How to Select the Working Current Density for Molybdenum Electrodes

What decides how long a molybdenum electrode lasts is often not the material grade but how many amperes each square centimeter of its section carries. Current density ties output, life, capital cost and power supply conditions together, and it is a parameter that has to be recalibrated repeatedly in the operating regime of an electric melting furnace. Set it too high and erosion and creep accelerate; set it too low and power falls short while the number of electrodes and the cost rise. Real selection means finding the balance between the two in the conditions of your own furnace.

Before discussing current density, settle the basis of calculation. In engineering practice it is usually calculated on the nominal cross-section of the electrode, that is, the current per electrode divided by one quarter of pi times the diameter squared; some plants instead convert on the surface area of the immersed section. The two methods give noticeably different numbers, so when quoting another plant's data or comparing with a supplier, confirm that the basis is the same; otherwise an apparently identical 2 A/cm² may not mean the same thing.

1. Why current density is the hub of the operating regime

As current passes through a molybdenum electrode, the Joule heat generated per unit length is proportional to the square of the current density. Part of that heat maintains the melt temperature and part is lost through the electrode, the cooling water and the furnace wall. As current density rises, the temperature of the electrode tip and of the gas-melt interface section climbs, and chemical erosion of molybdenum by the glass melt, volatilization of molybdenum oxides and high-temperature creep all accelerate. Current density is therefore not an independent electrical parameter: it simultaneously regulates the thermal field, the flow field and material loss.

In a glass melt, molybdenum loss comes broadly from several paths: molybdenum is oxidized and enters the glass as oxide; glass components dissolve and react with molybdenum at high temperature; oxidation and scouring where gas and melt alternate near the melt line; and mechanical damage from thermal shock and creep. A higher current density intensifies the first three paths at once and indirectly amplifies the fourth through higher temperature. The electrode thinning seen in service is therefore usually the combined result of several mechanisms, and current density is the lever that can be adjusted deliberately.

2. Empirical range and zone corrections

Long experience in glass electric melting has produced a working range of roughly 1-3 A/cm². This range is not a universal quota but a starting point that must be corrected by zone according to electrode position, glass properties, cooling conditions and operating strategy. The table below gives reference ranges for common duties; these are typical values, and actual figures should be confirmed by thermal calculation, operating data from similar furnaces and the contract.

Basis for verifying and judging the above figures: electrode products are controlled to YS/T 1312-2019, Molybdenum Electrodes; molybdenum rods follow ASTM B387 / GB/T 17792; molybdenum plate and disc follow ASTM B386 / GB/T 3876; a quality certificate accompanies each shipment and batches are traceable.

The essence of zone correction is to identify which section is the weak point of the whole electrode. On the same electrode the deeply immersed section may wear very slowly, while the section near the melt line, in contact with glass melt, furnace atmosphere and foam layer at the same time, usually thins first. Calculating current density on the average section alone overestimates the safety margin. A safer method is to divide the electrode along its length into segments, estimate temperature, atmosphere and current distribution for each, and take the most severe segment as the controlling section.

3. Design calculation from power to diameter

Forward calculation in selection usually starts from power. First determine the total electrical power from the furnace heat balance, output and the electric boosting ratio, then derive the current per electrode from the supply voltage, the connection method and the number of electrodes. For a single-phase circuit, power divided by voltage and then by the number of electrodes gives an estimate; for a three-phase circuit, phase voltage, number of phases and load balance must also be considered. Use the actual operating voltage, not the no-load voltage on the equipment nameplate, because short-network voltage drop and power factor change the result noticeably.

Once the current per electrode is known, back-calculate the required cross-section from the target current density and convert it to a diameter. Take a current of 8000 A per electrode as an example: at a target of 2.0 A/cm² the required section is about 4000 mm², corresponding to a diameter of about 71 mm, so a 75 mm size can be selected, giving an actual current density of about 1.8 A/cm²; at a target of 2.5 A/cm² the required section is about 3200 mm², corresponding to a diameter of about 64 mm, so a 65 mm size can be selected, giving an actual current density of about 2.4 A/cm². These are calculation examples; actual selection must also add corrosion allowance and respect the standard size series and installation space limits.

Corrosion allowance is the step most easily omitted. An electrode thins continuously over a campaign, and if design only satisfies the current density at the initial diameter, the actual current density will rise later in the campaign. The safer practice is to check against the permissible end-of-campaign diameter: first estimate the diameter loss over the campaign, then calculate the current density at that end diameter and make sure it still falls in an acceptable range. For a furnace planned to run three to five years, this check often moves the initial size up one step, but the extra life gained usually outweighs the additional material.

4. Matching the power supply and electrode layout

The power supply method determines how current is shared among the electrodes. Line-frequency supply is simple and has relatively fixed short-network losses, but the current distribution is strongly affected by electrode spacing and the resistance of the glass melt; medium- or high-frequency supply can raise power density and reduce equipment size, yet places higher demands on short-network layout, insulation and filtering. Whichever method is used, verify with the actual current waveform and phase current balance rather than dividing theoretical power by the number of electrodes.

Electrode layout also changes local current density. In a single-phase arrangement the current concentrates in the melt channel between two electrodes, and the current density in that channel may exceed the average calculated on the total section; in a three-phase arrangement the current is unevenly distributed between and within phases, and edge electrodes often carry a different load from center electrodes. Design should draw on electric-field and flow-field simulation, or on measured phase currents from similar furnaces, so that some electrodes are not permanently overloaded while others run permanently underloaded.

Short-network impedance and contact resistance are also variables that cannot be ignored. Oxidized clamps, uneven contact surfaces or insufficient clamping force all raise contact resistance, and the resulting local heating accelerates oxidation in a vicious circle. If the clamp temperature of one electrode is clearly higher than that of others in the same group, check the contact condition first before discussing whether the current density is appropriate. Mistaking an electrical problem for an electrode material problem is a common diagnostic error on site.

5. Matching the glass recipe and furnace duty

The redox state of the glass melt directly affects molybdenum loss. In systems refined with sulfate, in glasses with higher iron content and in strongly reducing batches, the reaction paths at the electrode surface differ from those in oxidizing glass, and the safe boundary of current density shifts accordingly. High-viscosity melts have weaker convection, so local overheating is not easily removed and values should tend toward the lower-middle part of the range; low-viscosity, strongly convecting melts dissipate heat better but erode faster, and should likewise not be pushed upward blindly.

Melt level is an easily overlooked operating variable. A falling level exposes what was an immersed electrode section to hot furnace gas, raising both oxidation and thermal shock risk, while the control system often continues at the original power, so the actual current density rises passively. A reasonably safe interlock logic is: when the level falls below the set value, reduce power first, then check the exposed length of the electrodes; after the level is restored, load up in steps. Similar logic applies to batch changes, stopping of bubbling and reheating after maintenance.

The output strategy of the furnace must also enter the selection. A main furnace running continuously for many years is better served by a low-to-medium current density in exchange for long life and stable operation; a furnace running short campaigns or seasonally may push slightly higher within its life budget, but diameter measurement and temperature monitoring must then be tightened. Using one current density for every furnace looks simple to manage, but in practice it transfers the differences between duties onto electrode life.

6. The cost at both ends: too high and too low

Choosing a current density is essentially trading one cost against another. The table below sets the typical consequences of the two ends side by side for use in design reviews. The entries are qualitative trends; the magnitude varies with glass system, cooling conditions and electrode size.

It should be noted that low does not mean safe. Long-term underloading leaves the melt temperature unevenly distributed, and local areas compensate by raising voltage, which creates overheating elsewhere; at low electrode surface temperature, more glass adheres, cleaning frequency rises and the risk of mechanical damage increases. The reasonable range is a band, not a point, and operation should fine-tune within it according to monitoring data.

7. In-service monitoring and adjustment

Whether the current density is right has to be verified by operating data. At least three kinds of records are recommended: periodic measurement of electrode diameter, with measuring points specifically at the melt line and in the deep section, to calculate diameter loss per thousand hours; current drift in constant-power mode, since a current that keeps rising at unchanged power usually means a shrinking section or a change in spacing; and the temperature of clamps, wall-penetrating sections and water-cooling jackets, collected periodically by infrared or embedded thermocouples.

1. Record the baseline at start-up: number, initial diameter, insertion depth, initial current and voltage of each electrode.
2. Observe closely for the first two weeks: log current, power, melt level and cooling water temperature daily and confirm there are no abnormal hot spots.
3. Once operation is stable, review monthly: calculate the diameter loss rate and compare it with the design estimate.
4. When the deviation exceeds a preset threshold, check the batch recipe, melt level, contact resistance and cooling condition first, then decide whether to lower the current density.
5. After each adjustment, keep at least one week of comparative data so that duty fluctuations are not mistaken for a trend.

The order of adjustment also deserves emphasis. When loss appears fast, do not rush to lower the current density; first confirm whether the glass recipe, melt level and cooling have changed, and only then consider derating. Derating directly affects output, so it is usually done in steps with a week of observation rather than as a single large change. For electrodes already in the middle or late stage of their life, derating can only slow the loss and cannot reverse section already lost, so the replacement window should be assessed at the same time.

8. Common misoperations and commissioning acceptance

• Copying the current density of a furnace with the same tonnage while ignoring differences in glass system, electrode depth and cooling.
• Dividing total power by the number of electrodes without considering voltage taps, short-network drop and three-phase imbalance.
• Using the rated current of the equipment instead of the actual operating current, which makes the calculated value too high.
• Running at full load to the original setpoint after a drop in melt level or a batch change without recalculating.
• Pushing the current density higher while cooling water flow is insufficient or clamps are oxidized, turning an electrical problem into electrode overheating.
• Running at the upper limit of the range for long periods and mistaking can deliver the power for suitable for long-term operation.
• Comparing life between furnaces without unifying the basis on which current density is calculated.

At the commissioning acceptance stage, proceed in the order of low-power preheating, stepped loading, baseline recording, troubleshooting and alarm setup. First confirm at lower power that the electrodes heat evenly with no local hot spots, then raise step by step to the design value; allow each step to stabilize before moving to the next, so that thermal shock and current surge do not coincide. After acceptance, write the design current density, measured values, alarm thresholds and adjustment authority into the operating procedure, so that operators adjust within clear boundaries rather than by feel.

There is no once-and-for-all answer to current density. It changes with furnace age, glass recipe, electrode wear and power supply conditions, and design calculation, in-service monitoring and periodic review must be closed into a loop. Treating this parameter as a manageable process lever rather than a fixed number on the nameplate usually improves electrode life and furnace stability at the same time.