options to deliver accuracy better than 1 µs with network topologies typically found in substation automation systems. This “Power Profile” also defines a Management Information Base (MIB) for the Simple Network Management Protocol (SNMP) that allows key parameters of Power Profile devices to be monitored with industry standard network management tools. The “health” and performance of a time synchronisation system can be monitored in real-time, with alerts raised if there are any issues or abnormalities.
    - - This profile incorporates performance criteria for transparent clocks that require no more than 50 ns of error be introduced by each transparent clock. This is to ensure that the 1 µs performance target is met with 16 Ethernet switches (e.g. a ring network topology), while allowing for up to 200 ns of GPS clock error. This covers most substation networks that use a ring (as opposed to star) topology.
    - - The Power Profile requires that “peer to peer” transparent clocks be used for all Ethernet switching of PTP messages, and that all messages transmitted using multicast “layer 2” Ethernet frames. “Peer to peer” means that each PTP device exchanges messages with its neighbour to measure the path delay between them, rather than each slave clock communicating directly with the active grandmaster clock. The overall network delay is calculated by adding together the path delays and switch residence times between the grandmaster and each slave clock. This has two benefits:
        
        The PTP system automatically compensates if a network link fails and an alternative path is used. Path delays are measured on all network links, even those that are blocked to normal traffic by spanning tree protocols.
        
        The network traffic seen by the grandmaster clock does not increase as the network gets larger. The grandmaster only communicates bidirectionally with the Ethernet switch (transparent clock or boundary clock) that it is connected to.
    - - Not all manufacturers of PTP equipment support the C37.238 “Power Profile”, however the “default” peer-to-peer profile specified in Annex J.4 or IEEE Std 1588-2008 can achieve the required accuracy if configured appropriately. If non Power Profile devices are used there is no guarantee that information useful for substation applications, such as time error and local time zone offset, will be made available to client devices, or that performance has been tested and found to meet accuracy specifications (Annex J.4 does not specify performance).
    - - One such application would be to translate between the ITU-T Rec. G.82651.1 Telecommunications Profile (used for wide area networks between substations) and the IEEE Std C37.238 Power Profile (used within a substation).
    - - WANs can provide a redundant source of time in case of GPS receiver failure, at which time the grandmaster becomes a boundary clock.
  - - - Synchronisation accuracy is not affected by other network traffic, provided PTP messages are not lost due to overloading. This allows the same network infrastructure to be used for PTP and for synchrophasors, sampled value process buses, IEC 61850 (GOOSE and/or MMS), DNP3 and so forth.
    - - PTP messaging rates have been optimised to meet the 1 µs performance requirement of power system applications without placing excessive traffic on the shared network, or requiring overly complex slave clocks.
    - - Fibre-optic or twisted pair Ethernet can be used, and it is a matter of selecting Ethernet switches with the appropriate port configuration.
    - - A single time reference is used, so there are no configuration issues regarding UTC or local time. All Power Profile devices use International Atomic Time (TAI), which avoids leap seconds and daylight saving time issues.
    - - The Power Profile transmits the local time offset, so there is no need to configure the local time zone on protection relays. In addition, any changes to the dates of operation of daylight saving time only need to be made to the grandmaster rather than to every protection relay. The mechanism used is defined in IEEE Std 1588, so is compatible with devices that may not support the Power Profile.
    - - Redundant grandmaster clocks can be used, with automatic failover if the active grandmaster suffers loss of network connectivity or degradation of performance.
    - - Protocols that enable redundant Ethernet connections, such as Rapid Spanning Tree Protocol (RSTP), Parallel Redundancy Protocol (PRP) and High-availability Seamless Ring (HSR), can be used to improve the reliability of network connections between PTP devices.
    - - Networks can be expanded without placing undue network load on the grandmaster clock.
    - - Propagation delays resultingfrom long cable runs are automatically compensated for, avoiding the need to hand-tune merging units and phasor measurement units in the field.
  - - - Ethernet switches used for PTP with the Power System Profile should have specific Power Profile support if time error reporting is to be meaningful. Not all peer to peer transparent clocks will meet the requirement to introduce no more than 50 ns of error, or be able to estimate time inaccuracy.
    - - There is limited native support for PTP in protection relays, but this is improving. A number of manufacturers have released protection relays with PTP support since 2013, but this may be an option that must be specified at the time of order.
    - - Not all PTP grandmaster clocks or slave clocks (including translators) are designed for use in high voltage substations, even though they may support the Power Profile. Substation equipment should be tested for higher levels of electromagnetic compatibility (EMC) than office or light industrial equipment.
    - - Time synchronisation is critical to the operation of synchrophasor monitoring and most sampled value process buses. It is essential that only authorised people have the ability to change the configuration and operation of PTP clocks, either through dedicated configuration tools, embedded web servers or via SNMP. If clocks can be configured from the front panel then this should require the use of a password. Policies and procedures in place for protection relay configuration management should be adopted for timing systems (master clocks, transparent clocks and boundary clocks).
    - - each optimised for certain applications. The needs of substation automation systems are best met by the Power Profile, but default profiles may work, but with no certainty that this is the case. Other application specific profiles, such as the Telecoms Profile or IEEE Std 802.1AS Audio Video Profile, are most likely not going to work as the application requirements are simply too varied.