Distribution Monitoring:
5 Signals That Predict Outages Before They Happen

Why feeder-level visibility is becoming the foundation of modern distribution reliability

You already know the pressure. Load is growing faster than most planning cycles assumed. DERs are pushing power in directions the original protection schemes never contemplated. Storms keep getting worse. And the same regulators and customers who want fewer outages also want lower rates.

So the industry has responded the way it usually does — more analytics, more automation, more “grid mod” line items in the capital plan. But there’s a quieter problem sitting underneath all of it, and most of us in operations have felt it for years: You can’t analyze electrical behavior you can’t see.

Substations and transmission are instrumented to the teeth. The distribution feeder, on the other hand, still runs largely on inference — SCADA at the breaker, AMI at the meter, and a long stretch of conductor in between, where we mostly find out something was wrong after the lockout. That visibility gap is where most reliability events actually start.

The good news is that the gap is closing. A new class of distribution monitoring is making it possible to watch how circuits are actually behaving — continuously, not just when something trips. And the utilities pulling ahead are the ones capturing the fault precursors: the small electrical, thermal, and mechanical signals that show up well before a sustained outage.

Why Distribution Outages Are Processes, Not Events

Most failures are processes, not events. A loose compression connector arcs intermittently under load for weeks before it finally parts. A limb makes contact in gusts long before it brings the conductor down. Thermal cycling walks a splice toward failure across a season. The feeder is producing diagnostic signals the entire time. We just haven’t historically had the instrumentation density to hear it.

That’s the case for feeder-level visibility — and it’s why the five signals below matter more than the lockouts they eventually cause.

Signal #1: Repeating Momentary Faults on the Same Section

Every operations group has the same conversation: a momentary isn’t a “real” outage, it doesn’t hit SAIDI, the recloser did its job, move on.

Fair enough — for any single event. But the pattern is the signal. When you start seeing repeating momentary faults clustering on the same section, especially correlated with wind, temperature, or load, you’re almost always looking at one of a familiar short list of issues:

  • Vegetation making intermittent contact
  • A degrading connector or splice arcing under load
  • Conductor slap in wind
  • Insulation breaking down on a piece of apparatus

None of that is news to anyone who’s chased a recurring blink. The harder problem has always been catching the pattern before the sustained fault, when the only data point is a SCADA counter at the substation that can’t tell you where on the feeder it happened. Continuous, location-aware monitoring on the line itself is what makes the pattern legible — and what turns momentary fault data into early fault detection.

Signal #2: Harmonic Distortion That Wasn’t There a Year Ago

Harmonics used to be mostly an industrial-customer conversation. That’s changed. Solar inverters, VFDs at commercial sites, EV chargers, and the occasional rogue switching supply have shifted the harmonic profile of feeders that used to look clean.

Rising distortion — particularly at the 5th, 7th, and 11th — tells you something about what’s connected and how it’s interacting:

  • Load composition is shifting in ways your planning data hasn’t caught up with
  • A transformer is experiencing greater thermal stress than the nameplate suggests
  • You may be approaching a resonance condition with a capacitor bank
  • Power quality complaints are about to show up at the call center

Snapshot PQ studies capture a moment. Continuous waveform-level monitoring captures the trend — which is what utilities actually need for capital planning and customer power quality discussions.

Signal #3: Voltage Instability and Recurring Sags

Sags are one of those things that everybody downstream notices and nobody upstream sees clearly. If you’re getting repeated sag events on a feeder, the feeder is telling you something specific:

  • A conductor or transformer is running closer to its rating than the loading study suggests
  • A regulator or LTC is hunting or wearing out
  • A large motor start somewhere is dragging the bus
  • A DER is tripping on its own ride-through settings and yanking support away during transitions

With electrification load showing up faster than reconductoring projects can keep pace, this is going to get worse before it gets better. The feeders that matter most aren’t always the ones with the highest peak demand — they’re the ones with the most volatile load shape. You can’t see that from a monthly peak reading.

Signal #4: Phase Imbalance That’s Been There Long Enough to Feel Normal

Imbalance is one of the most under-monitored conditions on the distribution system, partly because it rarely causes a hard failure on its own. It just quietly steals capacity and shortens the life of every transformer on the feeder.

Common drivers, all of which you’ve seen:

  • Single-phase load growth that nobody re-balanced after the last few service additions
  • DER concentrated on one phase
  • An open delta or single-phase regulator that’s drifted
  • A reconfiguration after a storm that never got cleaned up

Without continuous per-phase data along the feeder, imbalance gets discovered during the next thermal failure or the next IR scan. With continuous monitoring, you can see it the day it shifts.

Signal #5: High-Frequency Transients and Partial Discharge

This is the one most utilities can’t see at all today, and it’s arguably the most useful.

At sub-cycle and microsecond-scale intervals, failing insulation and degrading hardware are loud. Partial discharge in cable splices, intermittent arcing across a cracked insulator, an arrester nearing end of life — they all radiate distinct high-frequency signatures well before the asset fails. The substation relay doesn’t see them. AMI doesn’t see them. A monthly PQ meter doesn’t see them.

You have to be on the line, sampling fast, and you have to be doing it continuously. When you are, you get one of the most genuinely predictive signals available on the distribution system — and the foundation of real predictive grid analytics.

Why the Distribution Visibility Gap Persists

None of the five signals above are new. Every protection engineer reading this could draw the waveform on a whiteboard. The reason we don’t operationalize them isn’t a knowledge gap — it’s an instrumentation gap.

What Modern Distribution Line Sensors Actually Measure

Most distribution line sensors on the market measure a handful of parameters — current, maybe voltage, fault detection, and call it a day. That’s useful, but it leaves most of the diagnostic signal on the table.

EGM sensors continuously monitor more than 40 electrical, physical, and environmental parameters at the feeder, including:

  • Voltage, current, harmonics, power factor, phase angle
  • Fault current and waveform capture
  • Conductor temperature
  • Vibration and conductor movement
  • Line sag
  • Ambient environmental conditions

The point of that breadth isn’t more dashboards. It’s that real reliability problems almost never sit inside a single variable. They sit in the correlations — and that’s where decades of chasing failures teaches you to look:

  • Harmonics climbing and conductor temperature climbing on the same span → you’re cooking a section, not just seeing PQ drift
  • Vibration and repeating momentaries during wind events → conductor slap, not vegetation
  • Humidity and high-frequency transient activity around a specific apparatus → contamination flashover risk before the next dew point

Any one of those signals in isolation is ambiguous. Together, they can form a diagnosis.

The Future of Distribution Grid Reliability Is Observability

Reliability used to be primarily an infrastructure question — bigger conductors, more ties, better protection. It’s still partly that. But increasingly, the differentiator between utilities that hit their distribution grid reliability targets and those that don’t is going to be operational awareness, not asset spend.

The feeders are already producing the signal. The question is whether we’re instrumented to hear it.

If you work in distribution operations or planning, this is the kind of thing we write about regularly — feeder-level visibility, predictive reliability, and what’s working in the field. Read more insights on feeder-level visibility, predictive reliability, and modern distribution operations from the EGM team.