5—Flow Pattern in Vacuum Distillation 
\ 
o000000000 
. 
~~ 
ol ololole.o.olele%6 
pS SE 
0000000000 
“Bubble tray 
Tracing Flow Patterns 
Petroleum-stock movement in refining operations fre- 
quently follows a complex pattern. Sometimes relative 
partition of a stream among two or more paths may be a 
matter of some technical or economic importance about 
which quantitative information is difficult to obtain. In 
this situation, radiotracers can be used advantageously to 
gather the information. 
Flow distribution in a large vacuum-distillation column 
(9) was investigated as shown in Fig. 5. Each distillation 
tray is divided by baffles into eight channels. Overhead 
stream color, plus the presence of heavy metals, raised the 
suspicion that bubble caps in some trays were not flooded, 
permitting unwashed vapor to carry entrained liquid up 
the column. Such partial flooding could be caused by 
nonuniform liquid flow over the trays. Flow uniformity 
was tested during operation by injecting radioactive 
Tracing Tar Entrainment 
In contrast to the wide and varied radiotracer applica- 
tions to operations that involve stock movement or mixing 
(see the section on ‘‘ Determining Mixing” on next page), 
160 
triphenylstibine into the column. After the injection the 
response of eight G-M counters was observed simultane- 
ously. Each was attached outside the column opposite 
one of the eight channels on a single tray. Four of the 
counters, on one side of the column, showed a strong re- 
sponse as the tracer flowed down; the other four, on the 
opposite side, showed little or no response. The conclusion 
was that liquid was flowing over the trays on only one side, 
leaving the bubble caps on the other side open to pass 
unscrubbed vapor up the column. 
Solid flow pattern. Tracer applications in catalytic- 
cracking plants have yielded valuable operating data that 
could not be obtained in any other way. Catalyst-bead 
flow patterns have been determined with radioactive beads 
(7). Effectiveness of an intricate baffle system in pro- 
moting uniform flow through a large sloping pipe at the 
bottom of a furnace in one cracker (Fig. 3) was a disputed 
subject. 
The controversy finally was settled by placing a pair of 
G-M counters, one at each end of the pipe, successively on 
the top, on the bottom, and on each side of the pipe and 
observing the speed and relative number of radioactive 
beads that followed these paths. The great difference in 
speed and number of transits between the different paths 
showed that the baffles did not serve their intended 
function. 
In the same plant, relative load carried by the gas lift 
through three parallel pipes was measured by attaching 
the counters in turn to each of the pipes and observing the 
number of radioactive bead transits for several hours. 
In a large regenerating furnace in a similar plant, tracer 
beads were introduced in different paths near the periphery. 
Differences in bead flow rate down the furnace sides were 
noted and used as a basis for adjusting bottom orifice plates 
to obtain more uniform flow. 
only a few applications have been made to operations in 
which petroleum-stock components are separated. One 
study (Fig. 6) using this kind of tracer application was of 
tar entrainment in evaporator and distillation-column 
overhead streams. Tar cannot be tolerated in refined 
products; also, it must be excluded from furnace-tube feed 
to reduce coke formation and increase the plant oper- 
ating time. , 
A plant-scale test was made of tar entrainment in an 
evaporator in a thermal-cracking plant (7). Co®-naph- 
thenate, as a tar tracer, was injected in a short surge into 
the evaporator feed. A counter attached to the pipe 
carrying the still bottoms to the tar tanks recorded the 
activity of the bulk of the radiocobalt. Another counter 
attached to the overhead stream from the evaporator 
showed the cobalt activity in that stream. By integrating 
the counts on both streams and making corrections for the 
different flow rates, a quantitative measure was obtained 
of how much tar was going into the distillate. Values 
ranging from zero up to a few per cent, depending on 
operating conditions, were found. 
