Key Points

 It seems likely that a brief arc fault caused a 30-millisecond voltage dip, leading to the
simultaneous reboot of all devices.
 The grounded neutral in an otherwise unearthed system may have contributed to this
issue, making the system more susceptible to faults.
 Research suggests no physical evidence was found, but a transient fault could explain the
lack of tripped breakers or logged faults.
 The evidence leans toward a power distribution issue, possibly in the ACDB or cabling,
rather than a UPS failure.

System Overview

Your plant uses a 110V single-phase UPS system with two 50 KVA parallel redundant UPS
units supplying power to an ACDB, distributing to DCS controllers, PLCs, and field equipment
like SOVs and transmitters. All devices rebooted simultaneously for about 30 milliseconds, with
no fuses blown, MCBs tripped, or UPS fault logs recorded.

Likely Cause

It appears a transient arc fault in the power distribution system, such as the ACDB or cabling,
caused a brief voltage dip. This dip, lasting around 30 milliseconds, was enough to reset sensitive
devices but too short to trigger protective devices or be logged by the UPS. The grounded
neutral, despite the system being described as unearthed, may have made the system more
vulnerable to such faults by providing an unintended path for fault currents.

Recommendations

Consider installing power quality monitors to detect brief voltage dips and inspect the ACDB
and cabling for wear or potential arc paths. Verify the grounding configuration to ensure it aligns
with the system’s design.

Detailed Analysis and Survey Note

Introduction

This report investigates the simultaneous reboot of all DCS controllers, PLCs, and other loads
connected to a 110V UPS supply in your plant, occurring on an unspecified date prior to May 24,
2025. The system comprises two 50 KVA parallel redundant UPS units supplying 110V to an
ACDB, with a total load of 100A shared equally. No protective devices tripped, no fault logs
were recorded, and inspections found no physical evidence of faults. The system is described as
unearthed, yet the neutral is connected to ground, presenting a potential anomaly. This analysis
aims to identify the root cause and provide recommendations for further investigation.

System Configuration and Observations

 UPS System: Two 50 KVA single-phase UPS units operate in parallel, redundant
configuration, supplying 110V to the ACDB. Each UPS shares 50A of the total 100A
load, with a static switch inside the UPS before the common output.
 Load Distribution: The ACDB distributes power to DCS controllers, PLCs, and field
equipment (e.g., SOVs, annunciation panels, transmitters), all protected by MCBs. Some
field equipment is connected via underground cables.
 Incident Details: All devices rebooted simultaneously, with a 30-millisecond gap in the
DCS event log, indicating the total time from power loss to reboot. No fuses blew, no
MCBs tripped, and no UPS fault logs (e.g., inverter issues, load sharing problems, or
synchronization glitches) were recorded.
 Inspections: Thorough checks of the ACDB and UPS found no short circuits, hotspots,
or signs of vermin. Cable connections were tight, with no micro-arcing observed. Power
quality checks showed no abnormalities, and infrared scans confirmed no hotspots.
 Anomaly: The system is described as unearthed, but the neutral is connected to ground,
which could cause issues if a phase gets earthed, potentially leading to voltage dips.
However, no undervoltage was recorded in the UPS system.

Potential Causes and Analysis

Several hypotheses were considered to explain the simultaneous reboot without triggering
protective devices or logging faults:

1. Voltage Dip or Sag:

o A brief voltage dip could reset sensitive electronic devices like DCS controllers
and PLCs. The 30-millisecond disruption aligns with durations that can cause
resets, especially if the dip was deep (e.g., 50% or more). However, the UPS
should mitigate input voltage dips, suggesting the issue was downstream or
internal.
o Research from Impacts of Voltage Dips on Power Quality Problems indicates
voltage dips can result from faults like short circuits, often caused by switching
operations or insulation failure, leading to equipment malfunctions.
2. Grounding Anomaly:
o The neutral being grounded in an otherwise unearthed system is a significant
concern. In an unearthed system, no intentional ground connections should exist,
but here, the neutral is earthed, potentially creating a path for fault currents. If a
phase-to-ground fault occurred, it could cause a voltage dip without tripping
breakers, especially if the fault was high-impedance or brief.
o Understanding Neutral Earthing in UPS Systems discusses how grounded neutrals
can affect system behavior, particularly in transformer-based UPS, which may
apply here given the power rating.
3. Parallel UPS Synchronization Issue:
 oIn parallel redundant UPS systems, synchronization is critical for load sharing. A
momentary loss of synchronization could cause a brief power interruption.
However, no sync glitches were recorded, and both UPS units continued
operating, making this less likely. Comparing UPS System Design Configurations
highlights synchronization as a potential issue, but the lack of logs suggests it was
not detected.
4. Transient Arc Fault:
o A brief arc fault in the power distribution system, such as the ACDB or cabling, is
the most plausible explanation. Arc faults can cause voltage dips without leaving
physical traces if they are self-clearing and brief. The 30-millisecond disruption
fits the profile of such events, and sensitive devices could reset due to the voltage
dip. The Problem of Arcing Faults in Low-Voltage Power Distribution Systems
details how arc faults can occur in low-voltage systems, often due to loose
connections or insulation damage, and may not trip breakers if short-lived.
o The grounded neutral could exacerbate this by providing a return path, increasing
the likelihood of such faults affecting the system.
5. External Disturbance:
o An external event like a lightning strike or grid transient was considered, but no
input supply disturbances were recorded, making this less likely. What Causes
Arc Flash? Electrical Arc Blast Explained mentions voltage transients from
switching or lightning as causes, but the stable input supply rules this out.

Evaluation and Conclusion

After evaluating the above, the evidence leans toward a transient arc fault as the root cause.
This fault likely occurred in the ACDB or cabling, causing a 30-millisecond voltage dip that
reset all devices. The lack of tripped breakers or logged faults is consistent with a brief, self-
clearing arc fault, which may not leave physical evidence. The grounded neutral, despite the
system being described as unearthed, may have contributed by making the system more
susceptible to such faults, providing an unintended path for fault currents.

The 30-millisecond duration aligns with known effects of voltage dips on sensitive electronics,
as SMPS in DCS and PLCs typically have hold-up times of 10-20 milliseconds, and a complete
power loss for 30 milliseconds would likely cause resets. The absence of physical evidence (e.g.,
hotspots, blown fuses) supports the transient nature of the fault, as arc faults can be too short to
cause lasting damage detectable by standard inspections.

Detailed System Considerations

 Load and Protection: The total load is 100A, with a single 100A isolator, and each UPS
shares 50A. Field equipment is connected via underground cables, protected by MCBs,
which did not trip, suggesting the fault current was limited or brief.
 Power Quality and Inspections: No abnormalities were found in power quality checks,
and infrared scans showed no hotspots, reinforcing the transient nature of the issue. All
cable connections were tight, with no signs of micro-arcing, but a brief arc fault could
still occur without leaving traces.
  Timing and Impact: The 30-millisecond gap in the DCS event log indicates the system
was offline for that duration, likely due to the voltage dip. This short duration is
consistent with arc faults, which can cause rapid, self-clearing disruptions.

Recommendations for Further Investigation

To confirm the root cause and prevent future incidents, consider the following:

 Power Quality Monitoring: Install devices like oscilloscopes or power quality analyzers
to capture brief voltage dips and transients, helping identify future arc faults. This can
provide data on voltage profiles during such events.
 Detailed Inspection: Conduct a thorough inspection of the ACDB and cabling, focusing
on potential arc paths, even if no evidence was found initially. Look for signs of wear,
insulation damage, or loose connections that could predispose to arc faults.
 Grounding Verification: Verify the grounding configuration to ensure it aligns with the
system’s design (unearthed vs. earthed). If the neutral should not be grounded, investigate
and correct this anomaly to reduce susceptibility to faults.
 Protection Upgrades: Consider upgrading to Arc Fault Circuit Interrupters (AFCIs) or
similar devices, especially for critical systems, to detect and interrupt arc faults before
they cause disruptions.

Table: Summary of Key Findings and Actions

Aspect Finding Action

Incident 30 milliseconds, as per DCS Monitor for similar brief disruptions using
Duration event log gap power quality analyzers.
Protective No fuses blown, no MCBs Investigate why no tripping occurred; consider
Devices tripped AFCI installation.
No fault logs, no sync glitches, Verify UPS logging thresholds for brief
UPS Logs
stable input supply events; ensure proper configuration.
Physical No short circuits, hotspots, or Re-inspect ACDB and cabling for subtle signs
Inspection arcing observed of wear or potential faults.
Grounding Neutral grounded in unearthed Verify and correct grounding configuration to
Anomaly system align with design.

Conclusion

The simultaneous reboot of all devices is most likely due to a transient arc fault causing a 30-
millisecond voltage dip, with the grounded neutral contributing to system vulnerability.
Implementing the recommended actions will help confirm this cause and enhance system
reliability.