How to Assess the Remaining Service Life of Molybdenum Electrodes

When a molybdenum electrode should be replaced is a question that demands a serious answer in the operation of an electric melting furnace. Replace too early and an asset is wasted; replace too late and the electrode breaks in the furnace, possibly causing an unplanned shutdown, damage to refractories or even a furnace leak. The scientific approach is not to guess from experience but to estimate the remaining life from geometry, electrical and inspection data, so that replacement falls in a planned window rather than at the moment of an accident.

The object of the assessment is not the average condition of the whole electrode but the section under the most severe conditions. On the same electrode, the melt line, the threaded joint, the weld heat-affected zone and the deeply immersed section may differ several times over in loss rate. The conclusion must be anchored to the weakest part; otherwise the average hides the real risk.

1. The object of assessment is the weakest section

In a glass furnace, molybdenum loss is not uniform along the length. The zone near the melt line, in contact with glass melt, furnace atmosphere and foam layer at the same time, usually thins fastest because oxidation and scouring combine; the deep section is hot but the atmosphere is relatively stable, so loss is often slower; the wall-penetrating section is protected by water cooling and runs cooler, but poor cooling or a failed seal can also cause abnormal oxidation. Joints and welds, with their uneven structure and stress concentration, form another class of weak point.

Before assessing, therefore, draw a risk profile: mark the melt line, joints, welds, wall-penetrating section and electrode tip, then choose a measuring method for each. Measuring only the diameter of the exposed section cannot reflect the true condition of the immersed part, and measuring in one direction may miss oval wear. For critical electrodes, measure in at least two perpendicular directions and record the distance of each measuring point from the tip.

2. Three classes of basic data and the record system

Reliable life assessment rests on continuous data. It is advisable to divide the data into geometry, electrical and inspection classes, with operating variables added as explanatory items. The record need not be elaborate, but measuring points, methods, dates and duty conditions must be fixed, or successive data cannot be compared.

A note on the applicable standards: electrode products are controlled to YS/T 1312-2019, Molybdenum Electrodes, for composition and tolerances; molybdenum rods follow ASTM B387 / GB/T 17792; molybdenum plate and disc follow ASTM B386 / GB/T 3876; special requirements are implemented under the technical agreement between both parties.

The key to the record is one file per furnace and one number per electrode. Where several sizes are used in the same furnace, confusing numbers make it impossible later to tell which electrode wore fast and which was a spare. The initial diameter and measuring point positions before installation are especially important; without them, later assessment can only guess the condition from current drift and its accuracy drops sharply. Electrodes reused across campaigns should have a fresh baseline established rather than carrying over the curve from the previous campaign.

3. Diameter loss rate and remaining life calculation

Diameter loss rate is the most direct quantitative index. Take the diameter difference at the same measuring point between two dates, divide by the interval, and convert to loss per thousand hours. Ensure that the measuring point, direction and instrument are the same; if calipers were used on one occasion and ultrasound on the other, confirm the systematic difference between the two methods first.

Remaining life can be estimated by dividing the remaining permissible thinning by the recent loss rate. Take an electrode with an initial diameter of 80 mm: after 6000 hours in service the measured diameter at the same point is 72 mm, a total loss of 8 mm at an average rate of about 1.33 mm per thousand hours; if the permissible lower diameter limit is 50 mm, the remaining allowable loss is 22 mm, which at the current rate gives about 16500 hours. Allowing for duty fluctuations and measurement error, a safety factor of 1.3-1.5 puts the planned replacement window at 11000-12700 hours. This is a calculation example; actual figures should be set with the electrode size, current density, glass system and maintenance plan.

Where periodic diameter measurement is not yet available on site, a conservative estimate can be made from current drift, the appearance of the exposed section and the cooling water temperature, but such conclusions are only a lead for investigation and cannot replace measurement. For groups of electrodes connected in parallel, the current of each electrode should also be compared with the group average; a widening deviation often appears before the total power changes and is the signal to start diameter measurement.

The average rate cannot simply be extrapolated to the end of life. Loss usually follows a pattern of faster initial wear, a stable middle period and acceleration at the end: as the section shrinks, local current density rises, temperature climbs and loss speeds up again. The safer practice is to use the rate from the most recent interval rather than the whole-life average; if the rate has risen for two consecutive intervals, revise the remaining life downward and bring the inspection forward. For an electrode with an annular groove near the melt line, reassess on local thinning even if the average diameter still has margin.

4. Drift criteria from electrical parameters

In constant-power mode, a shrinking electrode section raises resistance, and the control system increases current or voltage to hold power; a current that keeps rising at constant power is therefore an indirect signal of section loss. Contact resistance, phase current imbalance and voltage fluctuation should be read together with it. A widening deviation of one electrode's current from the group average usually indicates an abnormal electrode or joint.

Thresholds must reconcile measurement accuracy with duty fluctuation. In engineering practice a grading such as this can be used for reference: a current rise of about 5% against the baseline means observe and check the joint and cooling; a rise of about 10% means schedule a diameter measurement or reduce load; severe phase imbalance, a sudden current surge or abnormal voltage means reduce power and investigate immediately. The specific numbers are a reference range and should be based on the history of this furnace and the capability of the equipment, not copied.

Electrical data have their limits too. A rising current may come from electrode loss, but also from a change in glass resistance, a drop in melt level, contact oxidation or an adjusted supply voltage. Check these variables at the same time so that a recipe change is not mistaken for electrode thinning. A better approach is to plot the current curve on the same time axis as melt level, batch lot and cooling water temperature and look for correlation, rather than reading a single curve.

5. Combining inspection and nondestructive testing

Inspection provides qualitative leads and nondestructive testing provides quantitative data; the two should be used together. Visual inspection reveals discoloration, oxidation grooves, bending and adhering glass on the exposed section; a borescope can show parts of the immersed section inside the furnace; infrared imaging reveals abnormal hot spots at joints, the wall-penetrating section and the cooling jacket; ultrasonic thickness or diameter measurement assesses the section without removing the electrode; and penetrant testing can be used where necessary to check for surface cracks.

• Visual inspection and borescope: check for grooves near the melt line, tip shape and electrode spacing.
• Infrared imaging: compare the exposed section and clamp temperatures of electrodes in the same group to find abnormal hot spots.
• Ultrasonic thickness: measure the remaining section at fixed points, taking care with probe coupling and surface condition.
• Mechanical measurement: re-measure the exposed section diameter with calipers or a dedicated fixture as a check on the ultrasonic data.
• Penetrant or ultrasonic testing: for joints, welds and suspect crack areas.
• Cooling water monitoring: flow, return temperature, pressure and signs of leakage, to rule out false overheating caused by cooling.

Each method has blind spots. Infrared sees only surface temperature, and internal defects do not necessarily create a hot spot; ultrasound is sensitive to surface roughness and coupling, and inconsistent measuring points introduce error; visual inspection cannot see the immersed section. Cross-verifying the results of several methods reduces misjudgment. For critical electrodes, keep photographs and measuring point marks from before installation and re-measure in the same positions later.

6. Condition grading and the replacement window

Once the data are assembled, they can be managed by condition grade. The purpose is not to label electrodes but to allocate limited maintenance resources to the positions of higher risk. The table below gives an adjustable reference framework; the specific percentages should follow the electrode duty, the importance of the furnace and historical data.

The replacement window also involves non-technical factors. Spare-part lead time, the shutdown plan, maintenance manpower and production orders all affect the decision; for critical positions carrying a high share of the power, the assessment interval should be shortened and the safety margin increased. If the estimated remaining life is shorter than the interval to the next planned maintenance, replace ahead of time rather than waiting to look again.

7. Safety margin, spares and the shutdown plan

Remaining life is not the same as the time available for replacement. The calculated value reflects the trend under current conditions, and changes in duty, measurement error and incidental events all alter actual life. The safety margin should cover three kinds of uncertainty: model error, duty fluctuation and supply-chain delay. For a main furnace, replacement is usually prepared one maintenance cycle ahead of the predicted exhaustion point; where a broken electrode would have serious consequences, the margin should be larger still.

The spare-parts strategy is directly linked to the assessment result. If a given electrode size has a long lead time, procurement should start when loss reaches the watch grade rather than the planned-replacement grade; for sizes shared by several furnaces, stock can be pooled to reduce inventory cost; for custom threads or welded extensions, drawings and processes should be confirmed in advance so that rush machining does not delay maintenance. Writing the spare-part lead time into the assessment table turns how many hours remain into an executable plan of when to order, when goods arrive and when to shut down.

For furnaces with an uncertain shutdown window, spares can be graded: keep one electrode of the same size in stock for critical positions and schedule non-critical ones to order, which avoids tying up inventory while leaving a buffer if the assessment suddenly deteriorates. Old electrodes removed at replacement should not be scrapped at once: check the fracture, the thinning distribution and the groove positions, and compare measured loss with the assessment curve to calibrate the prediction model for the next campaign; electrodes with remaining margin can be downgraded to non-critical positions or kept as emergency spares.

8. Common misjudgments and assessment discipline

• Using the average diameter instead of the weakest section, ignoring local thinning near the melt line.
• Not recording the initial diameter and measuring points at installation, leaving no basis for later calculation.
• Extrapolating with the whole-life average rate and ignoring end-of-life acceleration.
• Attributing every rise in current to electrode loss without checking melt level, recipe and contact resistance.
• Relying on visual inspection alone without diameter measurement or flaw detection, missing internal defects.
• Not rebuilding the baseline when electrodes are reused across campaigns, carrying over the old curve.
• Comparing life directly between furnaces with different current densities without normalization.
• Leaving assessment results out of the maintenance plan so that the data never leave the report.

Making assessment a system rather than an emergency action is the key to reducing the probability of unplanned shutdowns. Fixed measuring points, fixed methods, fixed frequency and periodic review, with geometry, electrical, inspection and operating variables on one sheet, turn remaining life from a vague judgment into an executable replacement plan. Utilization of the electrode asset improves as a result, because every replacement has a basis rather than a guess.