Viability of formcoke explored as a cupola fuel option.

This experiment shows formcoke to perform on par with conventional coke, giving cupola melters another fuel to consider as questions arise about future domestic coke supply.

The vast majority of coke used as the fuel for cupola furnaces is produced in slot ovens. However, due to some concerns about conventional coke's cost and supply in the next century, alternative fuels such as formcokes [produced by forming and subsequently heating a mixture of carbon (C) and binder] may offer advantages for cupola foundries.

Formcoke potentially can offer benefits of performance, cost and environmental impact derived from the uniform size and low porosity of the briquettes, the use of low-cost carbon sources, short heating cycles, and continuous and more uniform manufacturing conditions.

Numerous variants of formcoke exist, depending on the type of C and binder used. Few formcokes are in commercial production, however. Most are too small to be used effectively as cupola fuel. Extensive studies carried out 20-30 years ago of a large formcoke developed for cupolas indicated that it was more susceptible to degradation by impact and abrasion than conventional coke. A 1995 study of the properties of another commercially produced formcoke for cupolas by B.C. Young and MA. Musich also indicated that the properties could not match slot oven coke.

An unpublished portion of their study, however, did find properties equivalent to slot oven coke for a U.S.produced formcoke. A description of the process and the properties of the U.S. formcoke were also published in 1995, and the formcoke that is the subject of this article was developed by Coal Technology Corporation (CTC), Bristol, Virginia, which was recently acquired by Antaeus Energy, Charleston, West Virginia, the U.S. subsidiary of Greenfields Energy Corp., forming a new division called Antaeus CTC.

Several production tests were conducted in two large foundries with formcoke replacing 10-50% of the conventional slot oven foundry coke charged (see sidebar). These tests indicated that the formcoke performed on par with the conventional slot oven coke. Because of the difficulty in accurately assessing relative performance in the production environments, further evaluation was made in a 72-in. cupola dedicated to carrying out controlled experiments. Two cupola heats served to compare formcoke as the only fuel in the cupola against a baseline condition using only conventional slot oven foundry coke for the acid melting of gray iron.

Manufacturing Comparison

Conventional slot oven foundry coke is typically manufactured by blending 36 different coals and gravity-charging the blend into a coke oven battery through ports located on the top. The charged coal blend is leveled in the oven and then indirectly heated to 2000F (1093C) for 28-36 hr. At the end of the coking cycle, the coke is pushed from the oven into an open car and quenched with water. After quenching, the coke is sized and ready for shipment.

The new formcoke process first produces char by mildly gasifying one or more coals in a twin screw reactor indirectly heated to 1200-1400F (649-760C). The char is combined with coke breeze, coal and bitumen binders and formed into green briquettes. The briquettes are coked through a tunnel kiln at 1800-2000F (982-1093C) in an inert atmosphere. At the end of the coking cycle, the formcoke is dry-quenched in a cooling apparatus and ready for shipment. The continuously operated formcoke process is completed in about 4.5 hr in a totally enclosed system.

Property Comparison

The proximate and sulfur analyses of the formcoke and slot oven coke are presented in Table 1. There was no significant difference in the composition of the ash from each coke type. The test results indicated that, on average, the formcoke had the following absolute composition differences compared to the slot oven coke: 1.3% higher moisture, 1.6% higher volatile content, 2.6% lower fixed C, and 0.14% higher sulfur (S). The slot oven coke properties showed less variation than the formcoke.

The formcoke manufactured for this test was made in a 4 x 5 x 6 in. pillow-shaped die and on average measured 5.7 x 4.6 x 3.5 in. The mean slot oven coke size was 5.1 x 4.0 x 3.3 in. The formcoke used had a much lower variation about the mean size.

The unpublished portion of Young's study reported that the domestic formcoke had an apparent density 28% higher than slot oven foundry coke. In this experiment, formcoke had bulk [TABULAR DATA FOR TABLE 1 OMITTED] density measurements 54-100% greater than that of slot oven coke, depending on the size of the container used to measure the property.

Research Cupola Description

The research cupola is a water-cooled, refractory-lined, externally fired hotblast cupola with a divided blast tuyere arrangement and a straight shell. The cupola shell [D is 72 in., and it was lined to a working diameter of 61 in. at tuyere level. The three upper and three lower tuyeres are separated by 15 in., and the upper tuyeres and charge door are separated by 21 ft. The cupola is tapped continuously with a front slagging arrangement. The off-gas take-off point is above the charge door. Oxygen ([O.sub.2]) enrichment of the blast was carried out by direct injection through the tuyeres.

The Controlled Experiment

Two cupola heats were conducted using similar operating conditions such that the results from using formcoke as the only fuel in the cupola could be evaluated against a baseline condition using only slot oven furnace coke. In both heats, data was collected from the time the first metal flowed over the iron dam until the last charge entered the cupola stack. Charge materials were sampled throughout the heats, and every charge weight and time was recorded. An iron composition sample was taken every 10 rain and a slag composition sample every 20 min. Other cupola operating variables were measured and recorded every 10 min throughout the heat. These included: iron temperature, blast air rate, [O.sub.2] enrichment flow rate, blast temperature exiting the blast heater, blast temperature in tuyere number two, offgas composition (% CO, % C[O.sub.2], and % S[O.sub.2]), off-gas temperature, dew point and temperature of the air in the cupola blower inlet duct, and back pressure in the cupola well and wind dram.

The target coke charge weights were adjusted so that the ratio of charged metal to charged coke fixed C would be the same in both heats. This was necessary because the mean fixed C content of the two cokes were different. Each heat was conducted under the same refractory, bottom, coke bed and iron dam conditions.

The targeted cupola operating conditions for the two heats were: 2% [O.sub.2] enrichment by injection through the tuyeres, slot oven coke in the baseline heat charges equal to 10.5% of the metal weight and an equivalent amount of formcoke in the second heat on the basis of fixed C, a 2530 scfm actual blast air rate, and a charge mix consisting of 68.5% scrap gray iron, 29.4% fragmentized steel and 2.1% lump ferrosilicon (50% Si).

Controlled Experiment Results

Time segments representing steady state conditions for each heat were chosen by studying plots of all measured variables against time. Means and standard deviations were then calculated for the data from these time segments. The conditions were verified as steady state by demonstrating that the ratios of C input and output (C balance) for each heat were essentially the same. Within the experimenters' ability to control them, the input variables were essentially identical in both heats (Table 2), so a comparison of the mean output variable values (Table 3) for these time segments was considered suitable for use in assessing the relative performance of formcoke and slot oven coke.

Despite the much higher bulk density of the formcoke, the back pressure measured inside the cupola well during the formcoke heat exceeded the pressure in the slot oven coke heat by only 0.8 in. of water on average. There were no problems with slag control during either heat. On average, the two cokes produced identical C [ILLUSTRATION FOR FIGURE 1 OMITTED] and S pick-ups in the iron. Higher Si and manganese (Mn) recoveries were observed in the formcoke heat.

The higher Si and Mn losses observed in the slot oven coke heat may have been due to higher iron oxide generation rates, which are known to be responsible for higher alloy losses. Oxidation of a skull that formed during the slot oven coke heat, variations in the rest content of charge materials and other unknown factors may have contributed to higher iron oxide generation rates during the slot oven coke heat. As a result, further experiments are needed to verify the superior performance of formcoke with respect to alloy recovery.

An assessment of coke durability was made in these heats by measuring the coke fines screened while charging. The fines generated while charging slot oven coke during two periods represented 2.5% and 7.8% of the total coke weight, compared to 7.3% during the period studied while charging formcoke. Although the amount of data regarding durability is small, the fines generated by both cokes were similar.

Start-up with Formcoke Bed

Due to the much higher bulk density, cupola start-up procedures had to be altered to ensure a sufficiently burned-in bed when using formcoke. Alterations from the normal practice included using twice the normal amount of wood when building the bed, leaving the tap hole open during burn-in and tap out, increasing the blast air rate during burn-in, and blowing three times longer than was the practice with slot oven coke.

Conclusions

The controlled experiment demonstrated that the formcoke performance was on par with slot oven coke. The properties of the formcoke used in these tests were similar to the slot oven coke properties, with the exception of apparent and bulk densities, which were significantly higher for formcoke. Experience in tapping a cupola with a 100% formcoke bed indicated that the everyday practice of using [TABULAR DATA FOR TABLE 3 OMITTED] formcoke in a production environment would likely require some adaptation or revision of practices based on the use of slot oven coke.

RELATED ARTICLE: 50% Substitution at Intermet-Radford

A 50% substitution of this same formcoke was examined at Intermet Radford Shell Shop, Radford, Virginia. The foundry was interested in the alternative fuel because it was being produced by a promising new technology. with properties comparable to conventional cokes.

The foundry was most concerned with the formcoke's higher density, and what effect it would have on the light-off, melt rate, oxidation and C pickup. For that reason, an initial blend of 25% was studied.

After initial tests determined that 25% addition of formcoke had no impact on cupola performance, a substitution rate of 50% was studied with the goal of observing the durability of the formcoke and determining if there were any advantages to using it.

Test results indicated that the C level was consistent with the benchmark run. The Si and Mn level did decrease slightly and indicated the bed may have been slightly lower for the first run. As with the initial test, the S level was lower than the benchmark ran. In addition, based on a visual estimation of the fines collected during the test, it was estimated that the formcoke had 25% less loss than conventional coke.

The test didn't provide insight on whether the bed height should be adjusted to compensate for higher density nor whether light-off procedures should be changed if the weight of the coke is increased to keep the bed height the same, which can be answered only through additional testing. based on its limited study, however, it was concluded that a 50/50 blend of slot oven coke and the formcoke used did not adversely affect cupola performance.

- excerpted from a presentation by Richard Martin, Intermet Radford Shell Plant, at the 1997 AFS Southeast Regional Conference