Collecting and Comparing Electrode Life Data

When the same batch of electrodes lasts six months in one furnace and only three in another, the comparison is often meaningless. Furnace type, current, glass composition and start-stop practice differ, and the life figures lose their meaning when set side by side. To make the records usable, unify the definition first and then control the comparison conditions.

Define End of Life First

End of life is not only when an electrode breaks. Common criteria include: diameter worn to the limit, surface cracks grown to the point of affecting conduction, contact resistance at the joint rising, end deformation preventing clamping. A plant should select a set of conditions and specify who confirms them and how the date is recorded.

With inconsistent criteria, the same electrode may be recorded with different lives, and later analysis loses its basis.

Which Fields the Operating Record Needs

• Electrode number, batch, installation position and installation date.
• Operating current, voltage and power, recorded by shift or by day.
• Furnace temperature, glass level and atmosphere condition.
• Output: tonnes of glass or tonnes of fibre, matched to the electrode operating period.
• Reason and duration of stoppage, and a description of the appearance at change-out.

The record need not cover everything, but it must support later conversion. For example, with days in service but no current or output, it is impossible to tell whether a difference in consumption rate comes from load or from material.

Handwritten records miss items easily, so the key fields can be made into a form attached to the shift handover book for the operator to tick and complete; where the plant has the systems, current and temperature curves can be exported from the control system, filed daily and aligned with the output report.

Variables to Control in Cross-Furnace Comparison

When records from different furnaces are put together, at least the following conditions must be checked: glass composition, especially alkali content; furnace temperature regime; electrode current density; protective atmosphere; cooling method; start-stop frequency. Records with large differences in these conditions serve only as a trend reference and should not be ranked directly.

A workable approach is to pair furnaces with similar conditions, or to convert life into consumption per unit of output and then correct for load. The correction method should be agreed in advance so that a favourable basis is not chosen afterwards.

Another common bias is an inconsistent statistical period. Some summaries are by calendar month, others by campaign, and a cross-year comparison mixes in the load differences of different seasons. State the statistical interval and the record source in the table header to avoid misreading.

Summary Analysis and Review

Once enough records have accumulated, group them by batch, position and season and look for a pattern of concentrated failure. If electrodes at one position repeatedly fail early, check installation coaxiality, cooling conditions and local overheating first, rather than simply attributing it to the material.

The review meeting should involve production, maintenance and purchasing together, reach conclusions and feed them into the selection or acceptance terms for the next batch.

The value of life records is not in producing a ranking but in making every change-out accountable. Fix the criteria, the record fields and the comparison conditions, and the records can play a part in material selection.