Liquefaction of coal by SRC-II process: Part V: Dynamic thermal behavior of the reactor

Criteria for identification of the unstable or stable nature of the steady state based on the reactor temperature response to step changes in feed temperature, and a dynamic thermal simulation of SRC-II reactors are developed. These are used in the analysis of a dynamic SRC-II reactor experiment to confirm its unstable operation under normal process conditions. The simulation is used further to study the sensitivity of reactor temperature to changes in feed temperature and to variations in the total heat capacity of the reactor vessel and insulation materials. It is shown that, under normal SRC-II process conditions, it would take about an hour for the reactor temperature to change by ± 10°C, if no controls were used. With quench gas manipulation, the reactor temperature could be maintained within ± 1°C even for a sustained small perturbation in the process conditions. Therefore, in spite of its unstable character, the SRC-II reactors can be readily operated with quench gas manipulation without incurring any major temperature control problems.     

MULTIPLE STEADY STATES IN AN ADIABATIC COAL LIQUEFACTION REACTOR—ROLE OF PREHEATER

A general model is developed to predict the occurrence of multiple steady states in a coal liquefaction reactor. The effect of preheater variables under coal liquefaction conditions, on the occurrence of the multiple steady states is examined. An analytical expression is developed for the stability of the steady states for the case of a constant flow stirred tank reactor.     

MULTIPLE STEADY STATES IN AN ADIABATIC COAL LIQUEFACTION REACTOR—ROLE OF PREHEATER

A general model is developed to predict the occurrence of multiple steady states in a coal liquefaction reactor. The effect of preheater variables under coal liquefaction conditions, on the occurrence of the multiple steady states is examined. An analytical expression is developed for the stability of the steady states for the case of a constant flow stirred tank reactor.     

Free radicals in coal and coal conversions. 6. Effects of liquefaction process variables on the in-situ observation of free radicals

To determine the effects and relative importance of process variables in coal liquefaction, a uniquely designed and fabricated high-pressure/high-temperature electron spin resonance (e.s.r.) apparatus is used to monitor the in-situ formation and behaviour of free radicals, which are generally assumed to be the key factor. It is concluded that the temperature is the most significant single process variable that affects free radical formation; for Powhatan No. 5 coal there is a 9-fold increase in going from 400 to 460 °C. At 460 °C the other process variables tested can affect significantly the free radicals significantly, but at 400 °C these variables have essentially no effect on free radicals formation. The next most significant effect is due to the combination of solvent nature and residence time. Tetralin and the SRC-II heavy distillate quench the free radicals from Powhatan No. 5 to the same extent with one significant difference. In tetralin the maximum concentration is observed shortly after the slurry achieves its highest temperature, whereas in the SRC-II heavy distillate experiments the concentration is still increasing, at 460 °C, even after 1 h of reaction. The heating time, pressures and types of gas used affect the free radical concentration to a much smaller extent. The conversions obtained in the in-situ e.s.r. experiments using SRC-II heavy distillate as the solvent are somewhat lower than those obtained with tetralin as the solvent. The corresponding oil yields with tetralin are considerably higher than with SRC-II heavy distillate.     

Liquefaction of coal by SRC-II process: Part IV: Steady-state thermal behavior of the reactor

An axial dispersion thermal model of an SRC-II reactor is developed. The model is validated by simulation of steady-state temperature profiles in the Fort Lewis (Tacoma, Washington) pilot-plant SRC-II reactor. The simulation shows that normal operations of the SRC-II reactors are thermally unstable. Changes in the process conditions can be utilized to achieve thermal stability for the reactor at normal SRC-II operating temperatures, i.e., reactor outlet temperature of less than 465°C and average reactor temperature of about 455°C. However, thermal stability of the reactor would always result in reduction in the rate of reaction. Regions of unstable steady-state reactor temperature over large ranges of ash concentrations in the reactor and reactor feed temperatures are also illustrated.     

Free radicals in coal and coal conversions. 8. Experimental determination of conversion in hydroliquefaction

The results of conversion determinations on the products from Powhatan No.5 coal liquefied in an autoclave and in a high-pressure, high-temperature e.s.r. cavity are reported. Oil, asphaltene and preasphaltene yields, and overall conversion have been determined for Powhatan No.5 coal samples liquefied in tetralin, SRC-11 heavy distillate, and naphthalene at temperatures from 400 to 480 ° C in both reactor systems. The concept of reaction severity is introduced and used to formalize the relation between the effect of temperature and reaction time on oil yield and conversion. Oil is the predominant product in liquefaction in tetralin or naphthalene, asphaltene is the major product of liquefaction in SRC-II heavy distillate. Retrogressive reaction (THF-insoluble product formation) becomes severe when SRC-II heavy distillate is the liquefaction solvent and residence time of >10 min are used at temperatures >450 °C. Preasphaltenes appear to be the only intermediate species in Powhatan No.5 liquefaction.     

Free radicals in coal and coal conversions. 5. Methodology for the in-situ investigation of free radicals in coal depolymerization under SRC-II preheater/reactor conditions

In a continuing series of publications, a systematic investigation of the free radicals during coal depolymerization under SRC-II preheater/reactor conditions has been made. A uniquely designed electron spin resonance (e.s.r.) cavity allows the monitoring of free radicals and how they are affected by residence time, temperature and pressure changes. This paper summarizes the methodology that has been developed to study systematically free radical formation and behaviour in-situ during coal liquefaction. The methodology involved is to be used extensively in subsequent papers that examine in detail the manifold parameters that affect coal depolymerization. Also, results are given on the dependence of free radicals on temperature, pressure and gases in the presence of tetralin. The free radical concentration can be measured 3–4 min after the initiation of the heating process to an accuracy of ± 20%, while the process variables may cause several-fold changes in the free radical concentration.     

Pyrolysis of coal liquids

The pyrolysis of process recycle solvent derived from Western Kentucky coal via the SRC-II coal liquefaction process was investigated to ascertain the effect of residence time and temperature on the production of olefins. The study was made using an alonized transfer line reactor operating at temperatures of 650 and 730 °C, essentially atmospheric pressure, and residence times up to 0.13 s. A comparison is made with previously published results for the pyrolysis of hydrotreated COED light and heavy coal liquids (subsequently referred to as COED light and heavy oils, respectively) derived from Western Kentucky coal and steam pyrolysis of a hydrogenated fraction of SYNTHOIL derived from Western Kentucky coal. Results indicate that in each case the preferential pyrolysis of the saturate fraction occurs under convential pyrolysis conditions. Ethylene, propylene, and methane were the dominant gas products in all cases. The liquid pyrolysates from the COED oils and SRC-II recycle solvent had lower $\text{H}\text{C}$ ratios and heating values than their respective feedstocks. Mass spectroscopic analysis of the liquid pyrolysates in each case revealed the presence of polycyclic aromatics that were not present in the individual feedstocks. This trend which increased with temperature is indicative of cyclization and/or recombination of free radicals during pyrolysis. It is therefore surmized that the yields of light olefins from primary coal liquefaction products can be improved by partially hydrogenating them prior to pyrolysis. Alternatively, sufficient hydrogen can be provided in the vicinity of cracking to suppress retrogressive reactions which lead to the formation of coke. The pyrolysis of COED oils and SRC-II recycle solvent was found to follow first-order irreversible kinetics. The activation energy for the pyrolysis of the COED light and heavy oil was found to be 76.1 and 70.85 kJ g-mol^?1, respectively.     

Kinetic rate model interpretation of the SRC-II liquefaction of an Ireland Mine coal

An experimental data base for characterizing the SRC-II liquefaction of an Ireland Mine coal is interpreted using an empirical rate model developed from SRC-II liquefaction data for a Powhatan No. 5 coal exclusively. The model predictions for hydrogen consumption and for the yields of the key components; C₁-C₄ gases, C₅-755 K liquids, and solvent-refined coal (SRC) are compared with the experimental values over a wide range of reactor operating conditions; temperature (703–743 K), pressure (10.2–20.4 MPa), and recycle ash (2.0–10.4 wt%). Model predictions are in good agreement with hydrogen consumption measurements at baseline conditions (728 K and 13.6 MPa) and at more severe conditions (743 K and 20.4 MPa) for a feed slurry containing 30 wt% coal and 10.4 wt% recycle ash (absolute error ± 6.4%). A similar result is obtained forthe C₁-C₄ gases and the C₅-755 K liquids at the baseline conditions; however, the error in the SRC yield prediction is higher (± 12.1%). The model predictions for the key components at the higher severity conditions deviate more than those for the baseline conditions.     

Stability of some coal and tar sands syncrude fractions

Three fractions from a SRC-II coal liquefaction process, two naphthas from Illinois No. 6 coal and Wyodak coal using the Exxon Donor Solvent liquefaction process, and three fractions from tar sands extract were characterized for physical and chemical properties. Selected samples of syncrudes were stored at 43 °C for 32 weeks in air. The tar sands materials more closely resembled the properties of present petroleum-based materials, whereas the coal-derived liquids from the SRC-II process demonstrated the need for considerable upgrading. Heavier fractions seemed to undergo polymerization to produce gums, whereas lighter fractions deteriorated via oxidation.     

Scale formation in coal liquefaction reactor using FeS catalysts

The origin and formation behavior of scale in a coal liquefaction reactor, in which disposable catalysts such as red mud-sulfur were used, has been investigated by means of model experiments. The scale formed in the continuous coal liquefaction reactor which operated at a throughput of 0.1 t/d was analyzed with an electron probe microanalyzer (EPMA), scanning electron microscope (SEM), X-ray and scanning Auger electron microscope (SAM) techniques. It was observed in the model experiments that the scale formed on the surface of stainless steel (SUS-316) at temperatures higher than 250°C. The amount of scale formed was found to depend on the reaction temperature, the exposure time and the sulfur content of the catalyst. The scale formed in the continuous reactor collected on the walls of the tube, preheater and reactor. It was found that the scale was derived from two sources, the coal paste (catalyst, inorganic components in coal and unreacted coal) and the sulfidation of reactor materials.     

SOME ASPECTS OF DESIGN OF COMMERCIAL REACTORS FOR DIRECT COAL LIQUEFACTION

In recent years, a number of direct coal liquefaction processes have been developed. All processes use a slurry type reactor. Although for lab-scale reactors of large length-to-diameter ratio the use of highly sophisticated slurry reactor model may be justified, simple considerations can meaningfully elucidate the behavior of industrial reactors. A simple analysis shows that the coal liquefaction is controlled by intrinsic kinetics. Both gas and slurry phases can be assumed to be completely backmixed in large diameter reactors. A simple analysis of the thermal behavior revealed multiplicity for a fairly wide range of operating conditions. In most cases, the intermediate unstable steady state is close to the temperature observed in adiabatic coal liquefaction reactors (with and without quench). Due to the unstable character of the operation, point pathological phenomena like runaway may be possible and a close feedback control of the commercial reactor may be required.     

Hydrogen mass transfer and mixing energy effects in SRC-II coal liquefaction reactors

This paper summarizes the experimental work performed on a bench-scale pre-pilot unit for investigating hydrogen mass transfer and mixing energy effects in SRC-II coal liquefaction reactors. Experiments were carried out with an Ireland Mine coal where the effects of mixing energy level (150–1000 rpm), method of hydrogen introduction (preheater flow and direct reactor sparging) and hydrogen treat rate (4 to 6 g of hydrogen/100 g of feed slurry) were investigated. Several runs using Powhatan No. 6 coal were also carried out where the effect of mixing energy level (200–1000 rpm) was investigated. Other run conditions were fixed to correspond to those likely to be used in commercial operation. The experimental results clearly indicated that below a mixing energy level corresponding to 400 rpm a significant cement-like solid deposition within the reactor (hydrogen mass transfer limitation) occurred. Below this mixing energy level the C₅+ liquid yield decreases, and the selectivity of the reaction changes, resulting in an increase in the C₁C₄ yield. This critical mechanical mixing level corresponds to a mixing energy per unit of reactor volume of ≈3500 ergs/cm³ s (350 watts m^?3). For the run conditions employed, increasing the preheater hydrogen flow from 4 to 6 g of $\text{H}{}_2\text{100 g}$ of slurry prevented the formation of solid deposits at a mechanical mixing energy level as low as that corresponding to 200 rpm. Furthermore, the highest C₅+ yield in the entire data set occurred when the preheater hydrogen flow was at the higher level.     

Liquefaction of Yallourn brown coal at low pressures and rapid heating rates

Liquefaction of Yallourn brown coal in solvents at high temperature for short contact times and low pressures has been studied. Very high asphaltene yields are achieved with hydrogen-donating solvents (hydrogenated Ashland pitch A240, hydrogenated anthracene oil, and hydrogenated pyrene). For hydrogenated pyrene, yields of almost 90% were obtained during reaction at 450°C for 10 min or at 510°C for 1 min. The average molecular weight of the asphaltene found was 270, with 40 wt% being accounted for by three-and four-ring polynuclear hydrocarbons. The effect of liquefaction temperature, time, and solvents on the asphaltene yield have been examined to clarify the properties required for the solvent under the present conditions used. The behaviour of the asphaltene during pyrolysis and hydrotreatment has also been studied. Some mechanistic aspects of high-temperature, short contact time liquefaction are discussed with regard to the reactivities of the brown coal and the solvents.     

The effects of iron sulphide surface area variations on coal liquefaction

The purpose of this work was to determine the effects of surface area variations of iron sulphides on coal liquefaction. Several iron sulphides were synthesized including pyrites (FeS₂) with 46.6 wt% Fe, pyrrhotites (Fe_1?xS) with ~ 60 wt% Fe and iron-sulphur compounds of unknown composition. Surface areas of the synthetic pyrites varied from 2 to ? 10 m² g^?1, pyrrhotite surface areas were 6 and 10 m² g^?1, and the surface areas of the iron-sulphur compounds were 40 and 80 m² g^?1. These iron sulphides were tested for catalytic activity in tubing reactor runs with West Virginia Blacksville no. 2 coal and SRC-II heavy distillate. All these sulphides showed catalytic effects as compared to runs with only coal and solvent, although the effects were not as large as those obtained with a cobalt-molybdenum on alumina catalyst. Large differences in surface areas before reactor testing did not cause any significant differences in conversion. The results from an additional series of tubing reactor runs, which was carried out to determine how iron sulphide surface areas change during liquefaction, showed that the surface areas were drastically changed during the two-minute heat-up of the reactor. Robena pyrite with a surface area of 2.0 m² g^?1 and the iron-sulphur compound with a surface area of 80 m² g^?1 yielded iron sulphides with surface areas of 5.2 and 10.8 m² g^?1 after a two-minute heat-up to 425°C and subsequent one-minute quench.     

Liquefaction of coal by SRC-II process: Part I: A new kinetic model

The kinetic experiments were carried out in a continuous stirred tank reactor. Practically important ranges of SRC-II reactor temperature (444–466°C), pressure (10.4–20.8 MPa), nominal slurry residence time (0.54–1.62 h), coal concentration in the feed slurry (25–35 wt%), and inorganic mineral matter concentration (4.75–13.43 wt%) were covered in a total of 43 experimental runs. In each of the experimental runs, the feed slurry was formulated by using vacuum tower bottoms from SRC-II pilot plants using the same feed coal (Powhatan No. 5), to obtain feed compositions similar to those obtained in SRC-II pilot-plant recycle operation. The kinetic model considers the overall conversion to be achieved in two stages. The first stage is the instantaneous dissolution of coal and in the rate controlled second stage all the reactive organic components in the liquid phase are initially assumed to react, each yielding components lighter than itself. The distribution of products in each reaction stage is considered to be independent of the operating conditions. The best rate controlled (second stage) reaction scheme and values of the unknown parameters are obtained by minimizing the overall difference (i.e. for all the components over all the runs) between the measured and model predicted mass fractions of the various components in the reactor. This analysis identifies the reaction of solvent refined coal (pyridine soluble organic matter boiling above 482°C) to be the only significant reaction in the second stage and its rate is determined to be ^-r_SRC = 1.567 × 10⁵ exp (-79.16/RT) · p^0.28 · X_ASH, kg/L h. Overall error in this analysis yielding the reaction scheme, r_SRC and values of product distribution coefficients for both the reaction stages is less than 8% absolute i.e. ±4%.     

Kinetics of short contact time coal liquefaction. Part I: Effect of operating variables

The liquefaction kinetics of Powhatan No.5 mine coal (Pittsburgh Seam) in the presence of SRC-II recycle solvent at short contact times (<10 min) and temperature and pressure ranges of 573–723 K and 10.3–13.8 MPa is examined in a well-mixed reactor. In the initial stages of liquefaction, while overall coal conversion (tetrahydrofuran solubles) increases with temperature, oil (pentane solubles) is lost with an increase in temperature. An increase in solvent-to-coal ratio results in an increase of conversion. The initial coal particle size distribution, total pressure, and nature of gas phase (nitrogen or hydrogen) have no significant effect on the production of any of the product of liquefaction for contact times up to 10 min. A lumped kinetic model is presented to describe the product distribution.     

Mechanism of coal hydrogenation—liquefaction; Effect of temperature and coal particle size

The effect of coal particle size, hydrogen pressure and temperature on the extent of coal conversion in an entrained flow reactor is presented. Coal hydrogenation is done by feeding dry coal with ZnCl₂ catalyst into a continuous stream of hydrogen. The hydrogen-coal stream enters a long, small internal-diameter reactor (coiled tube reactor) controlled at about 500°C and 12.4 MPa hydrogen. At these conditions the coal particles become plastic and sticky. The hydrogen provides the energy to force the sticky coal particles through the reactor. Conversion of 85% of the coal to liquids and gases is easily attained. A physical mechanism is presented based on the unreacted-core-shrinking model. This mechanism aids in the explanation of the effect of process variables on reaction rates. Projections beyond the range of the variables studied are presented. These projections indicate that the pressure of coal liquefaction processes may be reduced by (1) the use of dry coal particles and (2) the reduction of the particle size. Significant reaction rates may be attained at pressures as low as 0.7 MPa by proper adjustment of particle size and temperature.     

Kinetics of thermal liquefaction of coal

A model is presented for the kinetic study of the thermal liquefaction of Belle Ayr subbituminous and Burning Star bituminous coals with anthracene oil, hydrogenated anthracene oil and hydrogenated phenanthrene. All experiments were performed in a continuous-feed, stirred tank reactor, at a temperature of 450 °C and a space time of approximately 5 to 55 min. A kinetic model which includes a reaction: coal + oil→more reactive coal, correlates the data reasonably well. This reaction explains the net consumption of anthracene oil during the initial stages of liquefaction. Such a reaction may account for a portion of the swelling of coal at low space times and the sizable increase of viscosity of reaction slurry during these initial stages of liquefaction. It is also observed that the yield of oil increases when solvents of increasing hydrogen donor capacity are used.     

Thermal decomposition and ignition of coal

A study of the thermal decomposition and ignition of coal as functions of pelletizing pressure and dwell time has revealed that: (1) ignition and thermal behaviour are related to the apparent density of the pelletized coal; (2) for a given apparent density of pelletized coal, the ignition temperature is related to the rate constants of thermal decomposition. Isothermal decomposition in air at 550 °C has been shown to fit the Avrami-Erofeev equation for three-dimensional growth of nuclei.