Report on ceramic tube al melter available.

A report describing work performed on a project titled "Gas-Fired Immersed Ceramic Tube Aluminum Melter" has been published. The project was funded by the Gas Research institute, with Babcock & Wilcox as the primary contractor and Reynolds Metals Co a major subcontractor

Work was focused on the accomplishment of three goals:

* to test commercially available natural gas burners in ceramic tube/burner assemblies tor immersed tube melters;

* to verify the performance, life and ease of maintenance of commercially available ceramic tubes used in the burner/tube assemblies;

* to design, build, operate and evaluate the performance of a prototype immersed tube aluminum melter

Operational goals were a 1000 lb/hr melt rate, 1200 Btu/lb energy consumption, a 2% increase in metal recovery during the melting of aluminum scrap and acceptable maintenance requirements.

A commercially available single ended recuperative gas burner was identified which could operate as an immersed tube burner at the desired heat output. However, at that heat output, which was about 300% of the design output, burner efficiency was about 10% lower than nominal.

Burner operation proved to be very reliable but adjustment for optimum performance was difficult because of the design of the gas train. This difficulty was determined to be surmountable by improving the ability to adjust each burner individually.

Three of four ceramic tubes used in long-term life tests performed for between 61/2 and 1 1 months without failure. This exceeded the time required for economic viability. The fourth tube failed after 35 days for an undetermined reason. Because of operator error all six tubes in the prototype melter failed by thermal shock after two months of operation. A modification to the melter controls was made to prevent a recurrence of the thermal shock problem.

The prototype melter was operated in a production environment for six months, during which about 410,000 lb of used beverage can (UBC) scrap was processed. The melter exceeded standard metal recovery by 0.2% during a one month period of operation. NO[.sub.x] emissions from the combustion system were shown to be about 60% less than from state of the art reverberatory furnaces. The metal produced contained about 35% less dissolved hydrogen than that produced in reverberatory furnaces melting UBC scrap.

Because of several problems unrelated to the immersed tube melting concept, the overall performance goals for the melter were not met. Overall metal recovery was lower and energy consumption higher due to substandard performance of the charging system, and poor scrap quality. The scrap contained only about 88% metal instead of the typical 94-95%. Particle size was also smaller than typical, which increased the tendency for oxidation.

Burner efficiency was somewhat lower (60% instead of 70%) because of operation at 300% of nominal output. Dross formation was a major hindrance to optimum performance.

In spite of the operational problems, economic analysis indicated that an immersed tube melter could be economically competitive with reverberatory furnaces. Assuming a tube cost of $1000 for a 60 in. long tube and a usable tube life of six months, metal recovery would have to exceed standard recovery by only 0.1 %. However, burner efficiency at high output would have to be improved to the levels obtained at design output.

Based upon operation of the prototype, the concept for a full scale 10,000 lb/hr melter was designed. The factors hindering commercialization of the full scale melter were identified and recommendations for improvement were made. The technical issues appeared to be resolvable with further work. Many were thought to be a function of melter size and therefore would not be important in a full size furnace.

A copy of the full report is available from the National Technical Information Service, 5285 Port Royal Rd, Springfield, VA 22161. For specific information, call Steve Sikirica or Chris Stala at the Gas Research Institute, 312/399-8200.