I.P the Centrifugal Ironmaking Process had been designed to operate continuously The three continuous experiments, conducted

under the conditions discussed in the previous section, did, in fact, represent the first test of the plant's ability to operate for long periods of time and thus represented the first real test of the plant’s design. Although some of the design aspects tested in this way had been determined prior to the start of the work covered by this thesis, it is pertinent to use this opportunity to review the operation of the plant and of its ancillary equipment under continuous conditions. This review will be carried out in three sections dealing with the ancillary

engineering equipment, the design of the reactor itself, and the functioning of the ironmaking process within the reactor.

7.6.1.(a) Ancillary Engineering Equipment

manner. Concern before the continuous experiments were started that the rotating gear would be unable to stand up to the continuous

accelerations and decelerations involved in variable speed stirring proved unfounded. Similarly, the twin mechanisms for the continuous monitoring and feeding of solid materials to the furnace had not been operated in tandem before, and the entire solids handling equipment had not been required to handle such large total quantities of material. Fears that this equipment might fail under long term operation were also shown to be unfounded. On the other hand, the continuous experi­ ments showed that the water cooling circuits to the various water cooled probes and cooling jackets could become blocked as fine scale particles built up in the circulating system. However, each probe or cooling jacket had been provided with a double filter system allowing one filter to be removed for cleaning whilst the other filter remained in operation. This arrangement allowed the filters to be cleaned at adequate intervals so that water cooling could be maintained.

The gas analysis probe and the suction gas pyrometer became rapidly blocked from the finely divided ash dust in the furnace atmosphere. Whereas it was normally found possible to clear the gas analysis probe by reversing the gas flow along it, or by 'rodding out1 in extreme

cases, the suction pyrometer became blocked so rapidly that gas temperatures could never be measured. The performance of the gas analysis probe can be seen in the results section, section 6.8, where

gaps exist in the continuous gas analysis record.

The continuous experiments, inadvertently, provided a test for the emergency shut down procedure, and for safety features that had been

incorporated into the original design. The test was occasioned when

the feed end dam failed during run 122 so that molten material started

to cast backwards into the feed end chamber. The emergency button which maintained rotation but caused all probes to be withdrawn and all feeds to be arrested and replaced by purge nitrogen or air was satisfactorily activated and the furnace brought to a standstill under manual control. The emergency hearth constructed into the feed end chamber proved

adequate to hold the molten material that drained from the furnace, and the air and nitrogen purges proved adequate to ensure that dangerous levels of carbon monoxide did not build up on the operating platform.

7-6.1.(b) The Design of the Reactor

One of the basic design features of the C.I.P. reactor was that its

refractory lining should be protected from slag attack by the layer of molten iron maintained against the lining by the centripetal pressure developed by rotation. Examination of the lining after the continuous experiments showed that this protection mechanism had worked in a highly effective manner. The only damage to the lining that could be observed was some minor spalling resulting from the emergency shut down procedure that had to be applied at the end of run 1 2 2.

The refractory dam at the casting end fared nothing like so well, unfortunately. After all experiments, the dam was so erroded that further operation of the reactor had become highly problematical, this erosion being due to slag attack since the molten iron layer could not be used to protect the dam in the same way that it protected the rest of the refractory lining. A further problem had been noted with the cast end dam, although this did not become manifest during the contin­ uous experiments. The cast end dam could not be secured by a steel

retaining ring at the outer end of the reactor, as could the feed end dam, because the operating temperature in the casting chamber was too high. Thus retention of the cast end dam against the expansion forces generated as the remaining refractory lining heated up depended entirely on the quality of the keying achieved when the cast end dam was built, or rebuilt. A new proposed design that will surmount both these

problems will be discussed in section 7-6-3*

The general design of the casting chamber did not prove satisfactory. Molten material could be cast from the reactor itself with relative ease, but a substantial proportion tended to solidify within the

casting chamber, frequently in a finely divided state against the walls and roof of the chamber. This material would then reoxidise because the atmosphere within the casting chamber was considerably more oxi­ dising than that at the end of the reactor, this reoxidation naturally reducing the yield of the process.

The continuous experiments naturally offered the first process test of the surge casting technique that had been developed in model studies

(see section 6.3*5)• In the event, the technique proved an initial

disappointment since it tended to demolish the entire carbon bed, cast slag out of the furnace in preference to iron, and to drag unreacted slag from the. feeder end of the furnace. However, the continuous experiments in themselves offered limited scope to gain operating experience of the technique and it is likely that further experience will allow improvements to be made in the control that can be exercised during surge casting.

7-6.1.(c) Performance of the C.I.P. Process

The continuous experiments indicated that the C.I.P. process was capable of steady operation and, although the supply of raw materials had limited the experimental time to a maximum of ten hours, the

process itself showed all the characteristics of being able to achieve true continuous operation. Since the experiments did not suffer from the build up of solid rings or from overoxidation, it appeared that the operating conditions chosen from the factorial experiments and the mathematical model came close to those under which the process could be operated as an efficient continuous iron producer.

Although the conditions under which the continuous experiments were carried out were chosen to produce high yields of iron, the large quantities of iron actually produced did allow some examination to be made of its quality. Carbon and sulphur levels were therefore deter­ mined in the iron that was discharged from the casting chamber during a casting surge or was obtained from the final bed collapse at the end

of one of the experiments. Experiments 118 and 119 were carried out

under more or less identical input conditions and the surge cast and

end cast iron obtained contained carbon contents varying between 0.56%