Misting 101: � How to maximize shaft retorting yields 30th Oil Shale Symposium October 17-20, 2010 Larry M. Southwick, P.E. Cincinnati, Ohio
Outline � Introduction Gas Combustion Process Misting Benefits and detractions Conversion to “ hard driving ” Other units The solution
Gas Combustion Process � US Congress in 1944 enacted Synthetic Fuels Act authorizing construction of demo plants Oil Shale Experiment Station at Rifle, Colorado Retorting processes studied depended on method of heat application: Thru wall - Pumperston Combustion in retort - Gas Combustion Heated gases or liquids - Royster Hot solids - TOSCO USBM Bulletin 635 reported on Gas Combustion Other studies by oil companies (6 Company, 17 Co)
Gas Combustion � Crushed and sized shale Rising hot gases retort shale and vaporize oil Carbonaceous residue is burned in combustion zone Oil product removed from gas, which is recycled back to retort Process works efficiently because of mist formed in product cooling zone
Oil Misting Basics � “ From the onset of experimental work, it was observed that the gas streams from the retorts usually contained shale-oil mist ” (not droplets, but fine mist) “ This fundamentally new concept led to the development of Gas Combustion Process ” “ If the oil is to leave the retort as a mist in the offgas stream, the droplets must be formed in the spaces between the shale particles and must be small enough so that inertial separation does not occur ” “ A refluxing problem occurs when the amount of oil on the shale is great enough to drip or flow down through the bed of shale ” Entrainment of droplets off of shale does not occur here because gas velocity is too low - thus oil collected on shale will descend with the bed = REFLUXING
Misting Section � Mist forms just above retorting zone Retort operates as a countercurrent heat exchanger No sharp demarcation between retorting and product cooling Assume 700°F shale temperature as dividing point
Refluxing � Refluxing of condensed mist causes oil cracking Alters heat distribution in misting section due to revaporization and secondary cracking Equilibrium is stable under refluxing and not-refluxing Often depends upon conditions at start of run Cracking produces lighter, less viscous oil, but loss of production is severe
Mist Formation � Mist is formed if: Oil vapor cools until gas becomes saturated Nucleation occurs Supersaturation, S, favors mist formation S is oil partial pressure in gas divided by its vapor pressure at shale temperature Mist forms when heat transfer to shale exceeds mass transfer of oil to shale Mass transfer depends on diffusion and on impaction Nucleation sites help form mists
Mist Dynamics � Mist flooding rate obtained by drawing mist from retort and feeding to external bed Raise cooling rate by lowering temp. of bed until refluxing When MassMeanDia = 2.5 µ Oil rate is 8-10 lb oil/MSCF Flooding mist MMD = 3.0 µ 5-6 lb oil/MSCF Thus lb/MSCF 3-4 lost Thus there is a maximum carrying capacity to gas Flooding vel ¼ x 1 ” = 2.7 ft/sec For 1 ” x 3 ” = 3.3 ft/sec
Mist Measurements � Refluxing caused by collision between mist and bed particle was incomplete explanation of refluxing – also unstable mist, mist growth and coagulation Mist impactor is standard test Stages, 16, 8, 4, 2, 1, 0.5 µ Considerable (50%) collected in piping and elbows off retort High dilution gas = oil loss from gas carrying capacity Collection efficiency increases Mist particle size goes up Gas velocity increases Mist loading increases Small shale particles High bed packing fraction (wide particle size range)
Mist Profile � Distance above air inlet, ft. 8 6 4 2 Droplet, mass mean dia., µ 2.36 2.28 1.82 Plugged with fines Loading, lb oil/MSCF 9.36 8.04 5.61 Temperature, ºF 140 300 470 800 Once nuclei occurs, no new nuclei form Mass balance confirms growth since larger diameter = more oil per particle = loading rate So oil is growing on existing nuclei Tests using injected nuclei did not resolve
Removal of Liquid � Refluxing liquid would cause accretions to form just above air distributor, blocking retort operation Use drawoff systems to collect refluxing liquid Worked well on small lab retorts, 1 ” , 2 ” , and 3.6 ” Variable results when applied to 150 TPD retort Drew off at zone where shale temperature is 600 ºF Two other options to eliminate refluxing Draw off unmisted hot, dry gas Draw off hot misted gas but above refluxing zone These point the way to “ hard driving ” of retort
Challenges � Minimize losses from impaction of mist on shale particles - ergo no small particles Maximize evolution of oil - ergo, the smaller the particles the faster the net retorting rate Testing found that particles as small as 1/8 inch could be used, but particles smaller than that caused significant oil yield losses Limiting the minimum size to greater than 1/8 inch provided no great advantage But retort still limited by oil refluxing, not easily controlled nor readily amenable to design The challenge was also what to do with the fines from crushing - ergo TOSCO process
Hard Driving Paraho � Capacity of iron ore blast furnaces were increased 150 years ago by “ hard driving ” Hard driving meant just feeding more and more ore They found blast furnace could handle ~30X feed So if remove shale retort bottleneck of oil refluxing, should be able to “ hard drive ” Thus eliminate mist formation or oil condensation 6 Company (1966) solution was an oil drawoff pan, which did not work (a typical “ boiling oil ” solution) Rather try one of the other two solutions not picked (pull off oil before it cools enough to begin refluxing) The oil would be cooled and condensed externally to retort in equipment similar to that used before This now oil-free gas can be reheated, re-injected above pull-off, and provide mist-free shale heating
Demo Plant Example � One shaft retort converted to this concept to make it operable Original Hytort scheme ran under conditions where normal mists did not form (high pressure, H 2 gas, small particles), and which enhanced condensation of oil vapor on solids Thus extract fumes before they condense, inoperable otherwise Scheme studied in cold flow model, had good zone isolation Hot tests were always just with retort zone, which worked well
Union B Retort � Spent shale gasifier had moving bed, rising vapors Hot shale still evolving gases - cools and condenses Mist or otherwise, refluxing became a problem
Zinc Fuming/Misting � Zinc ore (oxide) reduced and volatilized from retort Zinc metal fume will condense upon contacting cold downward flowing solids These vertical shaft retorts extracted hot fume or mist Process on left used splash condenser, right had labyrinth
Imperial Smelting � Shaft furnace, briquette feed (carbon + zinc oxide) Keep top of shaft hot (1000°C), so no mist forms Splash condensers inefficient, use four in series
Processing EAF Dust � Steel made in electric arc furnaces (EAF) by melting scrap Volatilizes zinc from galvanized steel, dust is hazardous EAF dust processed by heating to remove zinc Shaft furnace has similar zinc condensing problems
The Problem � Nature of oil shale retorting leads to shaft retorts Counterflow of oil vapor and cold solids leads to formation of very fine oil mists Mists can lead to refluxing of oil, net yields suffer Retort operation also suffers - accretions, flow blockage, channeling of gas and shale Fines WILL lead to oil losses, lighter oil and more gas and more coke Low top temperature can also cause yield losses The bottleneck to capacity is oil refluxing down the retort The higher the shale rate, the more likely refluxing will occur, which sets and limits the feed rate
The Solution - DryTop � Eliminate the oil refluxing bottleneck by removing mist or hot vapor before oil condenses onto shale Collect oil in devices similar to WetTop operation, then reheat and re-inject gases above drawoff Blast furnaces, zinc retorting and distillation, EAF dust processing, even modified Hytort concept provide examples of hard-driving operation Further, if shale feed is wet, the heat required to vaporize the water can be supplied by the re-injected gas, eliminating high temperatures in the retorting zone Shale feeding and withdrawal devices may have to be modified for the greater throughput
Recommend
More recommend
