A decomposition product result is the gas-side record of what a circuit breaker has been asked to do. Every arc in sulphur hexafluoride dissociates the gas, and the fragments recombine or react to form compounds whose concentration reflects the energy the gas has absorbed. That makes the measurement an event-driven one rather than a routine one, and it makes it interpretable only against the arcing history of the specific breaker.

It also makes the measurement a different question from the other two gas determinations. Dew point detects moisture. Purity detects air ingress. Decomposition products detect the arcing duty the gas has experienced. Reading the three together is what establishes whether a compartment is in a normal state for its service history.

When Decomposition Testing Is Warranted

The primary trigger is an abnormal operation. A breaker that has interrupted a high fault current, one that has operated following a nearby lightning strike, or one whose operation produced an alarm has performed work the gas has to absorb, and the measurement establishes how much.

The second trigger is an inconsistency. A compartment whose gas density has fallen faster than the leak history predicts, or one whose operating count has advanced in a way the maintenance records do not explain, justifies a measurement even without a specific event. The result resolves whether the difference is explained by arcing or by a separate problem.

The third trigger is the final measurement before a decision. Where a breaker is being prepared for an extended return to service, or where a compartment is being opened for any reason, the measurement taken during the work establishes the condition the gas was in and provides the reference for the next measurement. Taking that reference is worthwhile even when nothing suggests a problem, because it is the only way the next result becomes interpretable. A fourth case is a periodic programme on critical breakers, where the interval is set by the expected arcing duty rather than by the calendar.

The urgency differs by trigger. An abnormal operation justifies a measurement within days; a periodic check follows the programme interval. The re-use and handling framework that governs what to do with the gas afterwards is published as IEC 60480, with the gas handling requirements in IEC 62271-4.

Fault Types and the Compounds They Produce

An arc raises the gas to a temperature at which the molecule dissociates into sulphur and fluorine. As the gas cools, most of it recombines into the original molecule, but a fraction reacts with the decomposition products of earlier arcs, with oxygen and moisture present in the compartment, and with the electrode and enclosure metals.

The three mechanisms leave different signatures. High-energy arcing produces the largest quantity of reactive compounds and does so quickly, in proportion to the arc energy. Partial discharge, which involves far less energy, produces a smaller and different pattern over a longer period. Overheating at a contact or connection decomposes the gas more slowly still, and the compounds it produces reflect the lower temperature and the presence of metal.

What matters for interpretation is the relationship between the compounds rather than any single concentration. A result dominated by the products of high-energy arcing points at a fault interruption the records may or may not explain. A pattern consistent with low-energy activity over time points at something slower, which is why the ratio between the reactive products is more informative than the total.

The analytical method used should be named in the record, because results from different methods are not directly comparable. Cite the method applied rather than describing it generically, and where a laboratory is used, record its identification and the batch or run reference.

Sampling from the Gas Compartment

The sample has to be drawn from the gas space rather than from the bottom of the compartment, because reactive compounds and condensable material distribute differently through the volume. Sampling from a point where the gas flows freely during operation gives a more representative result than sampling from a dead end.

The sampling arrangement affects the result more than it does for an inert gas. Residual gas in the line, a connector that has been left open, or a container that has not been purged introduces air and moisture that react with the compounds being measured. Purging the sampling line before the measurement, and using equipment in good condition, is what makes the result describe the gas rather than the sampling.

Timing matters for the same reason. The reactive products do not accumulate indefinitely; they react further or are absorbed on internal surfaces. The interval between drawing the sample and completing the measurement should be as short as the equipment allows, and where a direct-reading instrument is used on site the exposure is shorter and the handling simpler.

Handling practice should follow the gas procedure the site operates under, which the handling requirements in IEC 62271-4 frame. Instruments for the gas-side measurements are grouped on the SF6 testing hub.

Interpreting Results Against Breaker History

A single result is meaningless without the arcing history. A level that would be unremarkable on a breaker that has interrupted several faults is a finding on a breaker that has done nothing but load switching for a decade. The history needed is the number of fault interruptions and the magnitude of the current interrupted, which the protection event records provide.

Where the accumulated interrupted current is available, it is the best basis for the comparison, because arc energy scales with it. Where it is not, the number and magnitude of interruptions form an adequate proxy. Either way, the measurement and the history are read together rather than sequentially.

The second distinction is step change against slow drift. A result that has risen sharply since the last measurement indicates an event. A result that has crept upward across several measurements indicates a continuing process or the residue of a series of events, and the two call for different responses.

A result higher than the recorded duty explains is the case that warrants investigation, because the gas has recorded arcing the records do not. That is a reason to examine the protection records and the breaker’s operation history before returning it to service, rather than to schedule a gas change and move on.

Combining Decomposition Data with Dew Point and Purity

The three determinations describe different contamination mechanisms, and the handling decision depends on all three. Dew point establishes whether moisture is present, purity establishes whether air has entered, and decomposition establishes what the arcing has produced.

A compartment with all three within limits returns to service without treatment. One with a decomposition increase but clean moisture and purity has absorbed arcing in a sealed, dry system, which is the expected outcome of fault interruption. One with a moisture increase alongside a decomposition increase has a chemistry problem as well as a containment problem, and the moisture path is the finding to pursue.

Moisture changes the decomposition pathway, which is why an elevated compound concentration in the presence of moisture points at degradation rather than at a simple arcing event. That interaction is the technical reason the three measurements are taken together rather than in sequence, and it is the reason a decomposition result interpreted alone can mislead.

Where all three have moved, the compartment has both a leak and an arcing history the records may not explain, and the combination justifies an internal inspection when access allows. The re-use specification that governs what the gas has to meet before returning is IEC 60480.

Actions: Reclaim, Service or Replace

The gas action and the breaker action are separate decisions. For the gas, the options are to return it to the compartment, to process it to remove moisture and reactive compounds before returning it, or to send it for reclamation. Which applies depends on the measured values against the limits for the equipment.

For the breaker, the response depends on what the gas and the other measurements jointly indicate. A decomposition result consistent with the interruption history requires no breaker action beyond the scheduled maintenance. A result that exceeds what the history explains warrants an internal inspection of the arcing contacts and the nozzle, because the gas has recorded arcing the records do not.

Where the compartment needs opening, the work should be scoped at the same time as the gas handling so that the gas is recovered, processed and returned in one sequence rather than in several visits. Where the result is marginal, the appropriate action is to shorten the interval and repeat the measurement rather than to intervene on a single reading.

The decision and its basis belong in the breaker file. A decision recorded with the values that drove it can be reviewed at the next measurement; one recorded as an action without its evidence cannot.

Records Needed for the Breaker File

The record should identify the compartment and the breaker, the date and the reason for the measurement, the analytical method used, and the results alongside the dew point and purity taken at the same time. It should also carry the interruption history over the period since the previous measurement, because that is what makes the result interpretable.

Added to those, the action taken and the values that drove it, including the processing outcome where gas was treated. Where gas was processed, the results before and after indicate the efficiency of the process and the state of the compartment.

Over a sequence of measurements the record becomes the breaker’s gas history, and it is what allows a future result to be interpreted without reconstructing the arcing duty from protection logs. Field practice with gas condition assessment is coordinated through CIGRE study committees, and the instruments for the measurements are grouped on the SF6 testing hub.

Testing Versus Monitoring

Testing establishes a value at a moment. Monitoring establishes whether a value is changing, and for gas condition the distinction matters because arcing is episodic. A breaker may perform no interruptions for two years and then several within a week, and only continuous or frequently repeated measurement captures that pattern.

Continuous monitoring of decomposition products is less common than density monitoring, because the reactive compounds require more careful handling than a density measurement. For a critical breaker the practical compromise is density monitoring with frequent automated trending, plus a decomposition measurement taken after any significant operation.

What monitoring cannot replace is the interpretation. A monitored value without the operating count for the same period says little, and a monitoring system whose alerts nobody reviews produces a record rather than a control.

The decision between the two approaches follows the consequence of failure and how much the operator needs to know before the next planned outage. A breaker whose failure is contained and whose replacement is straightforward can be managed with event-triggered testing; one whose failure would remove a critical transfer capability justifies the instrumentation that makes a continuous record possible.