Cover Story

Part 2

DISTILLATION:

Diagnosing Instabilities In the Column Overhead

Henry Z. Kister, Fluor James F. Litchfield, Consultant

roblems that involve the internals of a distillation tower have received a respectable amount of attention in the practical engineering literature.* But one cannot say the same for ones involving the tower overhead. This lack is surprising, because there is considerable potential for flawed design and for operating difficulties involving elevated equipment associated with distillation columns. Consider, for instance, the following three real-life examples. Although the specifics differ widely, these instances shared the following attributes: In each case, a minor flaw kept the equipment from achieving anywhere near its full potential A key to the correct diagnosis consisted of knowing the ultimate equipment capacity and recognizing that the equipment performance was well short of its capability Another key to correct diagnosis lay in performing a hydraulic analysis that combined basic calculations with field tests and observations In each case, the correct diagnosis led to simple, inexpensive modifications. Sledge hammers were not used (see box)

Understanding how the equipment works and keeping elementary physics in mind are the keys to eliminating effects such as siphons and unsteady vapor-collapse rates in reflux accumulators, condensers and other overhead components
TOWER FLAWS TYPIFY A MORE BASIC PROBLEM

n his preface to a 1997 book, A Working Guide to Process Equipment [1], author Norman Lieberman states: The general knowledge as to how process equipment really functions is disappearing from the process industries. This is not only my opinion, but the general view of senior technical managers, in many large corporations. Speculating about the root cause of this trend, Lieberman points that chemical engineers, the traditional guardians of process know-how, are spread thin over todays huge body of knowledge. This state of affairs dilutes their understanding of process equipment. Minor, seemingly unimportant design flaws slip by, later turning into major hidden bottlenecks that remain unresolved and limit production for decades. When the bottlenecks become entirely intolerable, a very expensive solution is often implemented: a sledge hammer is brought in to crack the nut. In one case involving distillation [2], a flashing feed entering a downcomer bottlenecked an entire olefins plant for 17 years. This bottleneck survived three failed fix attempts. The fourth, sledge hammer, attempt would have replaced all the column trays by packing. A last-minute investigation correctly diagnosed the problem and implemented a successful and inexpensive solution. The main text of this article presents three additional instances where minor flaws led to major bottlenecks, involving distillation-tower overhead systems. o

Case No. 1: AERATION DESTABILIZES REFLUX FLOW

In this facility, the reflux to a chemical-plant distillation tower flowed by gravity from a 30-in. (inside diameter) vertical accumulator through a vortex flowmeter and a flow control valve. The level in the accumulator was not automatically controlled.
*See, for instance, References [1] and [3].

During initial operation, the column reflux flowrate was very erratic. The operators dampened this behavior by keeping the control valve wide open and running the reflux accumulator at approximately 10% liquid level. Over time, however, the reflux flowrate dropped off, so the liquid level in the accumulator rose. The normal reflux flowrate was reestablished by stroking the control valve two or three times. However, this mode of operation destabilized the column and was an operating nuisance.

Figuring out the cause

The control valve was examined several times during brief outages. Each

time, the valve was found to be clean and in good condition. Then the fabrication drawings for the accumulator were reviewed. It was noted that at the 10%-liquid level, the entering feed dropped about 6 ft into a shallow pool of liquid at the bottom of the vessel (Figure 1). The liquid level was only about 18 to 20 in. above the liquid outlet nozzle. Descending like a waterfall, the liquid entrained vapor as it penetrated the shallow pool of liquid, creating very fine vapor bubbles. Some of these bubbles then became entrained in the discharging liquid, and a portion of those in turn became trapped downstream, at the control valve inlet.

56

CHEMICAL ENGINEERING WWW.CHE.COM SEPTEMBER 2004

FIGURE 1. A simple rerouting of pipe solved a problem, converting a waterfall into a modest fountain

FIGURE 2. The addition of a small siphon breaker drum to a seal loop eliminated recurring siphoning and erratic flow

These trapped bubbles limited the flowrate through the valve. The reflux pipe leaving the bottom of the accumulator had a diameter of 2 in. This size was generous for draining non-aerated liquid, but too small for liquid that was aerated. Draining the latter requires rundown lines that are sized for self-venting flow; in other words, flow in which liquid descends while any entrapped vapor bubbles disengage upward. An excellent correlation by Simpson [7] and Sewell [8] for self-venting flow is Figure 4.5 in Reference [3]. Based on that correlation, a 2-in. line can drain up to 7 gal/min of aerated liquid. Because the reflux flowrate in this case was 12 gal/min, the balance (5 gal/min) would build up in the reflux accumulator, raising the liquid level. At the same time, gas trapped at the valve would reduce the flow area through the valve and line, lowering the reflux flow rate. Stroking the valve vented the trapped bubbles and siphoned the accumulating liquid out of the accumulator. Plant operating personnel recalled that the accumulator had initially operated at levels much higher than 10%. At these higher levels, operation had been far more erratic. The explanation for that behavior is as follows: At those higher liquid levels, the waterfall height would diminish. This diminution and the greater pool depth would keep vapor bubbles from

reaching the accumulator outlet. The liquid at the bottom of the accumulator would degas, reverting to non-aerated liquid. In this condition, the liquid would easily siphon out, and the accumulator level would rapidly drop. But thereupon, the waterfall would again aerate the bottom liquid, and the aerated liquid flow would resume. The back-and-forth switches between aerated liquid flow and siphoning caused the initial erratic behavior.

The cure was simple

The accumulator happened to have an 8-in. hand-hole, located 15 in. above the bottom tangent line. The problem was solved by rerouting the 2-in. feed line so as to enter the drum by passing through the hand-hole cover, as shown in Figure 1. The rerouted line was configured in such a way that the portion inside the vessel extended to the drum centerline and then bent upwards, discharging upwards against a flat horizontal deflector baffle. This baffle redirected the incoming liquid, spreading it sideways. This arrangement eliminated the waterfall and aeration, and fully restored the stability of the reflux.

from a heavy liquid phase that then went to a distillation tower and other downstream equipment for final-product purification. The decanter was a horizontal drum, 4 ft in diameter by 8 ft long, that provided well in excess of 1 h of residence time for the phase separation. The maximum liquid level in the decanter was set by the 3-in. nozzle for light-phase drawoff, which was located in the decanter head, 6 in. below the top of the decanter. The heavy phase flowed through a block valve and an isolation control valve, and then upward through a seal loop. The elevation at the top of the seal loop was an inch or two lower than the elevation of the light phase draw nozzle. A 1-in. pressure-balance line connected the top of the seal loop to the decanter vapor space. After leaving the decanter, both phases flowed to their respective surge tanks at grade level, which was about 50 ft below the decanter elevation.

Problems
With the block and isolation valves wide open, the decanter proved to be susceptible to siphoning through the seal loop, creating erratic flow in this system. In fact, the seal loop had siphoned as much as 70% of the decanter liquid. Because of the erratic behavior, the decanter was unable to operate at its design temperature. Whats more, it
57

Case No. 2: SIPHONING IN DECANTER OUTLET PIPES

In this process unit, the decanter in Figure 2a separated a light liquid phase (for recycling back to a reactor)

CHEMICAL ENGINEERING WWW.CHE.COM SEPTEMBER 2004

FIGURE 3. In this example, the harmful pressure fluctuations could have been avoided in the first place by not ignoring the distillation-design literature

did a poor job of handling liquid surges from upstream heat exchangers.

Initial modifications
In the first try at solving these problems, the size of the pressure balance line was increased from 1 to 2 in. and that of the seal loop pipe from 2 to 4 in. These changes helped, but the erratic flow persisted. Operators were then able to lessen the siphoning by closing the seal-loop block valve halfway. But operation in this mode was not desirable, because whenever a feed surge occurred, some of the heavy phase was carried over into the light phase. Furthermore, closing the valve had little effect in helping the decanter to reach the design temperature.

Making a hydraulic balance

A hydraulic balance over the relevant equipment was made. The findings were illuminating. With the pressure-balance line doing its job, the static pressure PSL at the top of the seal loop would equal the static pressure PD in the decanter vapor space. The small, 1-to-2-in. elevation difference (hHP + hLP - hSL) between the liquid level in the decanter and in the top of the seal loop gave, as expected, enough driving force to overcome the friction head losses in the seal loop at normal flows. Calculations confirmed that the flow resistance through the seal loop, including the open valves, was extremely small. The 50-ft elevation drop from the top
58

of the seal loop to grade exerted strong suction at the top of that loop. With appropriate pressure balancing, enough vapor from the top of the decanter would have become entrained in the rundown line liquid to raise friction in the rundown line and make the pressure at the top of the seal loop the same as that in the vapor space of the decanter. But if there were no pressure balancing whatsoever, the suction at the top of the seal loop would have caused the liquid flow to rapidly increase and siphon out the decanter. The observation that siphoning was taking place meant that the pressurebalance line was not fully achieving its intended function. Increasing the line size from 1 to 2 in. had helped, but had not remedied the situation completely. The throttling of the liquid valve between the decanter and seal loop increased the pressure difference between the two, which dampened surging by further increasing vapor flow to the seal loop. However, the hydraulic balance had made it apparent that better vapor balancing and siphon breaking were required.

mize turbulence and short-circuiting in the decanter, the internal feed pipe in that vessel was increased in diameter from 2 to 4 in., and the feed was discharged against the head, as seen in Figure 2b. And for better control of the light-phase thickness, a weir was installed in the light phase just upstream of the outlet nozzle. As a result of these modifications, the siphoning was eliminated and the decanter outlet flow was no longer erratic. Most important, stabilization of the flow allowed the decanter to operate at its design temperature, which was 10 to 20F lower than the pre-modification temperature, leading to major improvement in phase separation.

Case No. 3: POOR HOOKUP OF HOT-VAPOR BYPASS PIPES

This installation consisted of the overhead for a new debutanizer column separating C3 and C4 hydrocarbons from gasoline. The column overhead vapor was totally condensed by a battery of four submerged condensers (Figure 3a). The reflux drum was elevated. Noncondensables, if any, from the condensers were vented to the drum using 1-in. vent lines (not shown). The tower pressure was intended to be controlled by a hot-vapor bypass, hooked up as shown in that figure. But when the tower was put into service, it experienced severe pressure fluctuations. Maintaining a constant column pressure was impossible, which bottlenecked the tower throughput.

The full cure

The seal loop was replaced by a 1-by4-ft vertical drum (Figure 2b) that provided good siphon breaking and pressure equalization with the decanter. The siphoning was broken by drawing the heavy phase from a side sump, into which liquid could only enter by overflowing a chordal weir. A 3-in. line was installed to balance the pressures between the top of the drum and the decanter vapor space. Two other changes improved the decanter operation further. To mini-

Diagnosing the problem

For successful pressure control by a hot-vapor bypass, correct piping is

CHEMICAL ENGINEERING WWW.CHE.COM SEPTEMBER 2004

Authors
Henry Z. Kister, a senior fellow and director of fractionation technology at Fluor Corp. (Aliso Viejo, Calif.; Phone: 949349-4679; e-mail: Henry.Kister @fluor.com), has over 25 years experience in design, troubleshooting, revamping, field consulting, control and startup of fractionation processes and equipment. Previously, he was Brown & Roots staff consultant on fractionation, and worked for ICI Australia and Fractionation Research Inc. (FRI). Author of the textbooks Distillation Design and Distillation Operation, plus 70 published articles, he has taught the IChemE-sponsored Practical Distillation Technology course 260 times. A recipient of Chemical Engineerings 2002 Award for Personal Achievement in Chemical Engineering, he is also a member of that magazines Editorial Advisory Board. He holds B.E. and M.E. degrees from the University of NSW in Australia. A Fellow of IChemE and a member of AIChE, he serves on the FRI Technical Advisory and Design Practices Committees. James F. Litchfield, a consultant residing in Ventura, Calif. (email: [email protected]), has 37 years of experience. He was with CFBraun/Brown & Root Braun for 34 years, including 15 in research and the last two as a staff consultant. For over a decade, he was the firms technical representative to the Particulate Solids Research Institute. His expertise is in chemical engineering design and multiphase flow; he has optimized or upgraded numerous phase separators (for two or three phases), inlets to vacuum and atmospheric columns, and lines for two- or three-phase flow. He holds a B.S. from the University of California at Berkeley, and an M.S. from the University of Idaho, both in chemical engineering.

mandatory. Bypass vapor must enter the vapor space of the reflux drum (Figure 3b), the bypass line should be free of pockets where liquid can accumulate, and any horizontal pipe runs should drain into the reflux drum. Most important, liquid from the condenser(s) must enter the reflux drum well below the liquid surface. The bottom of the drum is the most suitable location, but instead, extending the liquid line downward to near the bottom of the drum (Figure 3b) is an acceptable alternative. These principles and recommendations were first published almost fifty years ago [4, 5]. Since then, they have been strongly endorsed by key recent sources addressing methods for distillation-column pressure control [3, 6]. The initial piping design for the debutanizer (Figure 3a) defies those principles, and violates the recommended practices for hot-vapor bypass piping. Subcooled liquid mixes with vapor at its dewpoint; the vapor collapses at the point of mixing; the rate of vapor collapse varies with changes in subcooling, overhead temperature, and condensation rate. It is the variations in this collapse rate that induce the pressure fluctuations experienced, and the consequent control-valve hunting. Problems similar to this were repeatedly described in the early literature mentioned above [4, 5]. But strangely enough, the incorrect hookup that is shown in Figure 3a keeps reappearing in modern distillation designs.

which contained hot vapor, and the bottom part, which contained subcooled liquid, simply by (cautiously) n touching the drum.
Edited by Nicholas P. Chopey

References
1. Lieberman, N. P., and Lieberman, E. T., A Working Guide to Process Equipment, McGraw-Hill, New York, 1997. 2. Kister, H. Z., Hower, T. C., Freitas, P. R. de M., and Nery Souza Neto, J. Problems and Solutions in Demethanizers and C2 Splitters with Interreboilers, 8th Annual Ethylene Producers Conference, New Orleans, La., February 2529, 1966. 3. Kister, H.Z., Distillation Operation, McGraw-Hill, 1990. 4. Whistler, A.M., Locate Condensers at Ground Level, Pet. Ref. 33 (3) , p. 173, 1954. 5. Hollander, L., Pressure Control of Light-Ends Fractionators, ISA J. 4 (5), p. 185, 1957. 6. Chin, T. G., Guide to Distillation Pressure Control Methods, Hydrocarbon Proc., 58 (10), p. 145, 1979. 7. Simpson, L.L., Sizing Piping for Process Plants, Chem. Eng., p. 192, June 17, 1968. 8. Sewell, A., Practical Aspects of Distillation Column Design, Chem. Engineer, 299/300, p. 442, 1975.

Switching to the sound design

To remedy the situation at the debutanizer, the liquid and vapor lines were separated. The vapor line was modified so that it introduced the vapor into the top of the reflux drum. As for the liquid line, it was extended downward, terminating a few inches above the bottom of the reflux drum. Figure 3b shows the modified system. These changes fully solved the problem: the tower pressure no longer fluctuated. Whats more, one could feel the differences in temperature between the top part of the reflux drum,
An earlier version of this article was presented at the 2003 annual meeting of AIChE, in San Francisco last November.)

adlinks.che.com/3646-37
CHEMICAL ENGINEERING WWW.CHE.COM SEPTEMBER 2004

59