INTERIM FINAL TECHNICAL REPORT

October 1, 1998, through October 1, 1999

Project Title: DEVELOPMENT AND DEMONSTRATION OF INTEGRATED CARBON RECOVERY SYSTEMS FROM FINE COAL PROCESSING WASTE

ICCI Project Number: 99-48202

Principal Investigator: R. Q. Honaker, Southern Illinois University

Other Investigators: B. K. Parekh and D. Tao, CAER/UK; and Latif Khan, ISGS

Project Manager: Wayne Bahr, ODCM/DCCA

ABSTRACT

Samples from two different refuse coal ponds were collected for use in this investigation. The refuse pond sites include the Humphrey Mine of CONSOL Inc. in northern West Virginia and Peabody's Randolph Mine located in Marissa, Illinois. The fine coal refuse from the Humphrey site was derived from the Pittsburgh No. 8 coal seam whereas the Randolph pond represents Illinois No. 6 seam coal. The specific sampling locations from within each pond were determined using existing borehole information for the Humphrey pond and random sampling characterization data for the Randolph pond.

Characterization of the Pittsburgh No. 8 fine coal refuse revealed that only about 2.6% of the total weight has a particle size greater than 16 mesh whereas 29.5% is less than 500 mesh. The ash and total sulfur contents of the sample are 31.63% and 4.73%, respectively, with an ash content of the -500 mesh size fraction being 50.7%. The calorific value of the head sample was found to be around 10,000 BTU/lb on a dry basis. According to washability results, the Pittsburgh No. 8 refuse material is considered to be moderately difficult-to-clean using gravity-based processes based on a 5% - 10% content of near-gravity material. The data suggests the possibility of producing a product containing 9% ash while recovering greater than 90% of the combustible material. Result from release analysis tests, which were conducted on -65 mesh material, indicates that froth flotation has the ability to produce a clean coal concentrate containing 10.7% ash with a combustible recovery of nearly 92%. Characterization of the Illinois No. 6 fine coal refuse, collected from Peabody's Randolph refuse ponds located near Marissa, Illinois, showed that about 4.4% of the total weight has a particle size greater than 16 mesh while-500 mesh size fraction is less than 36.2%. The ash and total sulfur contents of the sample are 33.11% and 2.21%, respectively, with an ash content of the -500 mesh size fraction being 65.6%. The calorific value of the head sample was found to be around 8600 BTU/lb on a dry basis. According to washability results, the Illinois No. 6 refuse material is considered to be moderately easy-to-clean using gravity-based processes based on a 5% - 10% content of near-gravity material. The data suggests the possibility of producing a product containing 7.5% ash while recovering greater than 95% of the combustible material for +325-mesh size fraction. Results from release analysis tests, which were conducted on -48, -325 and 48 x 325-mesh materials, indicate that froth flotation is rather ineffective to produce a clean coal concentrate with an adequate level of combustible recovery.

A statistically designed test program was performed to identify the crucial operational variables of the Falcon concentrator for both coal samples. Generally, four operating parameters, i.e., bowl speed, feed solids content, underflow rate, and feed flow rate, were found to affect the concentrate yield. A laboratory scale unit of the ISGS froth washer has been constructed and tests initiated to quantify the design and operating parameter effects on separation performance.

Surface characterization of the concentrates generated from the Pittsburgh No. 8 and Illinois No. 6 Jameson Cell products was conducted and revealed interesting information that will impact dewatering. The zeta potential versus pH relationship of the two coals was found to not follow the typical trends in that negatively charged surfaces were obtained at low pH values and positively charged at high pH values. Optimum dewatering corresponds to the point-of-zero charge (pzc) for both coals, which occurred at pH 7.2 for the Pittsburgh No. 8 material and 11.3 for the Illinois No. 6 concentrate. Baseline vacuum and pressure filter dewatering tests produced cake moistures of 25% and 15%, respectively, from the Pittsburgh No. 8 Jameson Cell concentrates. Using a combination of vacuum and pressure filtration, the cake moisture content was reduced to nearly 12%.

EXECUTIVE SUMMARY

To date, the most common strategy used to recover the carbon from fine-coal refuse ponds incorporate the use of spiral concentrators to treat the nominal +100 mesh material and column flotation cells to clean the -100 mesh size fraction. However, spiral concentrators have a relatively low mass throughput capacity of 3 to 3.5 tons/hour per start. Also, spiral concentrators typically have high separation densities (1.80) and average separation performances as indicated by probable error (E_p) values between 0.12 and 0.18. Due to these reasons, it is difficult to achieve low product ash contents when treating moderately difficult-to-clean coals. Advanced fine coal cleaning technologies have been found to provide separation performances that are near or approaching the theoretical maximum based on either the washability curves for gravity separations or the release curve for froth flotation processes. Since the coal particles in the fine fraction represent the cleanest particles in the run-of-mine coal, the application of the advanced fine coal technologies should provide the highest possible product grades and mass yields in an operating preparation plant.

The Falcon Concentrator is such an enhanced gravity concentrator that can provide efficient low gravity cut points. Development of the Falcon Concentrator, it's application to coal preparation and it's novel dense medium application is a product of research aimed at improving the efficiency of fine particle processing, funded in a large part by the State of Illinois. It has been found that the Falcon Concentrator has the ability to improve the rejection of both ash-bearing material and coal pyrite while maximizing the recovery of fine coal. The process effectively treats particle sizes from 16 mesh to as small as 325 mesh. The dense medium application of Falcon Concentrator has been able to provide gravity cut points of 1.42-1.49 at exceptional probable error values between 0.038-0.05.

Samples from two different refuse coal ponds were collected for use in this investigation. The sites include a refuse pond derived from a processing plant treating Pittsburgh No. 8 seam coal in northern West Virginia and a refuse pond comprised of Illinois No. 6 fine coal located in southern Illinois. The specific sampling locations from within each pond were determined using existing borehole information for the Pittsburgh No. 8 seam coal refuse pond and by random sampling characterization data for the Illinois No. 6 seam coal refuse pond.

Water-only Falcon

An experimental program based on a statistical design was conducted to evaluate the parametric effects and optimize the performance of the C10 Falcon. A total of 29 experiments based on a Box-Behnken design were carried out over a range of operating parameters. The parameters studied were underflow rate, bowl speed, feed solids content and feed volumetric flow rate. The metallurgical results obtained from the treatment of the Illinois No. 6 seam coal sample are compared with the washability curve in Figure 1(a). These results show that it is possible to obtain a 48 x 325 mesh clean coal having approximately 7.50% ash with about 80% yield from a feed coal having 13.33% ash with a combustible recovery of 84%. The ash rejection is about 55%. The total sulfur content in the feed is about 3.12%. The total sulfur content in the Falcon products is in the range between 2.45% to 2.75%. The small reduction in the total sulfur content in the Falcon product is mainly due the presence of higher percentage of other forms of sulfur as compared to pyritic sulfur.

It has been clearly demonstrated that the froth flotation process is unable to reject pyritic sulfur. One of the important advantages of the enhanced gravity separation in fine coal cleaning is the rejection of pyritic sulfur. Figure 1 (b) shows the metallurgical results of Pittsburgh No. 8 refuse coal. At about 72.49% mass yield, the water-only Falcon concentrator was able to reduce the ash content from a feed coal having 16.34% to 8.79%. The total sulfur content was reduced to 2.39% from 4.77%. Nearly a 60% rejection of ash bearing material and 63% of the total sulfur to the tailings with a combustible recovery of 78% indicate a relatively efficient separation.

Dense Medium Falcon

Experiments based on a Plackett-Burman design were conducted on the dense medium Falcon operation to identify the most significant process variables. Based on these results, four important process variables, medium density, underflow rate, lip width and feed flow rate were identified for a more detailed test program conducted according to a Box-Behnken design. Using a continuous C10 Falcon concentrator, tests were conducted using the Illinois No. 6 refuse coal sample. The dense medium was formulated using magnetite containing 99% minus 325 mesh particles. Prior to the tests, the coal sample was de-slimed to obtain a nominally 16 x 325 mesh feed coal sample. Experimental results (Figure 2) show that that a clean coal product containing 6.82% ash with 92% combustible recovery was achieved from the dense medium Falcon process treating the Illinois No. 6 refuse coal sample having a feed ash content of 13.83%. The ash rejection was about 60%. The performance of the dense medium Falcon process is very close to the washability curve, which represents the ideal separation possible by the physical separation.

The extensive evaluation conducted in this investigation indicates that the dense medium Falcon operation is a unique high efficiency and high capacity process for treating coal over a wide particle size range, i.e., 16 x 325 mesh. Apart from the techno-economic benefits, the dense medium Falcon may also provide operational flexibility such as the ability to achieve a wide variation in separation density cut points (d₅₀) without changing the feed medium density. For example, a range of density cut points from 1.4 to 1.6 was achieved by manipulating bowl speed and underflow rate while maintaining the feed medium density at 1.5.

Column Flotation

Experiments were conducted to evaluate the separation performance achievable by commercial column flotation technologies like the Microcel and the Jameson cell in comparison to the theoretically best possible performance as indicated by the release analysis procedure. Tests were conducted on the -48 mesh size fraction of both the original Pittsburgh No. 8 seam coal refuse pond sample as well as the Falcon concentrator product obtained from the treatment of the same sample. Performance curves (Figure 3) obtained from the experiments were used to evaluate the need and/or benefits that could be achieved from froth flotation in a circuitry arrangement for generation of high quality carbon fuel from coal refuse ponds.

An interesting aspect of the enhanced gravity concentration prior to column flotation is visible in Figure 3. Column flotation treatment of the -48 mesh particle size fraction of the Falcon concentrator product (ash content = 32.07%, total sulfur content = 2.38%) results in a far superior separation performance compared to the treatment of the same size fraction from the original pond sample (ash content = 37.04%, total sulfur content = 5.18%). The release curves indicate that pretreatment of the column flotation feed can provide an ability to obtain products with lower ash contents by as much as 6 percentage points at the same combustible recovery values. This is caused by the efficient rejection of middling particles in the Falcon concentrator that cannot be achieved through single stage cleaning using column flotation.

As part of the circuitry evaluation, a long-term test was conducted treating the Falcon concentrator product using column flotation in this fashion. Approximately 200 kilograms of froth product was generated and 100 kilograms of tailings collected. The product and tailings samples were individually mixed for homogenization and sampled for assaying purposes. Analysis was performed for product ash, total sulfur content and their calorific value. From a feed containing 32.07% ash, 2.38% total sulfur and a calorific value of 9949 Btu/lb, a concentrate product containing 5.62% ash, 2.35% total sulfur and 14,648 Btu/lb calorific value with corresponding tailings assay values of 73.22%, 2.42% and 2,667 Btu/lb was produced. These values correspond to a mass yield of 60.87% and a combustible recovery value of 84.57%. This clearly indicates the excellent cleaning potential of column flotation and demonstrates its use for desliming the Falcon concentrator for production of high quality fuels.

Froth Washer Testing

In the present reporting period, systematic statistically designed experiments, based on a 1/8 fractional factorial design, using the seven parameters were conducted to investigate their relative importance on the performance of the ISGS froth washer system. The critical parameters affecting the response variables of yield, combustible recovery, ash rejection, total sulfur rejection, product ash and product total sulfur were identified using a test of hypothesis involving the coefficient estimate of the parameter effect being equal to zero at a confidence level of at least 90%. Four critical parameters, namely, feed percent solids, feed volumetric flow rate, bias factor and froth washer angle, were identified to be the most significant parameters. These parameters were used to develop an empirical response surface model for prediction of the froth washer performance at any given set of parameter settings. The model was developed using a statistical Box - Behnken experimental design methodology. The model was then used to optimize the froth washer performance for achieving premium separation performance.

In order to quantify the advantages of the novel froth washing system, a direct comparison with the performance of a 2-inch Microcel flotation column fitted with a conventional spider-type washing system has been performed. The base separation performance and maximum carrying capacity value has been established for the treatment of the -48 mesh particle size fraction of the Pittsburgh No. 8 seam coal refuse pond sample. In addition, the ultimate separation performance achievable using froth flotation has been estimated using the Advanced Flotation Washability (Modified Release Analysis) method.

Preliminary results indicate a promise for the novel froth washer being developed as a part of this investigation. The sulfur rejection ability of the froth washer appears to be exceptional as seen in Figure 4 (a). The data was generated from the treatment of -48 mesh particle size fraction of the Pittsburgh No. 8 seam refuse pond coal sample (feed total sulfur content = 5.18%). The application of the new froth washer to a conventional open cell, such as the Microcel, resulted in products containing 3% total sulfur content at 70% combustible recovery. The best performance that was achieved from the conventional column was a 3.5% total sulfur content product at a combustible recovery of 48%. Given the fact that conventional column flotation technologies are fraught with poor sulfur rejection problems, the new froth washer-column flotation design can have a great impact on the coal industry. The ash rejection performance of the new froth washer column flotation design in currently under investigation. It is well known that conventional column flotation technologies can provide excellent ash rejection performance, especially for low-middling content feeds. The data obtained from the Microcel treating the -48 mesh Pittsburgh No. 8 seam refuse coal sample confirms the observation. Preliminary testing with the new froth washer-column flotation design indicates a very slightly inferior ash rejection performance based on the data obtained as part of the statistical evaluation and model development. However, optimization of the developed model predicts a slightly superior performance from the froth-washer column flotation. The optimized model predictions remain to be verified. The preliminary results obtained thus far are presented in Figure 4 (b).

High Efficiency Circuitry

The proposed advanced coal cleaning circuit (Figure 5) shows the novel use of a hindered-bed classifier referred to as the Floatex density separator. Hindered-bed settling devices have been commonly used for achieving particle classification. However, a recent trend in the use of hindered-bed classifiers is to take advantage of the fluidization of a particle bed to create conditions suitable for separation of multi-species suspensions having wide density difference

between the various particulate components. The goal of using the Floatex separator was to provide sufficient rejection of coarse reject to allow a high quality product to be obtained by screening the overflow stream with a 48 mesh sieve. As shown in Figure 7, the ash content produced from the sieve overflow after treatment in the Floatex was 6.45% compared to 18% in the +48 mesh size fraction in the feed. The total sulfur content in this fraction was also significantly reduced from 4.17% to 2.56%. By comparing the data with the feed washability results (Figure 1), one can see that the separation density (d₅₀) required to achieve this process performance would be 1.5 gm/cm³ or less.

The underflow of the 48-mesh screen was fed to the water-only Falcon. The Falcon overflow (clean coal) after screening through a 325 mesh screen contained about 7.25% ash and 2.43% total sulfur. The Falcon overflow (without screening) was fed to the flotation column to remove the ash-bearing slime material. The flotation column successfully deslimed the Falcon product to produce a 5.62% ash product while achieving a combustible recovery of 84.57%.

The overall circuit was found to produce a high quality fuel from the Pittsburgh No. 8 refuse material. The product ash and total sulfur contents were 5.81% and 2.40%, respectively, which equates to rejection values of 90% and 72%. Based on the 52.26% mass yield achieved by the circuit, the organic efficiency (=actual recovery/theoretical recovery) was a relatively low 80%. However, the circuit was not optimized with respect to the integrated performances of the various process units, which is the focus of an on-going project. However, the results do show the ability to achieve effective low separation densities, which allows the production of high quality fuels from fine coal resources.

The characteristic and process performance results achieved so far in this investigation prove that high quality fuels can be produced from the waste found in fine coal refuse ponds. Product ash contents below 10% and, in some cases, 5% were produced from the processing of two refuse pond materials. In addition, significant reductions in total sulfur contents were achieved for the fine coal refuse containing a relatively large amount of pyrite. Coal refuse ponds can contain a significant amount of middling particles and coal pyrite. For these materials, the integration of an enhanced gravity concentrator and an advanced flotation system was found to be necessary to achieve a high quality product. For the Pittsburgh No. 8 coal, flotation of the -48 mesh material limited the quality of the clean coal product to approximately 8%. However, pretreatment of the material in an enhanced gravity system allowed the production of clean coal containing less than 4% ash at equivalent combustible recovery values. Therefore, integration of the two processes in a circuitry arrangement provides the ultimate separation performance.

Enhanced Dewatering without Additives

The pzc of the Pittsburgh and the Illinois clean-coal samples were at pH 7.25 and 11.3, respectively. The pH of the Pittsburgh and the Illinois clean-coal slurry were 3.3 and 4.4, respectively. The conductivities of the Pittsburgh slurry was 978 µ /cm and for the Illinois slurry was 569 µ/cm. The base-line vacuum dewatering tests conducted on the Pittsburgh No.8 Jameson cell product indicated that using 60 seconds filtration time and about 6 mm thick cake, a filter cake with about 25% moisture can be obtained. The base-line pressure filter dewatering data indicated that using 40 psi (3 bar) pressure and 90 seconds filtration time a filter cake with about 15% moisture can be obtained for the Pittsburgh No. 8 Jameson cell product. A combination of vacuum and pressure filtration for a 4 mm thick cake lowered the filter cake moisture from 29 percent obtained with the pressure alone, to about 12 percent.