Vacuum Breaker Leaking? What Vacuum Loss Really Means
When a technician says a vacuum breaker is leaking, the phrase is almost always shorthand for something different from what it sounds like. A vacuum interrupter is a sealed ceramic or glass bottle holding a vacuum on the order of 10⁻⁷ Torr. Nothing leaks out of it. What happens instead is that air leaks in, raising internal pressure and eroding the dielectric strength that lets the bottle interrupt fault current at all. Once internal pressure climbs toward 10⁻³ Torr, dielectric strength falls off sharply and the interrupter should be pulled from service. So the real question is never whether the breaker is leaking. It is how fast the vacuum is degrading, what is driving it, and whether the bottle has enough remaining life to survive the next maintenance cycle.
What Leaking Actually Means Inside a Vacuum Interrupter
In an SF6 breaker, a leak is a maintainable condition. You weigh the gas, trend density against temperature, find the leak path, repair at the source, and refill. That workflow is the basis of SF6 leak detection and repair, and why SF6 gas handling and recycling procedures exist at all.
A vacuum interrupter offers none of that. Once atmosphere has migrated past the ceramic-to-metal seals or through a fatigued bellows, the only remedies are bottle replacement or full breaker remanufacture. Across our work in the 12 kV through 500 kV range, that distinction is the most common source of confusion in a mixed fleet: the diagnostic that works on a gas breaker tells you nothing about a vacuum one.
Why More Fleets Are Facing This Question in 2026
The vacuum population is growing because of regulation, not preference. Under the EU F-gas framework, new medium-voltage switchgear up to and including 24 kV can no longer use SF6 for insulation or arc quenching as of January 1, 2026, extending to equipment above 24 kV and up to 52 kV in 2030. In the United States, the California Air Resources Board finalized amendments phasing out SF6 in gas-insulated equipment beginning in 2025, and OEM portfolios have followed.
At medium voltage the replacement technology is overwhelmingly vacuum, and retrofits often drop vacuum interrupters into existing metal-clad compartments. Asset managers who spent a career trending gas density now own equipment where the failure mode stays invisible until you test for it. Our technical evangelist, Amie Wallace, co-chaired TechCon 2024 and runs the SF6 and Beyond seminar series because this transition opens diagnostic gaps faster than maintenance programs update.
The Four Ways a Vacuum Interrupter Loses Vacuum
Loss of vacuum is not one failure. It is four, and they progress on very different timelines.
Bellows fatigue. The bellows lets the moving contact stem travel while keeping the envelope sealed. It is a wear item with a finite operation count, and long strokes or misaligned mechanisms cause kinking that shortens life well before the rated count.
Seal and envelope cracking. Ceramic-to-metal seals fail from mechanical shock, mishandled bottles, and repeated thermal cycling. This one is usually introduced during storage, shipping, or a rushed retrofit rather than in service.
Arc-driven outgassing. Every interruption vaporizes contact material and heats internal surfaces, releasing trapped gas. Getter materials absorb much of it, but heavy fault duty accelerates the climb.
Manufacturing defects and virtual leaks. A slow leak present from new can take a decade to surface, which is why a pass result at commissioning proves very little about year fifteen.
What a Hipot Test Can and Cannot Tell You
The standard field check is an AC overvoltage test across the open contacts. ANSI/NETA acceptance and maintenance specifications both direct that it be performed in strict accordance with the manufacturer's instructions. For scale, a new 15 kV class interrupter is commonly tested near 36 kV AC, with in-service testing at roughly 75 percent of that, or about 27 kV.
Two limits matter. First, applying AC voltage across open vacuum contacts generates X-radiation, with meaningful emission beginning around 35 to 40 kV. NETA specifications call for safeguards against it, and exceeding the manufacturer's value is a personnel hazard, not a judgment call. Second, the test is binary. A bottle just past the point of failure can condition itself during the test and pass, so a clean hipot is evidence of one moment, not of remaining life.
Magnetron Atmospheric Condition Testing and Predictive Vacuum Assessment
Manufacturers verify vacuum at the factory using a Penning discharge measurement. With the contacts open, a high DC voltage is applied and a small leakage current flows. A strong magnetic field traps that current inside the envelope, and the resulting discharge is proportional to the gas molecules present: fewer molecules, better vacuum, lower current. Portable magnetron atmospheric condition test sets now bring that measurement into the field.
The value is the trend, not the single reading. Because the output approximates internal pressure rather than returning a pass or fail, successive measurements across scheduled outages extrapolate toward the point where the bottle will no longer interrupt reliably. That turns vacuum integrity into a planning input, the same logic behind partial discharge testing and transformer oil testing elsewhere in the station. Caveats: the breaker must be de-energized with contacts open, and readings only compare when test conditions repeat.
Contact Erosion Is a Separate Problem From Vacuum Loss
A bottle can hold a textbook vacuum and still be finished. Every operation erodes contact material, and that erosion is measured externally. Most designs provide an erosion indicator or gauge on the drive stem or drive insulator, read with the breaker closed, where accumulated wear shifts the indicator toward a marked limit.
The mechanical signature follows. As contacts erode, the open gap grows while wipe, the over-travel remaining after the contacts touch, shrinks. That is why timing and travel analysis belongs in the same visit. ANSI/NETA ATS-2021 and MTS-2023 set the requirements, and MTS-2023 Appendix B frames base intervals that are then adjusted by equipment condition and system reliability requirement rather than applied uniformly. Our founder developed asset management and maintenance standards at PG&E before starting this company, and the recurring lesson is that duty cycle, not calendar age, should drive the interval.
Action Steps for a Vacuum Breaker Population
Inventory which breakers are vacuum and which are gas, including anything added through a recent retrofit. Mixed fleets get maintained off a single checklist more often than anyone admits.
Pull the manufacturer's stated overvoltage test value for each frame into the work package. Never carry a generic number between models.
Add X-radiation safeguards to the job hazard analysis for any vacuum integrity test.
Record erosion indicator readings and wipe measurements every visit so wear is trended, not just observed.
Where outage windows allow, add magnetron atmospheric condition measurement to build a pressure trend instead of a pass or fail history.
Trigger an out-of-cycle inspection after any significant fault interruption rather than waiting for the next calendar interval.
Budget for bottle replacement or full breaker repair and refurbishment as a planned event, because a failed vacuum interrupter is not a field repair.
Frequently Asked Questions
Can a leaking vacuum interrupter be repaired in the field?
No. There is no field procedure to restore vacuum or reseal the envelope. Once vacuum integrity is lost, the options are bottle replacement or remanufacture of the breaker. On gas equipment, by contrast, a leak path can often be located, repaired at the source, and the gas replaced.
How often should vacuum integrity be tested?
Follow ANSI/NETA MTS-2023 intervals adjusted for equipment condition and system reliability requirement rather than a fixed number from a generic checklist. Annual testing is a reasonable baseline for standard distribution vacuum breakers. Increase frequency for high operation counts, and always test after an interruption above roughly 50 percent of rated short-circuit current.
Is it safe to apply an AC hipot across open vacuum contacts?
Only within the manufacturer's stated value and with X-radiation safeguards in place. Emission becomes a real concern around 35 to 40 kV, and NETA specifications call for protective measures. If the manufacturer does not explicitly approve the procedure for that model, do not perform it.
What internal pressure means the interrupter is finished?
A healthy interrupter operates near 10⁻⁷ Torr. Published work on vacuum degradation shows dielectric strength dropping rapidly as pressure approaches 10⁻³ Torr, which is the point at which replacement is recommended to preserve reliability. The useful signal is the slope between readings, not any single value.
Does a vacuum breaker need less maintenance than an SF6 breaker?
It needs different maintenance, not less. There is no gas to weigh, report, or reclaim, which removes a real compliance burden. In exchange, mechanism, contact erosion, wipe, and vacuum integrity checks carry more diagnostic weight, because the interrupter gives no warning through a pressure gauge.
Can vacuum loss be detected while the breaker is energized?
Not reliably. Both overvoltage testing and magnetron atmospheric condition measurement require the breaker de-energized with contacts open. Energized techniques such as infrared thermography and acoustic or electromagnetic PD screening can flag related problems in the cell, but they do not measure the vacuum inside the bottle.
Get a Straight Answer on Your Breaker Population
If you are carrying vacuum breakers with unknown history, inherited retrofits, or an interval nobody has revisited since commissioning, have that conversation before the next outage window, not after a failure. Substation Solutions works on this equipment nationwide from Woodland, California and Conner, Montana.
Start with breaker repair and refurbishment if you already know a bottle or mechanism is suspect, or with high voltage consulting if the question is bigger than one breaker and you need lifecycle planning, testing strategy, or a condition-based maintenance program built around real duty data. Call (833) 723-2723 or schedule a consultation.