When root causing and diagnosing problems within an electrical system, every millisecond counts.

By: Allan Evora

Time synchronization is often overlooked in mission critical power when specifying hardware, proper controls, and data acquisition. I’m going to explain why, with a little diagram of a double ended substation.

In this main-tie-main scenario we’re going to look at the importance of a PLC that’s controlling the main and tie breakers. This PLC is always looking upstream to make sure:

  1. It has two sources of power
  2. Feeder breakers are fed power by one of the two sources at all times.

Its responsibility is to determine when to open and close certain breakers to make sure power continues to flow downstream.

 

Why is time synchronization so important?

Time synchronization is most important when it comes to root causing or diagnosing a problem with controls or controlled hardware. Say the controls scheme fails, or the system doesn’t sequence properly. One of the first things an owner/consultant would want to look at is the SCADA system to determine which events occurred in what sequence.

In electrical systems, events occur very quickly. The cycle of electricity is 60 cycles per second.

If something beyond normal operation happened in your system, you’d want to see the exact point in time the system realized there was a loss of source power. How would you do that if the timing was seconds or even milliseconds off?

In order to know what events occurred first to last, it’s important that any pieces of hardware reporting events to your SCADA system are on a common time base. CPUs, redundant CPUs, and computer clocks tend to drift if they aren’t synced to a common time base.

GPS is a known time base accepted throughout the world. It’s important that your GPS clock distributes its time to the PLC sequence of events modules and other devices in the system. This distribution occurs through a protocol used in electrical and automation industries, called IRIG-B, which uses a coaxial cable.

The importance of IRIG-B

IRIG-B exists because it’s deterministic. Once the signal is sent out from the clock, we know exactly how long it will take to arrive and be recognized by the pieces of equipment connected to the clock.

In contrast is a common network protocol for synchronization of time called NTP – network time protocol. The reason the electrical industry shouldn’t use computers acting as NTP servers to serve time out to the CPU is because it serves it over Ethernet.

Ethernet is nondeterministic. When you send a message out via Ethernet, you don’t actually know when it will be received by the CPU. Common side effects of Ethernet like network latency and collisions almost always cause time synchronization issues. Using timestamps served over Ethernet is like trying to solve a crime without any facts.

That’s why IRIG-B was developed.

IRIG-B has a resolution of 1 millisecond. In electrical systems, a cycle of electricity is about 16 milliseconds. Having a 1 millisecond time resolution you can count on ensures that when you root cause and analyze events, you can trust the information in your report.

If the clocks on the internal mechanisms were off by a couple milliseconds, you might not be able to definitively tell which one occurred first because your time base is incorrect.

 

A quick note about IRIG-B distribution

Some people think, “I can reduce my hardware costs by specifying fewer clocks in the system, more distributed clocks, and shorter networks!”

Here’s the problem with that.

Yes, IRIG-B can be distributed to many devices and distributed over large distances. There are converters that convert the IRIG-B coaxial signal to fiber optic. But when you extend your network with fiber optic, you introduce network latency.

If you do decide to go that route, you must account for that delay within your time scheme design to ensure it doesn’t introduce synchronization issues.

 

Moral of the story…

When you design your time-sensitive applications and systems, make sure devices that record events (like sequence of events from PLC, smart meters, or smart relays) are synchronized using a GPS clock and IRIG-B protocol.