Reactions and Separations
Can We Believe the Simulation Results? Be careful of these key issues that may generate differences between a distillation-tower computer simulation and its actual performance. Simulations do not always square with nuts-and-bolts reality.
Henry Z. Kister, Fluor Daniel
P
revious surveys ( 1 1,, 2) collected case histories of tower malfunctions from the open literature. Many reports described simulations that did not reflect what a tower was actually actually doing. doing. Often, Often, the problem problem was was with the simulati simulation. on. Sometime Sometimes, s, the problem problem was that the tower did something something unexpected: unexpected: the simulation was actually actually correct, based on the the data fed to it. Finally, Finally, there were were instances instances where where both mishaps took place — the simulation had some serious problems, but there there were aspects of tower tower behavior behavior that were not fully understood initially and the simulation helped to explain. This article focuses on instances where problems were found in the simulation or where the simulation was instrumental in identifying a previously misunderstood problem (see table). The cases were extracted from surveys in Refs. 1 and 2 and have been updated with some recently reported cases. The original numbering for each case has been retained, retained, so that the reader reader may easily locate the cases in the previously referred to articles. The number of examples presented here is by no means a large enough sample for performing a statistical analysis of the main problems in troubleshooting distillation simulations. Nonetheless, the cases provide guidance on what to look for when troubleshooting a distillation simulation — and what to watch out for when carrying out the next simulation.
52
w w w .cepm agazine.org
O ctober 200 2
C EP
Three major findings This survey revealed three major issues that require attention in using simulations: 1. correctly predicting vapor/liquid equilibrium (VLE) 2. having the simulation match plant data 3. applying graphical techniques to troubleshoot simulations. These three issues are present in about two-thirds of the reported cases. In about 20% of the remaining cases, the process chemistry chemistry and hardware hardware efficienc efficiency y did not match what was true in an actual tower. Other items — correctly correctly modeling modeling feeds, feeds, obtaining obtaining the true vapor and liquid loads, reliably predicting the hydraulic hydraulic behavior, and finding and squelching bugs in the simulations — were found to be problematic, but to a lesser degree.
Problems with VLE data and predictions Most case studies falling into this category involve close-boiling components. The problems can be with two chemicals of similar vapor pressure (e.g., hydroca hydrocarbo rbons) ns),, or due to a non-ideality that pushes the volatilities of a pair close to a pinch. Correctly estimating non-idealities is another trouble-spot when it comes to VLE predictions. A third dilemma is characterizing heavy components in crude-oil distillation. This is a key problem in simulating refinery vacuum towers. Few reports were made regarding other situations. It seems that VLE prediction for pairs of components components that have reasonably reasonably high volatilities, volatilities, for example, methanol/ethanol, methanol/ethanol, is not not often often troublesome. troublesome.
Case No. -Ref.
Type of Column
Brief Description Section 1. How Good Are Your VLE Predicti ons?
102-3
Acetylene solvent/water stripper
Solvent losses were far greater than design. Unsuccessful extrapolation of VLE data was one of the causes. Increasing number of trays and raising reflux helped reduce losses.
129-4
Chemical AMS (alpha methylstyrene)phenol
Column pressure was lowered from 100 to 30 mm Hg to improve separation, and valve trays were replaced by screen trays to match the capacity. Separation did not improve. Extrapolating VLE from 100 to 30 mm Hg gave optimistic expectations that did not materialize.
118-5
Chemical super fractionator
A laboratory error gave incorrect VLE data based on which a tower with 200 theoretical stages was built where over 300 stages were required. With the 200 stages, product purity could not be achieved. The plant was forced to rerun the purified material a second time through the tower, effectively halving plant capacity.
140-6
Butadiene
1,2-butadiene is less volatile than 1,3-butadiene and leaves mostly through the bottom, but a commercial simulator predicted it would leave through the top. Problem was due to incorrect critical constants used in the equation of state.
109-7
Trichloroethylene (TCE)/carbon tetrachloride (CTC)
The concentration of TCE in CTC was higher than expected. A total reflux test showed that separation near the column bottom was worse than expected. Either VLE nonideality or decomposition of chlorinated ethanes at the reboiler temperature was the culprit.
123-8
Water/dichloromethane (DCM)
DCM concentration in the wastewater was very low. Based on ideal behavior, the vapor vent to atmosphere from the wastewater storage tank would have contained little DCM. Measured DCM at the vent was 27 mole %.
137-9
N -heptane/
Use of fractional-composition data from batch distillation showed that the popular VLE choice for this system gave poorer simulation of plant data than alternative VLE procedures.
toluene test system 314, 315
VLE error leads to mismatch between plant data and simulation, see Section 2.
122-2
A 2% difference in relative volatility in a low-relative-volatility (≈1.1) system accounted for a difference of 50% in the tray efficiency. The designer’s efficiency worked only with the designer’s volatility; the operator’s efficiency worked only with the operator’s volatility.
124-10
Ethylbenzene/ styrene
Identical column simulations using SIMSCI, HYSIM and ASPEN, all employing Soave-Redlich-Kwong (SRK) VLE, calculated entirely different product purities. Reason was small differences in the critical temperature and pressure and in the acentric factor for styrene.
117-11
Refinery vacuum
In five deep-cut towers, the design wash-oil flowrate was too small, leading to drying, coking, high-pressure drop, loss in gas quality and short runlength. The drying resulted from simulations that underestimatedthe fraction of wash oil vaporized. In all cases, inaccurate boiling point characterization of the heavy fractions of the crude led to these underestimates.
130-12
Refinery vacuum
Inaccurate TBP characterization of the heavy fractions of the crude led to a wash-oil flowrate that was too small to prevent coking in a deep-cut wash bed. Coke plugged the level bridle and draw nozzle on the slop-wax collector tray. Unable to drain, slop wax was re-entrained into the wash bed.
These findings pretty much match the author’s experience. The largest VLE problems seen are with the distillation of close-boilers, whether due to their vapor pressures or their non-idealities. Many problems have surfaced with VLE predictions for non-ideal pairs. In petroleum refining, characterization of the vacuum tower feed and bottom is the
most challenging job. On the other hand, we have seen a comparatively small number of problems in VLE predictions of medium- and highvolatility ideal systems.
Matching plant data A major problem appears to be obtaining a reliable, consistent set of plant data. It can be difficult to get correct numbers from flowmeters and proper data from laboratory analyses. The data may need to be checked and rechecked. The troubleshooter’s prime tool is compiling mass, component and energy balances, and checking the results from laboratory analyses to catch the lying flowmeter. A prime area where mismatches between simulated values and plant data have been reported is in chemical operations with two liquid phases. A close comparison of the temperature profile derived from a simulation with that measured in the tower is an excellent tool for establishing the presence or absence of a second liquid phase in a tower. For instance, a simulation may not predict that two phases will be present, while the measured temperatures may show that there are, in fact, two phases. Another hurdle is in simulating refinery vacuum towers. The most troublesome prediction to make is how much entrainment of liquid from the flash zone reaches the slop-wax draw and becomes part of the measured draw (overflash). This measurement is difficult, and is often established by a component balance. When this value is not measured, or measured incorrectly, a simulation can yield an incorrect reflux rate, which, when used in debottlenecking, can cause the wash bed to dry and coke up.
Graphical techniques Three key graphical techniques that can shed light on what the columns and simulations are doing are the McCabe-Thiele and Hengstebeck diagrams, multicomponentdistillation composition profiles, and, in azeotropic systems, residue-curve maps. These methods permit visualiza-
C EP
O ctober 200 2
w w w .cepm agazine.org
53
Reactions and Separations
Case No.-Ref.
Type of Column
Brief Description Section 2. Does Your Distillation Simulation Ma tch Plant Data? General
315-13
Aromatics
311-14, 16 Olefins demethamizer
Diagnosis based on the initial simulation was control instability. A simulation reality check against plant data exposed needs for better energy-balance data, a low reflux test, VLE review for one pair, and a surface-temperature survey. Once modified to adequately reflect plant data, the simulation pointed to a problem of unexpectedly low tray efficiency. The corrective action for a forthcoming revamp became improving trays, not controls. Two different simulation models matched plant data well. Both suggested efficient packing in the lower sections. One model suggested efficient, the other inefficient packings in the upper sections. Plant logs of the temperature-reflux dependence proved that the model predicting poor upper efficiency was correct. A revamp based on this model succeeded; had the high efficiency model been used, the revamp would not have met is goals.
312-15
Olefins water quench
A simulation based on a set of tower readings led to theories for explaining liquid carryover from the top. A detailed test invalidated the simulation and theories. A discrepancy between data and simulation, initially attributed to an incorrect temperature measurement, was proven in the tests to be due to error in flow measurement. This completely changed the explanation for the carryover.
314-16
Stabilizer
To develop a simulation for revamp, column was tested at high and low reflux. Low reflux data matched the simulation well, high reflux data matched poorly. A Hengstebeck diagram explained the mismatch in terms of a VLE inaccuracy.
329-17
Refinery, depentanizer
Basing a simulation on ASTM D86 gave optimistic tray efficiency and a misleading simulation. Matching simulated-to-measured bottom component analysis gave correct tray efficiency and good simulation. With Second Liquid Phase
1426-18
Chemical solvent dehydration by azeotropic distillation
Replacement of tower by a larger one caused instability, reduced capacity and yielded high solvent losses. The reason was refluxing of water in the cyclohexane entrainer. An undersized condensate line caused condensate buildup in the condenser all the way to its mid-point vent, from where it drained directly into the cyclohexane side of the decanter. Matching simulated temperature profile to plant data revealed excess water in the hydrocarbon reflux, pointing to a decanter malfunction.
1279-18
Chemical monomer and water separation from acid 15 ft I.D.
The column operated normally at close to maximum capacity until it suddenly became unstable at high rates. Cause was a crack in the decanter baffle plate that allowed water into the refluxed organic phase. The reflux water generated second liquid phase and, therefore, temperature instability on the trays. Matching simulated temperature profile to plant data revealed excess water in the organic reflux, pointing to a decanter malfunction.
Chemical acid recovery from organics, packed tower
Acid recovery was poor due to a malfunctioning collector that collected liquid from an internal condenser, splitting it into a heavyphase reflux and a light-phase acid product. The phase-separation overflow weir in the collector was too tall, and was installed in an incorrect location, and the product draw-nozzle was undersized. All led to poor decanting. Once these were corrected, the tower operated normally. Matching simulated temperature profile to plant data showed the presence of two liquid phases where only one was expected, pointing to a malfunction in split between the heavy-phase reflux and light-phase acid product.
877 -18
54
w w w .cepm agazine.org
O ctober 200 2
C EP
tion of the simulation and allow the engineer to use his or her judgment in interpreting the results. These diagrams are not drawn from scratch when troubleshooting simulations. They are simple plots of the composition profiles generated by the simulation. The methods for plotting these diagrams for distillation troubleshooting are discussed elsewhere (16, 26).
Incorrect or puzzling chemistry In chemical towers, reactions such as decomposition, polymerization or hydrolysis are often unaccounted for in a simulation. Or, sometimes, a component that is believed to be present in one chemical form turns out to be in another. In either case, the resulting separation will differ from what the computer simulates. When dealing with unstable chemicals, such as some nitro compounds, this can lead to exothermic decompositions and explosions. There are also cases in which the chemistry of a process is not well understood. One of the best ways to get a good simulation in these situations is to first run the chemicals through a mini-plant, as recommended by Ruffert (27). In quite a few cases of reactive systems in which pilot work was carried out by a client, undertaking a through engineering-based understanding of the chemistry led to redesign of a process that would otherwise would not have worked well into one that was trouble-free.
Inefficient efficiency estimates The table reveals no clear trend regarding estimating efficiency. The author has found that in established processes, such as the separation of benzene from toluene, ethanol from water or ethane from propane, estimating the efficiency is quite trouble-free for conventional trays and packings. Problems arise when deviating from these. For instance, in a
first-of-a-kind process, efficiency prediction can be an issue. This is tackled by running pilot tests, if schedules permit and the economics favor, or by gross oversizing, if the column costs are low or if the schedule is tight. Likewise, efficiency predictions run into trouble when engineers are on the steep portion of the learning curve for a new device.
Case No.-Ref.
Type of Column
Brief Description Section 2. Does Your Distillation Simulation Ma tch Plant Data? (Continued) Refinery Vacuum Tower Wash Se ctions
218-19
Refinery vacuum
The design wash-oil flow was too small, leading to coking of the wash bed. This resulted from a single-tower simulation model predicting low-wash dryout ratios. Segmenting the simulation model into a number of flash units with recycles gave the correct dryout ratio, requiring triple the previous wash rate. The revised simulation model correctly predicted plant data.
318-20
Refinery vacuum
Following replacement of trays by grid in the wash zone, the column experienced chronic coking leading to high-pressure drop, reduced gas oil yield, and high metals content of gas oil. The design allowed for little vaporization in the wash bed. In reality, all the wash oil supplied vaporized and the bed dried up. Problem solved by redesigning the spray header for 3– 4 times the original wash rate.
320-21
Refinery vacuum
Wash-bed coked within 6 mo following replacement of low-efficiency by high-efficiency packings. Asphaltene balance showed that practically all the overflash was entrainment. The efficient packing vaporized more wash liquid, causing unwetting and coking in the lower packings.
Feed entry A correct representation of the feed inlet is crucial if the number of stages between the feed and the first drawoff is small, especially if it is only one or two. One of the two cases reported here happened in a refinery vacuum tower in which the first major product exited the tower between 0.5 and 2 stages above the feed. The other happened with sponge oil returning to the main fractionator in a refinery catalytic-cracking unit immediately above a product draw. The feed entry issue is not unique to refining. It can be more severe in chemical towers, especially if some of the chemicals react in the vapor phase and not in the liquid state. Entry of the feed into the vapor space may give completely different results than entry onto the tray or downcomer liquid.
Have You Used Graphical Techniques to Troubleshoot Simulations? 313-16, 22
Olefins C2 splitter
Addition of an interreboiler caused column design to approach a pinch. Pinch was undetected by a simulation; the simulation converged and worked well. Pinch detected by a McCabe-Thiele diagram.
204-23
Refinery debutanizer
The column feed was rich (72%) in butane. A few degrees of extra preheat caused a large increase in feed vaporization, accompanied by a large drop in stripping vapor rate. This increased butane in bottom product. The problem was solved by controlling the flow of steam to the preheater.
317-24
Refinery alky Deisobutanizer, 8-ft I.D ., valve trays
Isobutane in bottom was 3– 4 times the design, and column capacity was restricted. Excessive subcooling of feed overloaded the bottom section. Also, a few trays above the feed had low liquid rates and could have been blowing. Several other problems were identified. Problem solved by adding a feed preheater, installing anti-jump baffles in the lower trays, and adding picket-fence weirs to the low-liquid-rate trays.
1559-5
Chemical
Feed to a two-column train fluctuated. Top purity of the first tower was held constant, which concentrated the disturbances in the feed to the second tower, destabilizing this tower. To overcome, a surge drum was added with 5-7 days residence time. Heat losses from the drum caused a 50°F drop in feed temperature to the second column, aggravating its reboiler limitation, which, in turn, necessitated a reduction in reflux ratio and, therefore, a lower purity of the top product.
210-16, 25
Chemical
A vapor-side product impurity content was 10% (vs. design value of 1%), due to a non-forgiving concentration profile. Over the eight design stages in the bottom bed, the concentration rose from 30% at the bottom to 50% four stages below the side draw, then dipped to 1% at the side draw. A miss by 1– 2 stages would bring the concentration to 10%.
212-26
Azeotropic column
Components A, B and C: Column separated a minimum-boiling AC azeotrope from a heavy C. Small amounts of light-boiler B (lighter than the azeotrope) escaped in the top product. Changes in the reactor produced much more B in the feed. The light B was expected to go up, but much of it ended up in the bottom. Reason was the formation of a much lighter AB azeotrope that distilled up, leaving a BC mix in the bottom.
Vapor and liquid loadings Calculated vapor and liquid loadings are the basis for all of the hydraulic evaluations of trays, packings and tower internals. Incorrect loadings mean that tower-capacity estimates will be incorrect. Usually, the hydraulic evaluation of a section of tower is based on the highest vapor and liquid loadings in that section. These are derived from the simulation. The cases in this sec-
C EP
O ctober 200 2
w w w .cepm agazine.org
55
Reactions and Separations
Case Type of No. -Ref. Column
Brief Description
Section 2. Does Your Distillation Simulation Ma tch Plant Data? (Continued) Refinery Vacuum Tower Wa sh Sections 213-26 Chemicals azeotropic column
Components A, B and C: Light product B was separated from heavy reactants A and C. With a feed rich in A and lean in B, pure B could not be produced due to minimum-boiling A/C and A/B azeotropes, the latter boiling 1˚C less than B. Problem solved by a column concentrating B (with some A and C) at the top, bottom being an A/C mix. C was added to the B concentrate en route to the next column, where it drew the A to the bottom (as an A/C mix), leaving B at the top.
214-26 Freon -22 (R22) Reflux column
Column separated a light HF/HCI/R22 mix from heavy HF recycle to the reactor. Two simulations, identical except for initial value differences of 0.15% HF, gave completely different results. This difference shifted the column across a distillation boundary, giving completely different end points.
220-27
C9 bottoms separated from ternary azeotrope that formed two liquid phases. Decanter acid/anhydride was distilled in second tower to remove C9. Residue curve map showed that when the feed is low on anhydride, there is no phase split in the decanter and the process fails. Cure was diverting low-anhydride feeds away from the towers.
Acetic acid/acetic anhydride/C9 alkane
Section 4. Are Your Chemistry and/or Process Sequence Correct? 110-7
Absorption of HF from HCl gas by wash with aqueous HCl
HF absorption was poor. HF escaping in the column overhead destroyed the downstream glass plant. The cause was that most of the "HF" in the feed was in the form of carbonyl fluoride. This component was sparingly soluble in water, but hydrolyzed slowly to HF.
141-28
Mini-plant Distillate from first two towers was all the phenol. Distillate from phenol and thirdshould have been phenol-free reactant, but contained 1.5% reactant recovery, phenol,formed by a previously unknown cracking reaction of the 3-tower train high boilers at the bottom. Solved by switching process sequence, so that high boilers are removed in the second tower and reactant separated from phenol in the third. Easy to switch in a mini-plant, almost impossible once a full-scale plant is built.
112-17
O-Nitro-toluene recovery
The residues were held at 150°F and air admitted. A previously unknown exotherm set in, causing an explosion.
134-30
Solvent/ residue batch still, vacuum
Column separating the reaction solvent and separation solvent from residue experienced excessive solvent losses to residue. Change in the upstream reactor, and using the same solvent for both reaction and separation, reduced feed inconsistencies, permitted semi-batch operation, and reduced solvent losses.
135-31
Pharmaceutical, batch distillation
Changing plant operation from a single to multiple distillation process rendered the residue more thermally unstable. This led to an unstable mixture in a reactor, which exothermically decomposed, caught fire and exploded, injuring one person.
Section 5. How Good is Your Efficiency Estimate? 307-7
Desorption of methanol, acetone and ammonia from water, using air
Column failed to achieve design separation. Design efficiency was predicted from air humidification and oxygen stripping studies in a single-plate laboratory column. Wall and downpipe mass-transfer enhanced efficiency in the lab column. This led to optimistic efficiency predictions.
308-32
Isopropyl alcohol/water azeotropic distillation using benzene and IP E entrainer
The 2-ft azeotropic distillation column using benzene and IP E entrainer with metal-mesh packing achieved 12-in. HETP (height equivalent of a theoretical plate). This was considerably higher than the design HETP. Column HETP was scaled up from small-diameter columns that had good and frequent liquid distribution and redistribution.
56
w w w .cepm agazine.org
O ctober 200 2
C EP
tion of the table show that a major weakness is that subcooling of reflux and feeds is not always properly allowed for in estimating the tower loadings. Further, the presentation of liquid and vapor in the simulation output is not generally user-friendly, and may often conceal higher vapor and liquid loads under superheated or subcooled conditions.
Trusting simulator hydraulic predictions The only amazing result in the reported cases is that only two cases involved problems with a simulation’s hydraulic predictions. Misleading hydraulic predictions from simulators is one of the major trouble spots, together with VLE data matching plant data and the need for scouting out problems via graphical techniques. This was stated in our previous work (16). Case 514, for example, is one of about a few dozen that we have seen in which simulations gave optimistic capacity predictions with packings. It is difficult to understand why only two cases were reported here (both by the author!). Many engineers regard it as a given that a significant number of the hydraulic predictions from simulators are untrustworthy. Others, especially those in high-tech companies, often prefer using proprietary prediction methods to those cached in the simulations, and have little incentive to take the simulator’s hydraulic predictions seri ously. The most troublesome hydraulic and efficiency predictions are for packed towers. There have been problems with optimistic predictions made by simulation software and by using a vendor’s program. When it comes to trays, jet and downcomer backupflooding predictions, at least some methods in simulators are reasonable, even good. Simulator predictions are far less satisfying for downcomer choke flooding, which normally limits high-pressure and foaming systems, as well as predicting tray efficiency.
Bugs in the simulator Fortunately, this is not a major issue, but it is encountered from time to time especially when simulation vendors come out with an upgraded version of their programs. Always be on the lookout for bugs. A bug in a VLE method in one simulator disappeared in the next upgrade, but returned in the following one. Corrective action calls for using an independent means of verifying equilibrium values and verifying the simulated mass, component and heat balances.
Lessons learned
Case No.-Ref.
Type of Column
Brief Description
Section 5. How Good is Your Efficiency Estimate?(Continued) 302-3
Acetylene solvent/water stripper
Some causes of excessive solvent losses were an incorrect efficiency estimate and excessive entrainment. Performance was improved by adding trays and mist-elimination pads under top section trays, and raising reflux.
304-4
Refinery vacuum
Tray efficiency was lower than expected. Column contained valve trays and operated at low liquid loads and with wide variations in vapor loads.
Section 6. Incorrect Simulation of Feed Entry Ma y Give Mi sleading Predictions 321-33
Refinery, FCC main fractionator, several towers
Following replacement of trays by structured packings, LCO/spongeoil draw temperature dropped 60˚F, LCO product contained 5% more gasoline, and the LCO stripper stopped stripping. Reason was that the two trays between the sponge-oil return and the LCO draw were eliminated. Gasoline-rich returned sponge-oil mixed with tower liquid to form the LCO product. Problem minimized by minimizing sponge oil. In one case, solution was returning sponge oil below the LCO draw, generating a pumpdown.
Examining these cases reveals seven key items that require vigil and attentiveness: 1. The major issues affecting the Section 7. Have You Specified the Correct Vapor and Liquid Loadings? validity of distillation simulations are After replacing trays by packing, column efficiency fell, incurring high getting good VLE data, having the 316-34 Olefins demethanizer ethylene losses. Gamma scans showed poor liquid distribution in the simulation match plant data, and upper two beds, flooding in the third, and poor vapor distribution using graphical techniques to trouin the bottom bed. The design made no allowance for vapor bleshoot simulations. Another key condensation by the highly subcooled (70˚F) feeds. This overloaded distributor capacities. Some improvement achieved by rerouting issue is obtaining good hydraulic presome of the cold main feed to an upper bed. dictions from the simulation. 330-12 Petrochemical Tower flooded 5– 10% below design because additional vapor and 2. Providing the correct chemistry liquid traffic induced by 100°F reflux subcooling was not accounted and the correct tray or packing effifor in the internals design. Solved by using bubble-point reflux. ciency is needed to ensure the validity The design vapor rate in the slurry section of a FCC fractionator did of the tower simulation, but is less 306-23 Refinery fluid catalytic not allow for vaporization that occurs when a bottom feed with troublesome than the factors mencracker 300°F superheat contacts column liquid. Column therefore tioned above. Modeling feeds, prefractionator prematurely flooded. Problem solved by injecting subcooled quench liquid to desuperheat the feed. At a later stage, subcooled quench dicting vapor and liquid loads, and was replaced by a lighter liquid that vaporized, and premature detecting and correcting simulation flooding reoccurred. bugs cause trouble in tower simulaSection 8. Simulator Hydraulic Predictions: To Trust or Not to Trust? tions, but to a lesser extent. 3. VLE predictions from commer514-16 Refinery 2-in. Pall rings were replaced by 3-in. modern random packings. vacuum Expected capacity increase was 30% but only 17% materialized. Both cial simulations are most troublethe default and supplier’s options in a commercial computer some with close-boiling composimulation were optimistic, leading to the high expectation. nents, non-ideal systems, or heavyA commercial simulator gave optimistic prediction of packing component characterization in 515-16 High pressure capacity because it allowed extrapolation of a good correlation well crude-oil distillation. Predictions for beyond its applicability limits. components that have medium or Section 9. Bug in Simulation high relative volatilities and no 326-35 Specialty Well-known commercial simulations with successful convergence major non-idealities usually do not chemical and no error messages had erroneous energy balances on all three cause a problem. towers. Cause was a bug in the default convergence software. Repeat 4. The major hurdle in matching with an alternative convergence procedure gave valid mass and energy balances. Using the original simulation, all three reboilers would have a simulation with plant data is obbeen grossly undersized and tower feed grossly mislocated. taining a reliable, consistent set of the plant data. Getting correct numbers from flowmeters and laboratory analyses is the major headache. Specific issues re5. The key graphical techniques invaluable for trouported here are situations when a second liquid phase is bleshooting simulations are the McCabe-Thiele and Hengpresent, and when simulating the wash sections of refinstebeck diagrams, multicomponent-distillation compositionery vacuum towers. profiles, and, in azeotropic systems, residue-curve maps.
C EP
O ctober 200 2
w w w .cepm agazine.org
57
Reactions and Separations
6. In chemical towers, problems can arise when reactions are not properly accounted for in the simulation and/or when a component believed to be in one physical form turns out to be in another. 7. Estimating tray and packing efficiencies is not a major issue for established processes operating in conventional hardware. Most of the prediction issues arise when C EP simulating new systems or hardware.
Literature Cited 1. Kister, H. Z., “Are Column Malfunctions Becoming Extinct — or Will they Persist in the 21st Century?,” Trans. IChemE, 75, Part A, p. 563 (Sept. 1997). 2. Kister, H. Z., “Distillation Operation,” McGraw-Hill, New York (1990). 3. Martin, H. W., “Scale-up Problems in a Solvent-Water Fractionator,” Chem. Eng. Progress, 60 (10), p. 50 (Oct. 1964). 4. Guy, J. L., and J. A. Bonilla, “Case History of a Retrayed Column: Troubleshooting Techniques and Methods,” paper presented at the AIChE Spring National Meeting, New Orleans, LA (Mar. 29–Apr. 2, 1992). 5. Sloley, A. W., et al., “Why Towers Do Not Work,” paper presented at AIChE Spring National Meeting, Houston, TX (Mar. 20–24, 1995). 6. Moura, C. A. D., and H. P. Carneivo, “Common Difficulties in the Use of Process Simulators,” B. Tech. Petrobras, 34 (3/4), (Jul./Dec. 1991), quoted in R. Agrawal et al., “Uncovering the Realities of Simulation,” Chem. Eng. Progress, 97 (5), p. 42, (May 2001). 7. Rose, L. M., “Distillation Design in Practice,” Elsevier, Amsterdam, The Netherlands (1985). 8. Staggs, D. W., “The Impact of Non-Ideal Vapor/Liquid Behavior on Solvent Emissions,” paper presented at AIChE Spring National Meeting, Houston,TX (Mar. 20–24, 1995). 9. Kalthod, V. G., et al., “Distillation Column Performance Testing: Continuous and Batch Approaches,” in “Preprints of the Topical Conference on Separation Science and Technologies, Part I” p. 225, AIChE Annual Meeting, Los Angeles, CA (Nov. 17–19, 1997). 10. Sadeq, J., et al., “Anomalous Results from Process Simulations,” paper presented at AIChE Annual Meeting, Miami Beach, FL (Nov. 1995). 11. Golden, S. W., et al., “Feed Characterization and Deepcut Vacuum Columns: Simulation and Design,” paper presented at AIChE Spring National Meeting, Houston, TX (Mar. 20–24, 1995). 12. Sloley, A.W., et al., “Troubleshooting Practice in the Refinery,” paper presented at AIChE Spring National Meeting, Houston, TX (Apr. 2001). 13. Kister, H. Z., et al., “Does your Distillation Simulation Reflect the Real World?,” Hydrocarb. Proc., p. 103 (Aug. 1997). 14. Kister, H. Z., et al., “Debottleneck and Performance of a Packed Demethanizer,” in “Proceedings of the 4th Ethylene Producers Conference,” New Orleans, LA, p. 283 (1992). 15. Kister, H. Z., et al., “Troubleshooting a Water Quench Tower,” in “Proceedings of the 7th Ethylene Producers Conference,” Houston, TX (1995). 16. Kister, H. Z., “Troubleshooting Distillation Simulations,” Chem Eng Progress, 91 (6), p. 63 (June 1995).
58
w w w .cepm agazine.org
O ctober 200 2
C EP
is director of fractionation technology at Fluor Daniel (One Fluor Daniel Dr., Aliso Viejo, CA 92698; Phone: (949) 349-4679; Fax: (949) 349-2898; E-mail:
[email protected]). He has over 25 years of experience in design, troubleshooting, revamping, field consulting, control and startup of fractionation processes and equipment. Previously, he was Brown & Root’s staff consultant on fractionation, and prior to that he worked for ICI Australia and Fractionation Research Inc. (FRI). He is the author of textbooks “Distillation Design” and “Distillation Operation,” as well as over 60 published technical articles, and has taught the IChemEsponsored “Practical Distillation Technology” course 240 times. He obtained his BE and ME degrees from the Univ. of New South Wales in Australia. He is a Fellow of IChemE, a member of AIChE, and serves on the FRI Technical Advisory and Design Practices committees.
HENRY Z. KISTER
17. Kister, H. Z., et al., “Sensitivity Analysis is Key to Successful DC5 Simulation,” Hydrocarb. Proc., p.124 (Oct. 1998). 18. Opong, S., and D. R. Short, “Troubleshooting Columns Using Steady State Models,” in “Distillation: Horizons for the New Millennium,” in “Topical Conference Preprints,” AIChE Spring National Meeting, Houston, TX, p. 129 (Mar. 14–18,1999). 19. Golden, S. W., et al., “Improved Flow Topology for Petroleum Refinery Crude Vacuum Distillation Simulation,” 44th Canadian Chem. Eng. Conference, Calgary, Alberta, Canada (Oct. 2–5, 1995). 20. Golden, S. W., et al., “Refinery Vacuum Column Troubleshooting,” paper presented at AIChE Spring National Meeting, New Orleans, LA (Mar. 31, 1993). 21. Golden, S. W., et al., “Refinery Analytical Techniques Optimize Unit Performance,” Hydrocarb. Proc., p. 85 (Nov. 1995). 22. Kister, H. Z., et al., “Problems and Solutions in Demethanizers with Interreboilers,” in “Proceedings of the 8th Ethylene Producers Conference,” New Orleans, LA (1996). 23. Lieberman, N. P., “Troubleshooting Process Operations,” 3rd ed., PennWell Books, Tulsa, OK (1991). 24. Sloley, A. W., and S. W. Golden, “Analysis Key to Correcting Debutanizer Design Flows,” Oil & Gas J., p. 50 (Feb. 8, 1993). 25. Kister, H. Z., et al., “Improve Vacuum-Tower Performance,” Chem. Eng. Progress, 92 (9), p. 36, (Sept. 1996). 26. Short, D. G. R., “Using Residue Maps for Solving Separation Problems,” paper presented at AIChE Spring National Meeting, Houston, TX (Mar. 9–13, 1997). 27. Partin, L. R., “Use Graphical Techniques to Improve Process Analysis,” Chem. Eng. Progress, 89 (1), p. 43 (Jan. 1993). 28. Ruffert, D. I., “The Significance of Experiments for the Design of New Distillation Column Sequences,” in “Distillation 2001: Frontiers in a New Millennium,” in “Proceedings of the Topical Conferences,” AIChE Spring National Meeting, Houston, TX, p. 133 (Apr. 22–26, 2001). 29. “Miscellaneous Case Histories,” in “Fire Protection Manual,” Vol. 2, C. H. Vervalin, ed., Gulf Publishing, Houston, TX, p. 29 (1981). 30. Williams, J. A., “Optimize Distillation System Revamps,” Chem. Eng. Progress, 94 (3), p. 23 (Mar. 1998). 31. “HSA Criticizes Hickson,” The Chem. Engineer, p. 5 (Nov. 10, 1994). 32. APV DH-682, “Distillation Handbook,” 2d ed., APV, Chicago, IL (no date of publication given). 33. Golden, S. W., “Case Studies Reveal Common Design, Equipment Errors in Revamps,” Oil & Gas J., p. 62, (Apr. 7, 1997 and Apr. 14, 1997). 34. Freeman, L., and J. D. Bowman, “Use of Column Scanning to Troubleshoot Demethanizer Operation,” paper presented at AIChE Spring National Meeting, Houston, TX (Mar. 31, 1993). 35. Le, N. D., e t al., “Doublecheck Your Process Simulations,” Chem. Eng. Progress, 96 (5), p. 51, (May 2000).
