CONDITIONS OF OPERATION DURING COKING WHICH LEADS TO HIGHER YIELDS OF VALUABLE LIQUID PRODUCTS

A 2-inch diameter, fluid bed coker using a 21 cm bed of Syncrude fluid coke was adapted to feed bitumen, bitumen-in-water emulsions plus steam and nitrogen. Ini tially feeding was through a 1/8-inch tube incorporated into a steam nozzle. This was l ater modified to provide feeding of the bitumen plus steam through a side port nozzle at various depths below the surface of the fluid bed. The reaction system was const ructed for once-through operation without recycle of heavy ends. Product liquids were a nalyzed for microcarbon residue (MCR) and so that calculations could be made to determin e the extent of "by passing" of the feed on its passage through the reactor. As a basis for comparison to determine progress being made, a series of eight runs were made under a set of standard conditions. Bitumen feed rates employed r anged from 1.0 to 3.9 g/minute. Steam provided for fluidization and atomization varied from 6.3 to 9.1 g/min. and nitrogen flows were 3.4 L/min. Reaction temperatures ranged fr om 485 to 540 .degree.C and calculated residence times from 0.50 to 0.64 seconds. Real liquid product yields ranged from 54.6 to 67.3% averaging 61.6%. (ii) Bitumen-in-water emulsions were employed as feed to supply the bitumen as small particles (60 .mu.m) to determine the effects on real liquid product yield. Oper ating conditions were similar to those described for the standard runs. The real liqui d product yields surprisingly were significantly higher ranging from 64.0 to 73.8% averagi ng 69.6%. The cost and availability of the surfactant to prepare this emulsion on a commer cial scale ruled out the possibility of using bitumen-in-water emulsions as feedstocks. To simulate the small particle sizes observed in the bitumen-in-water emulsions and achieve higher real liquid product yields, hypodermic syringe needles (No. 1 7 and 18 gauge) were built into nozzles which were introduced through a side port to inje ct feed 5.5 cm below the surface of the fluid bed. With the No. 18 gauge needle and cond itions similar to those in the standard runs, real liquid product yields ranging from 8 3.9 to 86.1% averaging 85.0% were obtained. Using the No. 17 gauge needle and feeding bitumen under similar conditions, real liquid product yields from 77.3 to 80.2%, averagi ng 79.0% were obtained demonstrating that particle size of the bitumen fed plays a major role in determining the real liquid product yield. It was found that when the syringe needle nozzle was located further belo