Why Milliseconds Decide Your Trip Investigation
Every trip investigation starts with one question. What happened first? If your DCS, PLC, and vibration monitor disagree about the time, the answer disappears. I once spent three days reconstructing a compressor trip by hand because two systems ran 40 seconds apart.
The instrumentationtools.com article "Synchronizing Time in Industrial Automation Applications" covers the core methods. It explains NTP, SNTP, and PTP with a clear plant example. The message is simple: pick the protocol that matches the accuracy your events demand. Moreover, anchor everything to one master source. I agree completely, and I would push it further. Time sync belongs in the safety management system, not in the IT afterthought folder.
Protocol Accuracy: Know Your Numbers Before You Choose
The article quotes figures every engineer should memorize. First, NTP achieves roughly 10 ms accuracy on a decent plant network. Second, SNTP — its simplified cousin that skips delay calculations — tops out around 50 ms. Moreover, PTP reaches about 1 ms per the article, and sub-microsecond in practice with boundary-clock switches.
- NTP: master-slave polling over Ethernet; stratum 0 (reference clock) to stratum 3. Good for DCS servers, SCADA nodes, and HMIs.
- SNTP: same model, no delay compensation. Use it only for small, non-critical islands.
- PTP (IEEE 1588): master pushes timestamps continuously; slaves correct offsets in real time. However, it requires PTP-capable switches and NICs.
Therefore, use NTP for the general plant layer. Reserve PTP for protection relays, phasor measurement, and high-speed machinery protection where event order matters at the millisecond scale.
Step-by-Step: Synchronizing a Yokogawa CENTUM VP DCS
Here is the deployment sequence I use on CENTUM VP projects.
Step 1
Install a GPS time server and set it as your stratum 1 master. Feed it UTC, not local time, and let clients apply timezone offsets.
Step 2
Point the highest-level station (typically the HIS or an automation server on the Vnet/IP domain) at the GPS server as its NTP source.
Step 3
Configure the FCS stations to receive time from the upper station through the CENTUM VP time synchronization function. This keeps field controllers aligned with the operator view.
Step 4
Force a time adjustment during commissioning and verify alarm timestamps in the HIS operation log match within one second.
Step 5
Document the stratum chain: GPS (stratum 0) — time server (stratum 1) — DCS station (stratum 2). Never create a second undocumented master.
Case in Point: Bently Nevada Machinery Monitoring
Machinery protection adds a second clock consumer. A Bently Nevada 3500 rack with a 3500/22M transient interface captures high-speed vibration data. However, that data is worthless without correct timestamps. When System 1 correlates a vibration spike with a DCS process alarm, both clocks must agree.
Step 1
Configure System 1 servers to sync from the same plant NTP master as the Yokogawa DCS.
Step 2
Verify the 3500 rack time matches System 1 by comparing event lists from a deliberate test point.
Step 3
After every rack power cycle, re-check rack time. Some configurations lose the clock unless the battery or sync path is verified.
Step 4
During trip reviews, export DCS and System 1 event lists into one timeline. Confirm cause and effect order within a one-second window.
Troubleshooting Drift Like a Professional
Check four things when clocks drift. First, confirm the GPS antenna has an open sky view. Second, check NTP stratum reachability with the server status page or w32tm /query /status on Windows hosts. Third, audit virtual machines — hypervisors often desynchronize guest clocks during host snapshots. Finally, look for duplicate NTP masters introduced by well-meaning IT staff. However, remember the article's core warning: a single wrong device can seed confusion across every time-stamped record in the plant.
Conclusion & Action Advice
Time synchronization is cheap insurance for expensive investigations. The instrumentationtools.com article gives you the protocol fundamentals. My field additions give you the deployment order. Finally, act now: inventory every clock-bearing device in your plant, define one GPS-anchored stratum chain, and audit it quarterly. When the next trip happens, your Yokogawa alarms and Bently Nevada events will tell one coherent story.