New protective relays win the race against time

Since the mid-1980's, current differential protection has been implemented in digital relays, and widely employed as the main protection for transmission and distribution networks worldwide.

Alstom has recently launched new additions to its P540 current differential protection range. The P545 and P546 use GPS time synchronism to allow the relays an accurate common timing signal utilised in the operation of the differential protection. These relays are also the first in the field to use a fallback technique to allow continued use even in the event of a GPS failure.

The basic operating principle of current differential protection is to calculate the difference between currents entering and leaving a protected zone. The protection operates when this difference exceeds a set threshold. The relays send each other data messages along the telecomms link to compare the currents. If there are out-of-limit differences in the currents, the protection will operate.

Electrical utilities are being forced to seek savings. One such saving is to lease a link on a telecomms network for use by the protection communications instead of installing and maintaining their own. The performance of this channel is critical to the reliable operation of the protection, but as the links are no longer under the direct management and supervision of the power utility, with no guarantee as to the path of data messages within the network, there is therefore no guarantee of communication delay times.

Traditional time alignment

To calculate differential current flow between line ends, it is necessary that the current samples from each end are taken at the same moment in time. With the traditional method, it is necessary to measure, and compensate for, the communications delay of the channel (typically at least 3ms). The relays at ends A and B sample signals constantly, but the sampling instants at the two ends will not, in general, be coincidental.

Assume that relay A sends a data message to relay B. The message arrives at end B after a delay, and relay B registers the arrival time of the message. Relay B also sends out data messages to end A. From the time tagged to the message and the registered arrival time, the relays are able to calculate the delay in the communication network.

In this traditional technique, the relays assume that the transmit and receive channels of the communications link follow the same path and so will have the same delay time.

Knowing this propagation delay, it is then possible to align the relay's messages and provide protection.

Limitations of the traditional method

The traditional method offers reliable protection provided that the delays are equal. For long distances (over 80km), a multiplexed telecomms link is used as it is the most economical. The proviso is that the delays must be very similar or identical. The transmit and receive paths for the relay data messages must therefore take the same routing within the telecomms network.

Synchronous digital hierarchy systems

SDH-SONET (synchronous ring network) hierarchy is becoming increasingly more commonplace in telecomms networks worldwide. Typically, ring network methods are employed and these have the capacity to switch routes in the event of a path failure.

Figure 3 (figures currently not available) shows a simple ring with 6 nodes. Two P540 relays are situated at nodes B and C. Under healthy conditions, equipment at B communicates with equipment at C directly and vice-versa. Thus, the communications delay between the relays will be the same and the tradit-ional technique could be used.

Now assume that the transmit channel (only) from end B to end C fails. The SDH method allows the route to be switched, with the messages from relay B to relay C now travelling via nodes A, F, E, D, and then to C. The delay increases dramatically, owing to the longer path.

It is possible that the failure which disrupted the direct transmit path from relay B to C did not, however, result in a break in the receive direction from relay C to B. Now there is a major difference in the delay times and the relay could mal-operate because it is comparing currents from different times.

GPS synchronism with receiver module

To allow the relay to calculate the propagation delays in each direction, ie transmit and receive, independently, the new technique involves the use of an accurate GPS clock signal as a standard time signal to all relays. Because each relay is provided with a simultaneous common time reference, the delay can be calculated accurately.

Standardised time is implemented by connecting the protection relays to the GPS receiver module. Typically up to four relays may be synchronised by one GPS receiver module, making a cost-effecive solution.

Each GPS module is connected to a small antenna fixed outside the substation building, or inside, near a window or skylight. The GPS module indicates how many satellites are being tracked, allowing optimum positioning of the antenna to synchronise with the greatest possible number of satellites.

Dependability of GPS signals

When GPS synchronism is used, the likelihood of GPS failure or outage must be considered. In practice, where the antenna is well-sited, the loss of GPS signals is rare. Even rare military jamming of satellites is usually limited, so the timing reference is unaffected.

However, as current differential protection is often main protection on transmission and distribution circuits, steps must be taken to ensure that differential protection is maintained, even when GPS synchronism is lost.

Fallback position

Loss of GPS is extremely unlikely to be followed by a second failure of telecomms that results in unequal delays.

Where current differential protection is GPS synchronised, it is still possible to run the traditional technique for propagation delay measurement. Thus, whether GPS is available or not, the delays can be determined. If upon loss of GPS the delays remain unchanged, it can be deduced that no switching of telecomms route has occurred. Differential protection can then continue, assuming the same delays as recorded prior to the GPS loss. It does not matter whether or not the delays were equal prior to the loss of GPS as the relays were synchronised with the last available signal.

It is only where a double failure occurs (GPS loss and route switching resulting in changed delays) that reliable time alignment can no longer be guaranteed. For a loss of GPS, followed by a switch in the telecomms path, the protection would then need to bring backup (non-differential) elements into service.

Provided that telecomms route switching has not occurred, (even an average of once a year is unlikely) the "fallback" technique described can remain in service indefinitely.