The effect of the solvent composition, with particular reference to the saturates content, on product yields and trace element concentrations in the filtered extract solution, produced in the first stage of liquefaction, has been studied. It was found that with up to 14% saturates in the solvent, conversions were of the same order of magnitude, with a possible maximum at 12% saturates. Above 14% the conversion quickly reduced. Preasphaltene and oil levels were found to generally increase with conversion, but asphaltenes decreased. Trace element levels in the filtered extract solution decreased overall, from an ash level of 364 ppm using a solvent with 8% saturates, to 209 ppm with 18% saturates. Most of the individual trace elements decreased in line with the ash level, except for manganese which showed a greater than average decrease, and titanium which, conversely, showed an increase.

The ash content, after filtration, of coal extract solutions produced using two different coals and two different samples of hydrogenated anthracene oil (HAO), and the consequential effect of varying digestion pressure and filtration temperature have been studied. It was found that if the digestion pressure is above 3.0 MPa, then the extract solution ash level is reduced significantly. However, this effect is modified by the coal or HAO used. This can be allowed for by reducing the filtration temperature for a given system. The condensed material vented during digestions was analysed, and it was found that there was no significant difference in the components present for different digestion pressures, only that a greater quantity was given off during lower pressure digestions. The vented material from low-pressure runs was added back to the digest prior to filtration, but it was found that this did not give low-ash extracts similar to those produced at higher pressure.

The trace element contents of the insoluble portions of a filtered coal extract solution treated with pentane, toluene and THF were determined using atomic absorption spectroscopy (a.a.s.) and atomic emission spectroscopy (a.e.s.). In all cases a large proportion of the trace elements were found in the insoluble portion, but there were no significant differences between the individual trace elements. Further experiments were carried out in which TiO2, MnCO3 and low temperature ash were added to normal digestions of coal in hydrogenated anthracene oil (HAO), and were also digested in HAO only. The filter cakes produced in the latter experiments were redigested in fresh HAO a second and third time. The results strongly indicate that Ti passes through the filtration stage of the liquefaction process as finely-divided mineral matter. Mn, however, undergoes some change during the process which enables it to pass through the filtration stage.

A hydrocracking rig has been constructed which involves the passing of coal-derived liquids and hydrogen over a trickle bed packed with a nickel oxide/molybdenum oxide catalyst. The ash and trace metal levels have been measured, before and after hydrocracking, and their variation with catalyst lifetime has been assessed. Al, Ca, Fe, Mg, Mn, K, Si, Na and Ti levels have been determined by a.a.s. and considerable reductions have been observed for Fe, Mg, Mn and Ti after hydrocracking. The ash level of hydrocracked extract increases for the first 50 h and then stabilizes at about 60% of the feed level. Metal levels determined for the catalyst itself show an accumulation of all the metals under consideration, especially for the sections of the catalyst which were hottest. Overall catalytic activity is not seen to decrease markedly but the ability of the catalyst to produce lower boiling material is considerably reduced after only 25 h. Comparisons between catalyst activity and metal deposition are made.

High- and low-temperature ashes from feed coal, coal extract solution and filter cake from a two-stage coal liquefaction process have been studied by X-ray diffraction (XRD) and scanning electron microscopy/energy dispersive X-ray fluorescence (SEM/EDX). Hydrocracking experiments using alumina support only, in place of the active NiMo catalyst on alumina, were also carried out, with trace metal analysis of the coal extract solution feed and hydrocracked extracts using atomic absorption and emission spectroscopy. The major mineral transformations occurring were of pyrite to pyrrhotite and the fixation of organic sulphur by calcium carbonate. Mineral particles were not observed in the coal extract solution ashes, even under high magnification, and the study indicated that size alone was not a determining factor as to whether a coal mineral was to be found in a coal-derived liquid. None of the trace metals was deposited on the alumina support under hydrocracking conditions, in marked contrast to the results obtained with the normal NiMo catalyst. These results lead to the conclusion that for the deposition of trace elements to occur a reaction must take place and hence the trace elements must be chemically bound in some form.