Francis P. Burke

Composition and performance of distillate recycle solvents from the SRC-I process

Abstract Recycle distillates (‘solvents’) from the Wilsonville, Alabama, SRC-I pilot plant have been analysed by a variety of methods, and the composition data related to the performance of the solvents in both batch and continuous unit coal liquefaction. Because of its low distillate yield, solvents in the SRC-I process evolve rather slowly from the ‘anthracene oil’ start-up solvent to steady-state, process-derived distillates. As the steady-state solvent evolves it increases in saturates and mono-aromatic hetero molecules (primarily phenols and indanols), and its quality as a liquefaction medium declines. Because it is low in condensed aromatics, the steady-state SRC-I solvent cannot be greatly improved by hydrogenation. Other analytical data are provided which describe the composition of SRC-I recycle distillates. The analytical methods used in this work are of adequate simplicity that they could serve as routine monitoring methods in SRC-I liquefaction.

Correlation of microautoclave and 1H n.m.r. measurements of coal liquefaction solvent quality

Abstract Some 467 coal liquefaction process oils, having nominal boiling points within the range 473–808 K (200–535 °C), were assayed for donor solvent quality in 577 microautoclave tests using three different sets of reaction conditions. Proton distributions were also determined on each sample by 1 H n.m.r. Microautoclave coal conversions and proton distributions, correlated using multiple linear regressions, give similar measurements of solvent quality over a wide range of solvents. Donor solvent quality (as measured by microautoclave coal conversion) increases with increasing hydroaromatic content and with decreasing aromaticity and paraffinicity. As they were designed to do, the different microautoclave tests measure different solvent properties. Above a certain level of solvent quality, microautoclave extractions are insensitive to solvent quality differences. However, 1 H n.m.r. can differentiate between these high quality solvents, and the data can be related to differences in process performance. Advantages of using 1 H n.m.r. for donor solvent quality measurements are discussed.

Liquid column fractionation: a method of solvent fractionation of coal liquefaction and petroleum products

Abstract A method is described for the solvent fractionation of coal liquefaction and petroleum products which is both reproducible and considerably more rapid than many conventional solvent fractionation techniques. This method involves sequential elution of a sample injected onto an inert liquid Chromatographic column. Applications of this method to coal liquefaction and petroleum products are given.

Agglomeration of low-rank coal as a pretreatment for direct coal liquefaction☆

Abstract Laboratory experiments demonstrated that low-rank coals can be cleaned by agglomeration with distillate coal liquefaction recycle oils. Materials tested included a lignite, three subbituminous coals, a bituminous coal, two petroleum oils, and three coal liquefaction recycle oils. Ash rejections obtained were as high as ~ 50% for Texas lignite and 15–20% for two Wyoming subbituminous coals. A Montana subbituminous coal showed no ash rejection, though physical agglomeration occurred. Organic recoveries for the low-rank coal tests were always > 98%. All three liquefaction recycle distillates tested were effective agglomerating agents for low-rank coals. In the few cases tested, adjustment of the slurry pH to 2.0 or addition of cresylic acid to the slurry had beneficial effects for the lignite agglomeration. Selectively retained in the product ash were Fe (a potential liquefaction catalyst), Ti, Ca, and Mg, whereas Na was selectively rejected.

Oil agglomeration as a pretreatment for coal liquefaction

Abstract Laboratory experiments demonstrated that a variety of distillate coal liquefaction recycle oils were satisfactory agents for cleaning Illinois no. 6 bituminous coal by oil agglomeration. Ash rejection up to 41% with 98% organic recovery was attained with conventionally cleaned coal, and ash rejection up to 67% with 90% organic recovery with run-of-mine coal. Agglomerates of > 1 mm average diameter were produced under a variety of conditions. Similar results were obtained in the scaled-up production of 268 kg of agglomerates. Oils with lower hydrogen aromaticities and higher hydrogen contents performed better than more aromatic oils. Fe, Ti and Mg were selectively enriched in the ash of the product coal, while Ca, Si, and Al were selectively rejected. The mineral pyrite was rejected only ≈ 30–40% as extensively as the bulk of the ashforming minerals. The coal cleaned by oil agglomeration performed similarly to the feed coal in batch donor liquefaction tests. In continuous hydroliquefaction tests, run-of-mine coal cleaned by oil agglomeration performed substantially better than coal cleaned to the same ash level by conventional means, because of the selective enrichment of catalytic iron minerals.

Improvement in coal liquefaction solvent quality by dewaxing

Abstract Recycle oils from the Integrated Two-Stage Liquefaction (ITSL), H-Coal and Solvent Refined Coal (SRC) processes were dewaxed by variants of commercial dewaxing processes—the ketone and the urea adduction techniques — yielding up to 47 wt % ‘wax’. Feed oils and product fractions were characterized by elemental analysis, 1 H n.m.r. and gas chromatography. The clean waxes were nearly pure mixtures of n -paraffins. The dewaxed oils were substantially better coal liquefaction solvents than the original (non-dewaxed) oils in batch liquefaction tests. For example, in one case, dewaxing improved the conversion of a bituminous coal to tetrahydrofuran-solubles under standard reaction conditions from 71 wt% (dafb) with the original oil to 87 wt % (dafb). These data provide a direct indication of the inimical effect of paraffinic components on solvent quality. The impact of solvent quality is particularly relevant to liquefaction processes in which thermal reactions proceed in a recycle solvent. In addition, the results indicate the technical feasibility of dewaxing coal liquefaction recycle oils by commercially available technology to improve solvent quality and to produce a useful by-product. Dewaxing could be applied in any liquefaction process that uses a deasphalted (preferably distillate) recycle stream.

Retrograde reactions in SRC-I liquefaction

Abstract Operating the vapour-liquid separators in the solvent refined coal (SRC-I) process at or near the reactor temperature is desirable to retain the heat in the liquid product stream which is necessary for subsequent distillation. However, adverse changes in the product composition resulting from high temperature operation (i.e., ‘coking’) would obviate this advantage. A study of these potential retrograde reactions using a 4.5 kg h −1 bench-scale coal liquefaction unit is reported. At high pressure, separator temperatures up to 722 K and residence times in the pressure let-down system of 15 to 30 min, retrograde reactions occur as conversion of a few per cent of the product SRC to insoluble organic matter. If sufficient agitation is present in the separators these retrograde reactions would have little effect on product yields or unit operability. Coking, manifested as the accumulation of significant deposits of anisotropic carbon in the separator vessels, occurs only with the extended time-at-temperature that the solids can experience in a poorly agitated vessel.

Stable carbon isotope analysis of coprocessing materials

The purpose of obtaining stable carbon isotope analyses of coprocessing products is to determine the amount of coal (or petroleum) carbon that is present in any reaction product. This carbon-sourcing of distillate fractions, soluble resid, and insoluble organic matter, etc. is useful in modeling reactions, and evaluating synergistic effects if they exist.

N.m.r. determination of aromatic carbon balances and hydrogen utilization in direct coal liquefaction

Abstract Solid- and liquid-state 13 C n.m.r. measurements were made on a suite of samples obtained from different stages of a coal liquefaction run at the Wilsonville Two-Stage Advanced Coal Liquefaction Research and Development Facility. The n.m.r. measurements were combined with elemental analysis and mass balance data to measure aromatic carbon balances and hydrogen utilization for Wilsonville coal liquefaction run 259G. This was a catalytic/catalytic integrated two-stage liquefaction run on a deeply cleaned, Pittsburgh seam, high volatile bituminous (hvAb) coal from the Ireland mine in West Virginia, USA. The n.m.r. measurements showed that on the basis of the feed coal, about 58% of the aromatic carbons in Pittsburgh coal were hydrogenated during two-stage liquefaction, and 55% of the hydrogenation occurred during the second stage. A net of 68.1 mol of hydrogen per 100 mol of coal carbon was consumed during the first stage of Wilsonville run 259G. This amounted to 69% of the total overall two-stage hydrogen consumption. Matrix cleavage and hydrogenation reactions accounted for 31% and 27%, respectively, of the total first-stage hydrogen consumption. Hydrogenation reactions accounted for most (69%) of the hydrogen consumed during the second stage, and accounted for 40% of the overall two-stage total hydrogen consumption. Most of the total hydrogen consumed for hydrocarbon gas generation (70%), and for heteroatom removal and heterogas production (71%) occurred in the first stage. Overall hydrogen consumption was 14% for hydrocarbon gas production and 27% for heteroatom removal and heterogas production.