Further Studies of the Thermal Performance of Glass Melting Tank Furnaces

JSGT 1946 V30 T243-T277

Data of the weekly glass production and weekly fuel consumption for a total of eight melting campaigns involving four different furnaces are studied in detail. The furnaces are: (a) An oil-fired furnace the melting area of which was 83 sq. ft. fur two campaigns and 93 sq. ft. for a third campaign; (b) a producer gas-fired furnace of 580 and 560 sq. ft. melting , area in two separate campaigns; (c) a producer gas-fired furnace of 590 sq. ft. melting area; (d) a furnace of 171 sq. ft. melting area, fired with coke oven gas for two campaigns. The importance is emphasised of plotting some measure of furnace performance against statistically determined limits, in order that changes in performance shall be detected at the earliest moment. In Section II the variations of the thermal efficiency index Q are discussed in some detail, with a view to providing evidence of the technical value of calculating this index weekly and of plotting it on a “control chart.” A formal analysis of regressions connecting the weekly fuel consumption with the age of the furnace and with the weekly glass production is given in Section III. Some statistical methods not used in an earlier paper are illustrated. A general discussion of the data in Section IV provides evidence that some of the variability in furnace efficiency arises from the fact that changes in glass output from the furnace and fuel changes are not synchronised.

For all practical purposes the calculated value of Q is as good a measure by which to estimate the thermal performance of a furnace as the more involved and formal analysis dealt with in Section III. It is also shown that the correction for furnace area recommended by the Furnace Committee is a good practical approximation. In spite of the wide divergence of size of the furnaces studied and of the fuel used, it is possible to make a reasonable estimate of the magnitude of the proportion of additional fuel required each week to maintain the furnace at the required melting temperatures and of the amount of fuel which is associated with the production of one pound of glass under the present conditions of operation. A table is given of values for the construction of provisional control charts. These charts would serve as a guide to reasonably good practice until data are available for any particular furnace and method of operation, in order to assist those wishing to introduce the control chart technique for watching furnace performance.

B. P. Dudding