Melter Bottom Thermocouples

The following guidelines are drawn from many years observation and experience. Evaluating these comments in light of your specific applications may prove beneficial in providing optimum life and performance from melter thermocouples.

Life and stability of in-glass thermocouples, when properly designed and installed are generally excellent for all traditional soda-lime and borosilicate glass compositions. All of the soda-lime glasses (except amber), found in container and float, as well as any of the fiberglass compositions should pose little threat to the sensor. Such is not the case with leaded, opal, high arsenic and other specialty glass.

High lead glass has a tendency to become unstable during the early melting and refining process and may locally decompose or reduce in the presence of precious metals. It is not fully understood at this time if the catalyzing ability of platinum and its alloys contribute to or promote this occurrence. In either case, the lead compounds react with the precious metal causing disintegration of the sheath and ultimate attack of the assembly by the glass constituents. Life expectancy of thermocouples in leaded glass tends to vary with lead content. As an example, 3 to 6 months for 24% Pb is typical, but great variations have been noted to exist. In general, once the glass has been refined and most of the SO2, CO and CO2 have been desorbed compatibility with the precious metal is re-established. This is noted in many applications where forehearth immersion thermocouples perform adequately whereas melter and to a lesser degree refiner units regularly fail. At this juncture dual sheath assemblies with or without precious metal liners, in dense dead end blocks is the only practical recommendation that can be made for monitoring or controlling melter temperature.

Precious metals are readily attached by the Halogens. Opal, and other glasses relying on fluorine compounds as a colorant will destroy immersion assemblies in a short time period. Even low level fluorine bearing glasses have a tendency towards metal attack. The life of thermocouples must be determined for each application.

Flatware and other products relying on arsenic or lithium additives as brightening agents have proven problematical on occasion. While concentrations of the additives are low, upsets in redox potential due to combustion control excursions have been noted to cause local reduction of the element and subsequent destruction of the precious metal.

Amber glass has been well documented in recent years for the cause of its incompatibility with precious metals. The sulfur within the coloring agents combine with the rhodium forming a glassy phase compound which dissolves readily in the soda-lime matrix. The use of ODS Platinum completely eliminates this occurrence and should be exclusively used in all amber tank applications.

The performance of platinum alloys in electric melters has been extensively investigated over the past twenty years. The dissolution rate of platinum sheaths and thimbles is known to be proportional to the exposed metal surface area; and the leakage current through the metal to tank ground, and inversely proportional to the frequency of the power mains. Recorded observations compared with theoretical calculations have been in excellent agreement. Leakage currents of 10 amperes has been documented to physically destroy a 2-inch immersion sheath in 120 days.

The prevention and cure of electrolysis failure is relatively simple. Using the following notes as a guide an instrument specialist can easily guard against common problems.

A few recommendations can be broadly stated for thermocouple installation, most of which are based on observations and tests conducted in the field.

• Thermocouples must be completely isolated from electrical contact with any tank metal work.
• Support of the sensor should be by use of flanges resting on non-conductive sheets.
• If shielded compensating lead wire is used, the shield should be wrapped with glass tape for that portion which is inside the terminal head. Many inadvertent groundings have been recorded because of frayed shield wires coming in contact with the inside of the head.
• Grounding of shield wires, if required, should be at one place at the end inside the main instrument panel. Before grounding, an isolation test, using a 50 vdc Megga should be conducted between the shield, conductors, and ground to insure no secondary ground loops exist.
• Compensating lead wire should not be run through the power line channels. Induced voltages with sufficient power to cause instrument problems have been known to exist.
• Ideally, compensating lead wire should be run through solid conduit. If a flexible connection is made to the terminal head it should incorporate an electrical isolator.

Traditional glass immersion sheaths have been produced with 10%, or 20% alloys of rhodium in platinum. Experience has shown that ODS platinum can be a direct substitute for these alloys and has been successfully used to replace rhodium and subsequent inherent cost. In addition, sheaths being free of rhodium are not subject to attack by amber glass. The use of ODS platinum permits direct immersion control to be obtained in this otherwise difficult to instrumentate process.

All BASF produced thermocouples for glass melters are fabricated with high purity alumina insulation. The alumina content is at a minimum of 99.7%. While it is commonly known that high purity alumina is needed for stable performance of precious metal thermocouples used at high temperatures, it must be realized that type and nature of the remaining impurities are equally important. BASF choses its high purity alumina with great emphasis on eliminating known deleterious contaminants such as Fe, Ni, Cr, Sb, among others. Equally as important in the selection process are the concentrations of certain glass formers and combinations of them such as B2O3, SiO2, Na2O, etc.

The wire used in all thermocouples is matched and calibrated to comply with ISA, ANSI, or IEC reference tables to within +/- ¼% of temperature from 400°C to 1450°C. All thermocouples supplied on a single order are fabricated from a single batch of wire insuring that all thermocouples provide equivalent data. Lot calibration, as well as specific unit calibrations data can be supplied at an additional cost.

Some thermocouple designs use Fibro Platinum with types R and S. Fibro is a proprietary wire making process which imparts an elongated grain structure to the metal. This structure in addition to strengthening the wire imposes barrier to the migration of contamination and subsequent weakening of the material. This material provides additional stability of calibration and increases the thermocouples effective life.

Precious metal thermocouples always have value - even when they are no longer usable. The metal content of bottom thermocouples can be recycled into new replacements; or monitary credit be given for their intrinsic value; or provide the user with book credit for use at a later date. All reclamations are made on a weight basis. For the most accurate and beneficial credit, spent thermocouples should be returned intact for disassembly at the plant so that proper material and alloy separation can be performed. Contact the customer service office for specific details on how to return spent thermocouples.