For twenty years ProgUSA has been putting test equipment in the hands of utility and industrial crews. In that time the single most common thing we hear after a transformer failure is some version of the same sentence: everything looked fine at the last inspection.
It usually did. That is the problem with a maintenance program built entirely on snapshots.
What 70B actually asks you to prove
NFPA 70B is a standard now, not a recommended practice. That shift arrived with the 2023 edition and carries through the current 2026 edition, which means mandatory language and an expectation that you run a documented Electrical Maintenance Program. Chapter 9 is where the teeth are.
Chapter 9 does not simply hand you a fixed calendar. It sorts equipment into condition levels, and it does that on three separate axes.
Physical condition is the one most people know. Condition 1 is equipment in like-new shape, in a clean and tight enclosure, with maintenance current, no open recommendations, and no unaddressed notifications from a monitoring system. Condition 2 is equipment showing drift: results deviating from past results, major component repairs in the previous cycle, active recommendations, or monitoring notifications since the last assessment. Condition 3 is equipment that has missed the last two successive maintenance cycles, or that shows changes in operation calling for action now.
Two more assessments sit alongside it. Criticality condition is assigned on operational reliability or business continuity, which is the lens most utilities already apply to a distribution transformer. Operating environment condition covers equipment running in harsh chemicals, contaminants or extreme conditions it was never rated for.
Your equipment takes the highest of the three. A transformer in perfect physical shape can still land in Condition 3 on criticality alone.
Those levels drive your intervals. Table 9.3.2 sets the baseline, and the pattern is easy to see: infrared thermography on all equipment sits at 12 months in Conditions 1 and 2, then tightens to 6 months in Condition 3.
You are allowed to move off that baseline. The standard permits altering an interval based on the potential risk to personnel or facility operations if the equipment fails to operate as expected. But where the deviation extends an interval, both the deviation and the justification for it have to be documented in the EMP.
That is the part worth sitting with. Stretching an interval is not a judgment call you get to make quietly. It is a claim you have to support. Which means the quality of your condition file decides how much flexibility you actually have.
Snapshots tell you the state. They do not tell you the trend.
Visual inspection, infrared thermography, insulation resistance, winding resistance and turns ratio. Every one of these is a good test and every one of them belongs in the program. We sell the instruments for all of them.
But they share a limitation. Each one is a measurement taken at a moment you chose, under whatever conditions happened to exist that morning. A thermographic scan run on a mild Tuesday at 40 percent load tells you very little about what that transformer sees during a July peak. An insulation resistance test tells you the state of the insulation on the day of the test. Neither one captures the switching transient at 2:40 a.m. that nobody was standing there to see.
The 70B condition framework is built around drift. Test results deviating from historical norms. Alerts from monitoring systems. Both of those phrases assume you have a record over time, not a folder of disconnected readings.
That record is what continuous power quality monitoring produces.
What a continuous PQ record adds to the condition file
Put a Class A monitor on a distribution transformer and leave it there, and within a season you have things a spot test cannot give you:
- Real loading, not nameplate. What the asset actually carries, hour by hour, through seasonal swing. This is the difference between guessing at remaining life and knowing.
- Harmonic content over time. Rising THD on a transformer feeding drives, UPS systems, or a growing nonlinear load is a thermal problem building slowly. It shows up as a trend line months before it shows up as a hot spot.
- Sag and interruption history with timestamps. When a plant calls to say their line dropped out again, timestamped event records settle the question of whether it originated upstream or inside the fence. That conversation costs a lot of money when nobody has data.
- Transient and switching event capture. Repeated switching stress is cumulative and invisible to periodic testing.
- Voltage unbalance and flicker. Both quietly shorten the life of rotating equipment.
- Surface and ambient temperature alongside the electrical data. Load and temperature on the same time axis is a far better condition indicator than either alone.
None of this replaces your periodic testing. It gives the periodic testing context, and it gives your condition assessment something to point at.
The Power Quality Monitoring Checklist for NFPA 70B
One page. What to monitor, how long to monitor it, and which records your condition file should hold when someone asks you to justify an interval.
The part almost nobody writes about: measurement class
Here is where a lot of power quality content stops short.
Not every instrument that reports harmonics and sags measures them the same way. IEC 61000-4-30 defines measurement classes, and Class A specifies the aggregation intervals, the accuracy, and the methods required for results to be repeatable. Two Class A instruments measuring the same signal are expected to agree. Below that class, they are not.
This matters the moment your data has to do more than satisfy your own curiosity.
If you are documenting a deviation from a standard maintenance interval, the evidence should be defensible. If you are in a disagreement with the utility about whose side an event came from, or supporting a warranty claim, or handing findings to an insurer, or building the capital case for a replacement, the measurement class is the difference between data that settles the argument and data that becomes the argument.
Class S has its place. It is a lighter specification, appropriate for surveys and general trending where the result does not have to hold up under challenge. Choose it deliberately, not by accident.
Where to start: the distribution transformer
If you are adding continuous monitoring to an existing program, distribution transformers are usually the highest-return place to begin. They are expensive, they have long lead times, they carry load you cannot easily reroute, and they are typically the least instrumented asset in the system relative to what they cost to replace.
The CHK Miro-F is built for exactly that job. It is a transformer monitor and logger certified to IEC 61000-4-30 Class A.
What it does:
- Three phase voltages and currents, temperatures, THD and flicker
- Surface and ambient temperature probes
- Gapless logging, so pulling data does not create a hole in the record
- 8 GB of logged memory with adjustable log interval
- Starts logging on power up
- Graphical color display, waveforms and phasor diagrams
- IP66 enclosure, powered from Phase A, internal backup battery with five minutes of standard backup
- Integrated 3G/4G cellular, with optional Ethernet, plus WiFi for the mobile app
- TCP-IP, Dyn DNS, FTP, DNP3 and MQTT
- Optional GPS and external antenna
- Optional auxiliary I/O module for DGA, hydrogen and bushing monitors, so the electrical record and the condition record live together
Data lands in AsMoSys, CHK's database platform, which is what makes this workable at fleet scale rather than one asset at a time.
CHK Power Quality builds for low voltage and medium voltage networks and their instruments are in service with utilities, railways, water authorities and industrial operators internationally.
Equipment is where it starts, not where it ends
A monitor on a transformer produces a great deal of data. Turning that data into a condition assessment your program can act on is a separate skill, and it is not one every maintenance group has on the bench.
ProgUSA has power quality engineers on staff. If you buy a Miro-F from us, you get access to them. That means help scoping what to monitor and for how long, help configuring the logging and reporting so it produces the records your EMP actually needs, and help reading the results when something in the data does not look right.
That is what we mean by Equipment plus Education plus Service. We are here to make your people better at this, not to do it instead of them.
The Power Quality Monitoring Checklist for NFPA 70B
One page. What to monitor, how long to monitor it, and which records your condition file should hold when someone asks you to justify an interval.
Frequently asked questions
Does NFPA 70B require power quality monitoring?
70B does not mandate a specific instrument. It requires a documented Electrical Maintenance Program with condition assessment, and Section 9.1.1 expressly permits continuous monitoring and predictive techniques to be used as a consideration when determining maintenance intervals. Continuous power quality data is one of the strongest ways to support a condition-based interval.
What does IEC 61000-4-30 Class A mean?
Class A is the highest measurement class defined in the standard for power quality instruments. It specifies aggregation methods, accuracy and measurement techniques such that two compliant instruments measuring the same signal produce results that agree. It is the appropriate choice when the data may need to support a dispute, a claim or a capital decision.
How long should we monitor before the data is useful?
For troubleshooting a specific recurring problem, days to weeks. For condition assessment and interval justification, plan on covering a full seasonal load cycle. Loading and harmonic behavior in February tell you very little about August.
Can we start with one transformer?
Yes, and most programs should. Pick the asset with the worst combination of replacement cost, lead time and consequence of failure. Prove the value there, then scale.
Talk to us about your program. Call +1 407 332 8678 or email info@progusa.net. We will help you figure out which assets to instrument first and what the data should be telling you.
ProgUSA has supplied electrical test and measurement equipment to utilities, NETA firms and industrial customers since 2005.
Equipment + Education + Service = Partnership