Environ. Scl. Technoi. 1903, 17, 714-777
Offgas Emissions from Simulated Modified In Situ Oil Shale Retorting Norman E. Hester,* Zoltan C. Mester,?and Yue G. Wangs Occidental Research Corporation, Irvine, Cailfornia 927 13 Experiments have been completed that were designed to determine the emission characteristics of modified in situ oil shale retorts. A series of pilot plant scale, oil shale retort tests using oil shale from Occidental's future commercial site have been carried out. The effects on emissions of the size of particle being retorted were investigated as well as the effects of using an air/steam mixture to support the combustion. The results demonstrate that hydrogen sulfide and ammonia are the dominant species of concern. These gases occur in the offgas in the thousands of a ppm range. Carbonyl sulfide and other organosulfur species were detected, but their concentrations were in the tens of a ppm range or lower. Measurements of arsenic and mercury levels in the offgas were also made. Arsenic and mercury levels averaged well below the ppm range and were not detectable in most samples. Introduction Oil shale in the western United States offers the potential of being a major source of domestic liquid fuels. In order to exploit the vast reserve of oil shale, Occidental Petroleum has developed the modified in situ (MIS) retorting technology to produce a crude oil from kerogencontaining rocks. The Occidental MIS process has been discussed in detail in a number of publications (1-3) and basically consists of mining a void in an area of rich shale, fracturing the shale with explosives, and then igniting the rock in place (in situ) to pyrolyze the kerogen in front of an advancing flame front. The combination of combustion, high-temperature reactions, and pyrolysis produce an offgas with a number of species that will need to be controlled. Occidental has proceeded from laboratory scale experiments through scale up to commercial-sized retorts. Experimental work on the full-sized retorta has taken place at Occidental's Logan Wash, CO, site. A considerable amount of both process and environmental data has been collected at this site; however, commercial development of the Oxy process is planned for federal lease tract Cb in Colorado. Early tests comparing the mineralogical and elemental analyses of Logan Wash oil shale w. Cb oil shales indicated that there were enough differences, particularly in sulfur levels, that significant problems might arise if Logan Wash emission data were used to estimate emission levels from Cb. It was the purpose of the experiments reported in this paper to examine in some detail the emissions from retorting of Cb shale so that proper decisions can be made regarding the size and type of emission control equipment. Experimental Procedures Simulated MIS retorting runs were done on Occidental's "Adiabatic Mini-Retort" which has been described in more detail in previous reports (4). Basically, the reactor consisted of a 15.2-cm (6-in.) diameter steel tube, 1.22 m (4 ft) long. Sixteen separate heating sections, each 7.6 cm (3 in.) high, maintained the reactor skin temperature equal to the center line temperature. Ideally, no heat was applied to the shale bed during retorting; only enough energy was supplied to balance heat losses. Typically, shale of less 'Present address: Union Oil Research Center, Brea, CA 92621. Simulation Sciences, Irvine, CA 92714.
4 Present address:
714 Envlron. Sci. Technol., Vol. 17, No. 12, 1983
than 4 cm (- 1.5 in.) in particle size was packed into the bed with thermocouples to monitor temperature and the movement of the retorting front. Ignition of the shale was done with a coil heater on the surface of the shale. A gas preheater was provided at the top of the mini-retort which also served as a steam generator. Gas flows were metered both into and out of the system. Part of the offgas was diverted through an oxygen meter. Oxygen content of the offgas was used to monitor bed ignition. Gas flow through the retort was maintained at 0.63 SCFM (3.5 L/min) per ft2of retort cross-sectional area, and unless otherwise indicated, runs have a 70%/30% air/steam ratio. Retort offgas samples were collected at 2-h intervals throughout the run in 500-mL gas sampling tubes and were analyzed by gas chromatographic techniques within the same day. Good reproducibility of analyses was obtained when the samples were analyzed within this few hour period. Trends in H2Swere followed by a Hewlett-Packard gas chromatograph equipped with a flame photometric detector on line during each run. Quantitation of the gas tube samples was done on a Tracor GC equipped with a electrolytic conductivity detector. The large H2S peak relative to minor organosulfur species made quantitation of these species extremely difficult; thus, mercaptans and sulfides were determined on a selected few of the gas tube samples by GC-MS. Ammonia was collected by absorbing it in a HC1 solution (pH
