Condition-based maintenance is a strategy that schedules substation service around the measured condition of each asset rather than a fixed calendar interval. Instead of overhauling a transformer or circuit breaker every set number of years, it uses data (dissolved gas analysis, infrared thermography, partial discharge readings, contact resistance, and online sensors) to act only when measurements show developing degradation. It beats time-based service when equipment fails randomly rather than on a predictable wear curve, and when an unplanned outage costs far more than continuous monitoring. With large power transformer lead times now stretching to several years, keeping the assets you already own healthy has become a reliability priority, not just a cost decision. Across our 18-plus years of substation field work from 12 kV through 500 kV, condition data has consistently pointed to problems a fixed calendar could overlook.
What Condition-Based Maintenance Actually Means
Three terms get used loosely in substation reliability work, and the differences matter. Time-based maintenance (also called calendar or interval-based preventive maintenance) assumes wear is predictable and services equipment on a fixed clock, every two years, every six years, and so on, regardless of how the asset is actually performing. Condition-based maintenance instead monitors condition indicators (temperature, dissolved gas, pressure, partial discharge, contact resistance) and triggers work only when the data shows a meaningful change. Predictive maintenance is the next step: it applies analytics to those condition trends to forecast how much useful life remains before a fault develops.
Sitting above all of them is reliability-centered maintenance (RCM), the framework that decides which strategy fits which asset. RCM and condition monitoring are not competing ideas. RCM tells you which assets justify the investment in monitoring, which can safely run to failure, and which still belong on a fixed interval. A sound program uses all three together rather than treating one as a replacement for the others.
When Condition-Based Maintenance Beats Time-Based Service
The deciding factor is how an asset fails. Time-based service works well for components with predictable, wear-out failure patterns, such as consumables and simple mechanical parts that degrade on a clock. Condition monitoring earns its keep where failures arrive more randomly but announce themselves first through detectable precursors. Power transformers, bushings, and breaker interrupters fit that profile: insulation degradation, gassing, and contact erosion all leave measurable fingerprints long before a hard failure.
The second factor is the cost of being wrong. When a single unplanned transformer loss can take a facility offline for months, continuous condition data is cheap insurance. That math has shifted sharply in the last few years. Lead times on large power transformers have stretched from roughly 30 to 60 weeks before the pandemic to an average of about 128 weeks in 2025, with some high-capacity units quoted at four years. Demand for substation transformers rose roughly 116 percent between 2019 and 2025, and generator step-up transformer demand climbed about 274 percent, driven largely by AI data center load and electrification. When you cannot simply order a replacement, extending the life of the asset you have stops being optional.
Time-based service still has its place. Safety-critical and regulator-mandated checks, assets without a reliable condition signal, and low-cost equipment where monitoring would cost more than the part are all reasonable candidates for a fixed interval. The goal is not to eliminate calendar work but to stop over-maintaining healthy equipment, since every intrusive overhaul carries its own risk of introducing a defect.
The Diagnostic Tools That Feed a Substation CBM Program
Condition-based maintenance is only as good as the data behind it. In substations, that data comes from a handful of established diagnostic methods, most governed by IEEE and NETA standards:
- Transformer oil testing through dissolved gas analysis (DGA). Often called the blood test for a transformer, DGA tracks gases such as hydrogen, methane, acetylene, and carbon monoxide dissolved in the insulating oil. IEEE C57.104-2019 is the guide for interpreting those gases, and trending DGA alongside load data can flag many developing faults well before they cause a failure.
- Online monitoring. IEEE C57.143-2024, the guide for applying monitoring equipment to liquid-immersed transformers, covers continuous bushing monitors and gas-in-oil monitors that bridge the gap between periodic offline tests. Worth noting: that standard covers how to apply sensors, not how to interpret the results, so field judgment still matters.
- Infrared thermography. A longwave IR camera finds loose or corroded connections on energized equipment before they overheat and fail. NFPA 70B now treats this kind of inspection as a maintenance requirement rather than a best practice.
- Partial discharge testing. Ultra-high-frequency and acoustic emission sensors detect the small electrical discharges that signal insulation breakdown in switchgear and GIS. It has become a gold-standard health check for high-voltage assets under a condition-based program.
- Circuit breaker diagnostics. Contact resistance, insulation resistance, and time-travel testing reveal mechanical and dielectric wear. Our companion guide on high-voltage circuit breaker testing intervals covers how often each design needs attention.
How Predictive Analytics Is Changing the Equation
The market is moving quickly toward turning that raw condition data into forecasts. Market research firms estimate the transformer monitoring systems market at roughly 2.6 to 3 billion dollars in 2024, growing at a high single-digit rate. Vendors and researchers are training machine-learning models on fleet failure data to forecast insulation problems months in advance, though published accuracy claims vary widely. Combined with the transformer supply crunch, that predictive lead time is exactly what lets an operator schedule an intervention on its own terms instead of reacting to an outage.
The caution is that analytics is only as trustworthy as the sensors feeding it and the expertise reading it. Our founder, Brad Webb, served as an Electric Standards and Work Methods Specialist at PG&E before launching Substation Solutions, and that background shapes how we approach this: data sets the trigger, but an experienced engineer still confirms the diagnosis and sets the repair scope. A dashboard that no one can interpret is not a maintenance program.
Action Steps: How to Start a Condition-Based Program
- Inventory and rank your assets by criticality. Use reliability-centered thinking to separate the transformers and breakers that warrant monitoring from the gear that can stay on a calendar or run to failure.
- Establish baselines. Capture a starting DGA sample, a thermography survey, and partial discharge readings so future measurements have something to trend against.
- Match the method to the asset. Decide which units need continuous online monitoring versus periodic offline testing, guided by criticality and replacement lead time.
- Set data-driven thresholds, not just pass or fail. Track the rate of change in gas levels, temperatures, and discharge activity, because the trend often warns earlier than any single reading.
- Keep a calendar floor for safety and compliance. Condition monitoring supplements mandated checks; it does not replace them.
- Review and refine intervals as data accumulates. The longer the program runs, the better you can extend service intervals on healthy assets and concentrate effort where the data points.
Frequently Asked Questions
What is the difference between condition-based and preventive maintenance?
Preventive (time-based) maintenance services equipment on a fixed schedule regardless of its actual state. Condition-based maintenance services equipment based on measured condition data, intervening only when readings show developing degradation. The first prevents wear-out failures on a clock; the second catches random faults that show detectable warning signs.
Is condition-based maintenance the same as predictive maintenance?
They are closely related but not identical. Condition-based maintenance acts on current condition readings. Predictive maintenance adds analytics to those readings to forecast remaining life and a likely failure window. Predictive maintenance is essentially condition-based maintenance with a forecasting layer on top.
Which substation assets benefit most from condition monitoring?
High-value, long-lead, hard-to-replace assets with detectable failure precursors benefit most: power transformers, bushings, circuit breakers, and gas-insulated switchgear. With transformer lead times running several years, these are exactly the assets where life extension matters most.
Can we stop time-based maintenance entirely once we adopt CBM?
No. Safety-critical inspections, regulator-mandated checks, and assets without a reliable condition signal still belong on a fixed interval. A mature program blends condition-based, time-based, and run-to-failure strategies, choosing the right one per asset.
What standards govern substation condition monitoring?
Key references include IEEE C57.104-2019 for dissolved gas interpretation, IEEE C57.143-2024 for applying monitoring equipment to liquid-immersed transformers, NETA MTS-2023 for maintenance test intervals, and NFPA 70B for inspection requirements including infrared thermography.
How early can condition-based methods detect a developing transformer fault?
It depends on the fault and the method. Trended dissolved gas analysis often reveals developing faults weeks to months before failure, and predictive models aim to extend that window further.
Build a Program Around Your Assets, Not the Calendar
Whether you are standing up a condition-based program from scratch or refining intervals on equipment you cannot afford to lose, the right starting point is an honest look at which assets carry the most risk. Schedule a consultation with Substation Solutions and we will help you map a monitoring plan across your 12 kV through 500 kV fleet. For ongoing support, see our maintenance services, and for lifecycle and upgrade planning, our high-voltage consulting team can help you decide where to invest first.
