Coal Preparation » Dewatering
High moisture contents in export coals increase transport and handling costs and invoke contract penalties. Considerable efforts have been focussed on fine coal dewatering, but relatively little research has been done on reducing the moisture content of the coarse product, which is up to 80% of the total product.
The first stage of this project established that there were four major factors controlling moisture retention. The present ACARP/BHPC funded project was established to investigate the use of reagents to decrease the propensity of water to remain on the coal surface.
The novel laboratory centrifuge technique, which was developed in the first stage of the project, has identified that there are three types of water associated with a coarse coal with respect to its potential for centrifugation;
- Water which has little relationship with the coal, and can be centrifuged off very easily.
- Water which is closely related to the coal surface, and the extent of its removal is affected by the surface physics and surface area of the coal.
- Water held internally within the coal which is related to the porosity of the coal.
A procedure was developed to allow the screening of reagents with respect to their potential to decrease the amount of water retained by coarse coal. This procedure was based on the laboratory centrifuge.
Samples were obtained from a number of Queensland and NSW minesites. These covered a broad range of rank from a maximum reflectance (R) of 0.6 to 1.7 .
The reagents classes used in the laboratory test were: (i) Anionic surfactants, (ii) Cationic surfactants, (iii) Non-ionic: particularly the block copolymer P103 and (iv) Commercial reagents: Dewat and Quadry.
A number of these reagents can lead to reductions of moisture content in the range of 0.8 to 1 percentage points. The effect appears to be greatest for coals of intermediate rank (R of about 1).
The results for surface tension measurement of solutions with and without coal, and zeta potential values obtained by an Acoustosizer indicated;
- The cationic surfactants adsorb strongly onto the coal surface and change the surface from negative to a relatively high positive charge. The surface tension of the solution in contact with the coal was that for water.
- The anionic surfactants adsorb to a small extent leading to a minor increase in the negative charge. The surface tension of the solution was lowered and it has a tendency to foam.
- The block copolymer did not change the zeta potential of the coal surface. Sufficient amounts of the compound remain in solution to lower the surface tension, and the solution had little tendency to foam.
The laboratory centrifuge results indicated that the additional water removed was that which tended to stay on the coal surface (NCMs) and the amount of water which is slow in its removal (ECM) is relatively unaffected.
A plant trial was conducted at Goonyella plant, using an on-line microwave moisture meter supplemented by manual sampling.
Quadry is a commercially available reagent which is based on water insoluble surface active chemicals dissolved in diesel. It has been previously demonstrated at Curragh and Goonyella that this reagent can lower the moisture content of centrifuged coal if sprayed onto the coal at the end of the drain and rinse screens.
Laboratory spray screen experiments linked with the laboratory centrifuge were able to show the same degree of moisture reduction as seen in the plant. It is interesting to note that the additional water which is removed is the water which is slow in its rate of removal (ECM), while the amount of water which tended to stay on the coal surface (NCMs) is relatively unaffected.
Three plant campaigns were carried out with the reagents being sprayed onto the product coal at the end of the two drain and rinse screen just prior to the coal dropping into the basket centrifuge.
Difficulties with the moisture meter on the first campaign meant that experimental work was dependent upon manual samples. This clearly demonstrated the difficulty of using this approach (uncertainty about ( 0.7% water) compared with the moisture meter (uncertainty about ( 0.25%). The problem was amplified by the low number of manual replicate samples which could be taken compared to the moisture meter which provided a data point every 3 minutes.
With the manual sampling it was not possible to statistically verify whether any of the reagents had an effect in lowering the moisture content of the coarse coal.
On the second campaign, the block copolymer P103 was found to reduce the moisture content of the coarse coal by about 1 percentage point on a number of separate trials. This result was relatively insensitive to a range of reagent addition levels and the quantity of coal on the screen.
The experiments with three other reagents, CTAC and Adogen (cationic surfactants) and Dewat (a commercial reagent) did not produce results which would demonstrate a moisture decrease.
Trials during the third plant campaign confirmed the result with the block copolymer P103.
It is considered that the results obtained with the block copolymer P103 are very encouraging and justify the conduct of a full plant trial to confirm the benefit. Clearly the commercial viability of using this reagent is dependent upon the value of reducing the moisture content of the coarse coal by this amount and the price which the potential chemical companies would charge to supply the chemical.
The laboratory investigation found that two other block copolymers also reduced the moisture content by a level similar to P103. Given these positive results with this reagent type, it is believed that a more comprehensive study should be done on the optimal combination of reagent hydrophobe to hydrophile functional groups with regard to level of reduction with the amount applied and the ease of application.
The is some evidence that block copolymers also assist in the flotation process hence this could be considered as well.
The laboratory coarse coal centrifuge test has been used as a tool for determining the dewatering properties of coarse coal since its development . However, to date only the CSIRO Division of Energy Technology and the University of Queensland Chemical Engineering Department have had access to the laboratory centrifuge.
So that the methods and equipment do not remain solely in the hands of the researchers, ACARP supported a project which documents the methods and equipment used to determine these properties. A document in the form of standard has been prepared, which will allow appropriate laboratory staff to build the necessary equipment and carry out the tests. The methods have been tested with three operators from different organisations and three coals of varying rank. This report describes in detail the results obtained, and includes the "standard" document as an Appendix.
The three types of moisture referred to in the results below can be described as follows:
- Exponentially Centrifugable Moisture (ECM) which is water closely related to the surface of the coal, and therefore results in a slow rate of moisture removal during centrifugation, (Exponential with time),
- the Non Centrifugable Moisture (NCM) which is the moisture content of the coal after centrifugation
- and the Non Centrifugal Moisture internal (NMCi) which can be interpreted as an internal moisture closely related to the pore structure of the coal. This water can only be removed by thermal drying.
The ECM results for all coals show a standard deviation between operators of 0.06, 0.06 and 0.10 for the three coals. The NCM has a standard deviation between operators of 0.10,0.18 and 0.10 and the NCMi of 0.22, 0.31 and 0.21.
Results show clearly that there is little deviation between operators. ECM, NCM and NCMi values were all within two standard deviations between operators.
The largest errors were found in the determination of the NCM for the larger size fractions. This is expected due to increasing sampling errors (less number of coal particles in the sample to make up the 1.5 kg), reduced surface area resulting in smaller and more variable quantities of water available for removal and changing petrography with size.
Comparison with plant results show the estimates from the method to be within the 95% confidence limits albeit on the top range of these estimates. The estimate, it must be emphasised, is the best possible from an optimised centrifuge with a normal centrifuge feed. Other factors such as the amount of fine particles adhering to the surface, the hydrophobicity of the coal and the size distribution will have a bearing on the centrifuge product moisture.