1996: New Urea Process Designs - IFFCO : KANDLA

New Urea Process Designs

The urea 2000plus process is the state-of-the-art technology with more than 2,000-mtpd capacity. Itcan reduce the investment cost of the synthesis section by retaining advantages of the CO2 stripping

process.

K. Jonckers, J. Mennen, J. Meessen, and W. LemmenStamicarbon bv, Geleen, The Netherlands

Introduction

This article discusses completely new ureaprocess designs, and it is interesting to high-light the presently existing highly competitive

CO2 stripping urea synthesis. This process has beenlicensed by Stamicarbon 115 times.

113 conventional CO2 stripping units1 pool condenser CO2 stripping unit1 pool reactor CO2 stripping unit

115 CO2 stripping unitsThe fourth section summarizes the technology of thepool condenser and the pool reactor in a modernStamicarbon urea plant. The advantages of the poolcondenser and pool reactor are as follows:

• Approximately 10% lower investment• Smaller or lesser number of HP equipment• More operations flexibility• Lower skyline• Total gravity flow in synthesis section

Stamicarbon has succeeded in vastly improving itsCO2 stripping urea process. The feedstock consumptionfigures are almost equal to the stoichiometric values forammonia and carbon dioxide, leaving no scope for anyfurther reduction. The utility consumption figures havealso been reduced to acceptably low values. Furtherreduction of steam consumption would require additionalinvestment in relatively new heat exchange methodsresulting in a more complicated process that is wise as aninvestment, but is expensive.

The Stamicarbon CO2 stripping process has extremelylow effluent and emission figures. Being almost zero, theyalso leave no scope for any significant improvement.

The new Urea 2000plus process includes a numberof significant improvements, which ultimately resultin lower investment costs for a new urea plant. Themain improvements are a superior reactor tray design,a pool condenser, and a pool reactor. The first twoimprovements are already commercially proven, whilethe pool reactor is to be incorporated hi DSM's newurea plant at Geleen, the Netherlands, commissionedin 1997.

AMMONIA TECHNICAL MANUAL 250 1997

Since the pool condenser and pool reactor technolo-gy is purely urea synthesis technology, this articledeals with the urea synthesis only.

History of CO2 Stripping Process

The CO2 stripping urea process was developed in olderto reduce the steam consumption of the conventional ureaprocess (which was about 1,600 kg/ton). This developmentresulted indeed in a steam consumption reduction by 600 to800 kg/ton. The process has been a success ever since andmore than 100 units were built. Another breakthrough inthe late 1960s was the introduction of type 25-22-2 tubematerial in the stripper. This material really lived up to itsexpectations and is still successfully used.

Typical features of the CO2 stripping process:• Reaction pressure about 140 bar.• Ammonia/carbon dioxide ratio in the reactor about 3.• Nonconverted NH3 and CO2 recycle at reaction pressure.The evolution of the CO2 stripping process, up to the intro-

duction of Urea 2000plus technology can be summarized asfollows:

• Small-scale plants with helicoil carbamate condenserabove the reactor (total gravity flow).

• Large-scale plants with vertical falling film condenser(total gravity flow).

• Large-scale plants with vertical falling film condenserand a low elevation for the HP scrubber (partial gravityflow).

Basically, this last plant type is still being built today.The basic flowsheet of this type of synthesis is shown inFigure 1.

Because many plants have been built with this flowsheetas a basis, this article will not elaborate on this synthesisconfiguration but will deal only with the innovative Urea2000plus technology.

The skyline of the Stamicarbon CO2 stripping processplant has changed dramatically over the years (Figure 2).

Stamicarbon Urea 2000plus

Pool condenser: an alternative to the falling film-type HP carbamate condenser

The pool condenser is used in the KAFCO urea

plant at Chittagong in Bangladesh. The plant has adesign capacity of 1,725 mtpd and was constructed byChiyoda of Japan. It came on-stream at the end of1994.

As described earlier, the pool condenser concept wasselected for this plant because of its location in thevicinity of Chittagong Airport. A low skyline was astrict requirement. The pool condenser allowed thereactor to be made considerably shorter because acompensating reactor volume would be available inthe shell of the horizontal carbamate condenser(Figure 3).

Basically, the pool condenser is a horizontal reactorvessel with a submerged U-tube bundle as shown inFigure 3. It is fabricated from carbon steel with a lin-ing in urea-grade type 316L stainless steel (similar tothe standard urea reactor).

All internals and the overlay welding on thetubesheet are fabricated from type 25-22-2 stainless.

The tubes are welded to the tubesheets by internalbore welding. This welding technique is proven formany years in carbamate service and no difficulties ofany significance have ever been encountered.

Operating Principle. This horizontal condenser-reactor is a gas-agitated vessel with a retention timeenabling the condensed carbamate to dehydrate par-tially into urea and water. The gases condense arounda bundle of U-tubes, generating low-pressure steam asin any other HPCC.

The stripper off-gas, the carbamate solution from theHP scrubber, and ammonia from battery limits areintroduced into the condenser. The liquid phase isthoroughly agitated by the gases from the stripper, andthe heat of condensation is utilized for producingsteam in the tube bundle.

The condensation capacity of the condenser is deter-mined by the heat-transfer coefficient, the surface areaof the tube bundle, and the temperature differencebetween the process side and the steam side.

The temperature difference can be controlled by thepressure of the generated steam. Moreover, the processtemperature is a function of the rate at which urea isformed with all other parameters, such as the pressure,the NH3/CO2 ratio and the initial H2O content, remain-ing constant.

The rate of formation follows from the installed


retention time. Our models accurately describe thetemperature as a function of the rate of formation.

The temperature in a falling film condenser in whichalmost no urea formation takes place is about 170°C.With 25% urea at 60% approach to equilibrium, thistemperature is raised to about 175°C.

As is normal, the temperature of the generated saturatedsteam is dictated by the steam pressure of 4.5 barrequired in the downstream sections of the plant.

The amount of heat of condensation is determinedby the involved chemical reactions.

The operating principles of the pool condenser canbe summarized as follows.

• Strip gases are condensed in a pool of liquid on theshell side, with LP steam being generated on the tubeside.

• Adequate residence time allows the reaction ofammonium carbamate to urea and water to proceed toup to 60% of equilibrium.

• The condensation temperature on the shell side is high as aresult of the formation of high boiling components (urea andwater). The formation of gas bubbles ensures a high degree ofturbulence and provides a large area for mass and heat trans-fer.

Development of the Pool Condenser. The pool con-denser took several years and a great deal of effort todevelop. Use was made of extensive experience gained inurea plants of our licensees in which similar condensersare used.

These condensers are: (a) the low-pressure carbamatecondensers applied in the 4-bar recirculation section of allplants designed since the late 1950s; (b) the medium-pres-sure carbamate condensers in revamped conventionaltotal-recycle urea plants operated at 18 to 20 bar. Thisexperience dates back to the early 1970s; (c) the high-pressure scrubbers hi the first-generation CO2 strippingplants operated at 130 bar; (d) the bottom section of theHP scrubbers in the CO2 stripping urea plants designed inthe 1970s and 1980s. The operating pressures in thesescrubbers range from 140 to 150 bar.

The first three heat exchangers in this list are quite simi-lar. They are installed in a vertical position, and ammonia,carbon dioxide, and water vapor are introduced in the bot-tom. The resulting condensâtes go overhead along withminor amounts of noncondensed gases evolving in thetop. Heat is removed via cooling water in U-tubes sub-

merged in the solution. The fourth heat exchanger (scrub-ber) is a vertical shell-and-tube exchanger, with gas andliquid rising cocurrently from the bottom to the top. In allthese condensers, pool condensation takes place wherebythe liquid forms the continuous phase.

A common feature of these condensers and the poolcondenser is that there is intensive mixing of the vaporand liquid phases, resulting in a highly turbulent heattransfer to the cooling water tubes. The difference is thatall of the above-mentioned condensers are vertical,whereas the pool condenser is horizontal.

Design of Pool Condenser. The heat-transfer coefficientused in this first pool condenser was conservatively cho-sen, and this ultimately resulted in a higher than anticipat-ed LP steam pressure in the condenser.

Further design criteria are presented here below:• Process design: shell side, 140-bar synthesis pressure

175°C; tube side, 4.5-bar steam/BFW; gas dividing sys-tem with flow deflector plates.

• Mechanical design: stainless steel; U-tube bundle;internal bore welding; tube supports.

Advantages of Pool Condenser over Falling-FilmCondenser. The technological advantages of the poolcondenser are shown in Table 1.

Another advantage of a urea synthesis equippedwith a pool condenser is that a smaller reactor can beused.

Many calculations done through the years indicatethat the optimum approach to equilibrium in the ureareaction is 95%. This is illustrated in Figure 5.

As can be seen from this curve, about 60% of theapproach to equilibrium is reached already within thepool condenser whose capacity is only 30 to 40% ofthe required reaction volume. As a direct consequence,the actual reactor volume can be considerably reduced.The improved heat transfer in the pool condenser incombination with the reduced reactor volume resultsin different equipment dimensions in the synthesissection.

Both the heat exchange area in the condenser andthe reactor volume will be smaller (Table 2). A smallerreactor volume obviously gives weight savings, mak-ing transportation to the site and site handling a loteasier.


From HPcsrfeamats pump

CompressorTo «circulation section

1. Reactor2. Stripper3. Condenser

4. Scrobbsr5. Pool cosden»6. Pool reactor

Figure 1. Falling film HP carbamate condenser. Figure 2. Skyline of the Stamicarbon CO2 strippingprocess plant has changed dramatically over the

years.

Ta reeltoulstlon

C-8t»l will with SS llnvr Tube »tractC-«t»l with SS Unir

U-tubls. dllllctor plitllend bit«»

H«.d C-»t«el

Figure 3. Pool condenser synthesis section. Figure 4. Pool condenser.


F.A.E. {%)

t180%

PoolCondenser

Tfthii«)

Inarts carbarn etapump

From HPNH, pump

From CO2

CompressorTo racirclntionsaction

Figure 5. Optimum reactor design. Figure 6. CO2 stripping synthesis (pool reactor).

C-steel wall with SS liner Tube «heetC-steel with SS liner

Gas divider

U-tubes, deflector platesand baffle»

HeadC-steel

Figure 7. Pool reactor.


Operational Experience with Pool Condenser.Basic observations:• From the very first startup of the synthesis, it

became clear that the fundamental design aspects ofthe pool condenser were correct.

• The LP steam pressure achieved in the condenseris about 1 bar higher than anticipated due to better heattransfer as compared to the design.

• The space around the tube bundle, forming thereactor volume in the pool condenser, was anticipatedto be equal to one continuously stirred tank reactor(CSTR) which has a mixed temperature of 175°C.However, the large increase in temperature along theshell indicates that the pool condenser behaves moreor less like a plug-flow reactor, resulting in a higherurea concentration in the top of the following reactoras a bonus.

The observed temperature increase (from 174°C toabout 180°C) at a pressure of 138 bar in the synthesisis a definite advantage if compared to 170°C at 140bar in a conventional HPCC.

• The final temperature in the top of the urea reactoris about 186°C at a pressure of 138 bar. We have neverexperienced this before in any of our plants, but wecan completely reproduce it in our computationalmodels.

The design of the reactor which comes after the poolcondenser is comparable with all other reactors, except, ofcourse, for its volume. Its feed, however, has already cov-ered 60% of the approach to the equilibrium composition.Channeling of the almost finished product will hardlyaffect the composition in the top of the reactor.

• Experience indicates that the pool condenser ismuch less sensitive to variations in the N/C ratio. Thisis quite understandable, since condensing off-spec gasalone is more difficult than condensing the same off-spec gas in a pool providing ample amounts of one orother failing component in the gas phase.

• The inverse response occurring in all our strippingplants has disappeared here, so that the level controlvalve can now be put on automatic.

Until now, on slightly opening the level control valve, thelevel in the reactor always increased a bit before it went down.This is due to increased gas generation in the stripper part.Initially, the gas is not completely condensed and returns tothe reactor. This phenomenon is called inverse response. In

the pool condenser, however, all of the gas condenses straight-away and the inverse response has disappeared.

Operational Experience at KAFCO (Bangladesh)since December 1994

• The fundamental concept of the pool condenser iscorrect.

• The heat transfer is even better than anticipatedresulting in 5.5 bar steam (goal was 4.5 bar).

• The pool reactor acts as a bank of CSTRs ratherthan as a single CSTR. This was evidenced by a tem-perature increase to 180°C from 174°C rather than auniform temperature of about 175°C (unexpected ben-efit)

• The inverse response in the synthesis loop has van-ished (unexpected benefit).

• Deviating N/C ratios do not affect condensation inthe pool condenser and thus the synthesis pressure(unexpected benefit).

• The pool condenser has a positive effect on thereactor temperature and pressure. It is now 186°C at138 bar (unexpected benefit).

• An inspection of the pool condenser after one yearof operation did not show any abnormal corrosion tovessel wall or internals.

Future developments

A major development resulting from the successful per-formance of the pool condenser in Bangladesh is theincorporation of all the required retention time for the ureareaction in one vessel, creating the pool reactor.Furthermore, the introduction of the pool condenser is astep closer to the N=3 low-energy process, in which strip-ping steam is used three times, reducing the energydemand of the CO2 stripping process.

Table 3 summarizes further potential improvements.The first pool reactor, for DSM in Geleen, is nowbeing fabricated for a plant capacity of 1,150 mtpd.This new plant is slated to come on-stream in 1997.

Conclusions concerning pool condenser

Summarizing the main features of the pool condenser, the fol-lowing conclusions can be drawn:


Table 1. Technological Advantages

Pool Condenser Falling Film Condenser

Heat/mass transfer High turbulence

Large transfer area (bubbles)Large LMTD due to ureareaction.Tprocess = 175°C

Restricted by laminarlayer between gas andliquid film.Limited area (tubes)Small LMTDTprocess = 170°C

Table 2. Investment Advantages

Equipment Dimensions in Urea Synthesis, Base Case 2,000 mtd plant capacity

Cylindrical reactor volumeCylindrical pool condenser volumeCondenser heat exchange area

Pool Condenser

150nri380m3

55%

Falling Film Condenser

230

100%

Table 3. Future Developments

1 . Further development of low-level arrrangement• Reactor bottom at ground level.• 2,000 mtpd urea plant with total elevation of max. 30 m.• Considerable investment savinas.

2. Application in N = 3 low-energy process• Energy supplied to HP stripper to be used three times using multiple effect heat

exchange.3. Further refinement of the pool reactor

• Substantial investment savings.


(1) The pool condenser reduces the urea reactor volume bysome40%.

(2) The pool condenser requires some 45% less heatexchange area.

(3) The pool condenser reduces the height of the steelstructure by some 10m.

(4) No stress corrosion can take place in the pool con-denser.

(5) The pool condenser is commercially proven andis operationally a success.

(6) The pool condenser paves the road to furtherimprovements such as the following:

• Further reduction of the height of the steel structure.• Further energy savings.• Development of the pool reactor.

Pool Reactor

After the successful commissioning of the pool con-denser at KAFCO with its process and operational advan-tages and high temperature at the outlet, it became obvi-ous that the step to the pool reactor was only a small one.Installing some more volume in the pool condenser, butwithout spoiling the reactants which are almost in equilib-rium with fresh condensed carbamate, the idea of the poolreactor was born and patent applications were filed.

As chance would have it, the DSM Melamine businessunit was planning to scrap its 40-year-old urea plant atGeleen and replace it by a new one.

Following a thorough evaluation process in whichall aspects of the pool reactor, the process, and thecommercial risks were considered, DSM Melamine, incooperation with Stamicarbon, decided to opt for thepool reactor for this new 1,150-mtpd urea synthesisplant.

Synthesis configuration

The principle of the pool reactor synthesis section isshown in Figure 6.

The urea synthesis section shown in this flowsheet nolonger includes the vertical reactor and basically compris-es a CO2 stripper, an elongated pool condenser now calledthe pool reactor, and the safety features of the HP scrub-ber. The heat exchanger in the HP scrubber is omittedalso.

Pool reactor

Construction of the pool reactor is similar to that of thepool condenser, and the same materials can be used. Theconstruction, the materials, and leak detection system arebased on Stamicarbon's experience with the fabrication ofurea reactors. The only difference is that the pool reactoris in a horizontal position while its predecessor was placedin a vertical position.

As in the pool condenser, the tubes are welded to thelined tubesheet using an accurately well controlled inter-nal bore welding procedure. A sectional drawing of thepool reactor is given in Figure 7.

Advantages of pool reactor

The expected advantages of the pool reactor configu-ration are as follows:

• The amount of transferred heat through the con-denser bundle in the reactor will be the same as in thepool condenser configuration (far better then in thefalling film condensers).

• The baffles in the reactor positively prevent back-mixing, so affording the highest approach to equilibri-um in the urea reaction and consequently the highestconversion and most economic downstream equipmentsizing.

• The same operational flexibility as with the poolcondenser, namely

(a) Synthesis less sensitive to deviating N/C ratios;(b) No inverse response in the synthesis loop.• Considerably lower investment because of:(a) Omission of an expensive HP vessel.(b) Omission of HP ejector.(c) Far less HP piping in the synthesis section(d) Lower steel or concrete structure.

Proprietary equipment

Stamicarbon/DSM have developed the pool condenserand reactor through extensive research efforts, coveringboth the process aspects and fabrication. Internal borewelding is a specific feature of fabrication.

Stamicarbon has extensive experience in HP equipmentfabrication, from which customers can benefit. In theinterest of our customers, the new vessels will be fabricat-


ed as proprietary equipment by selected manufacturersonly. Selection of these manufacturers is underway.

Availability for licensing

The UREA 2000plus technology is ready for licensingby Stamicarbon through its licensed contractors.

The pool condenser plant in Bangladesh has nowoperated for more than 20 months and the pool reactorplant of DSM will be on-stream well before a thirdplant can be commissioned. Consequently, anyteething troubles, if any, in the pool reactor plant atDSM will have been resolved in time. The UREA2000plus technology can also be used in plantrevamps and debottlenecking projects in order toimprove profitability.

Process Data

Consumption figures

Finishing technique: Prilling Granulation Unit

Raw materials

NH3 (100%)

CO2 (100 %)

Utilities

568

733

564

730

855MP steam(22 bar, 330°C)Electricity 14Cooling water (A = 10°C) 58UF85

Utilities production

805

50507.5

LP steam (4 bar sat.) 370 415Steam condensate export 235 280Process condensate export 570 245

kg

kg

kg

kWhtonkg

kgkgkg

All figures are per metric ton of final product andare given by way of example only. Local conditionsmay have a great impact on the optimum consump-

tions.

Scope of performance figures

The figures include:

• ammonia compression• synthesis section• recirculation section• evaporation section• wastewater treatment section« finishing section (either prilling or granulation).

The above figures exclude:• CO2 compression• storage and bagging facilities.

Product Characteristics

Nitrogen (N) content, wt. % 46.4Normal biuret content, wt. % 0.85Low biuret content, wt. % 0.25Moisture content, wt. % 0.25

Wastewater characteristics

Urea concentration 1 ppm wt.Ammonia concentration 1 ppm wt.

SI units are used.In this system 1 ton equals 1,000 kg.

Conclusion

Technology can always be improved upon.Stamicarbon has again proven that mature technology

can be innovated, while retaining all process advantagesof the CO2 stripping process at the same time substantiallyreducing the investment cost of the synthesis section.

Stamicarbon's target is to remain the world leader inUREA technology. With its UREA2OOOplus technology,Stamicarbon can license state-of-the-art urea plants withcapacities far above 2,000 mtpd for dependable servicewell beyond the year 2000.

Clients opting for UREA 2000plus technology will beready for the challenges of the future.


DISCUSSIONE. R. Kilian, Continental Engineering: Mr. Lemmen,congratulations. This is a great step forward. Can yougive me an indication about the investment reductionof a 1,700 ton plant with this technology combinedwith a 1,0001 ammonia plant?Lemmen: I do have an indication, Mr. Kilian. Theestimation department of DSM Engineering has beenlooking at this because for Stamicarbon, it's a veryimportant issue. The plant in the Netherlands, wherewe considered the three options, actually arrived at a100% figure for the vertical falling film carbamatecondenser, 97% for the pool condenser, and 90%-91%for the pool reactor. We can say for the pool reactor incomparison with the old concept that it was very closeto a 10% investment reduction for the plant inside batterylimits.Kilian: What about the second question about startupand shutdown. Any emphasis on that part?Lemmen: Yes. There are also big advantages simplybecause if you look at the pool condenser and the poolreactor, and you compare that with the vertical fallingfilm carbamate condenser, you see that starting up avertical falling film carbamate condenser alwaysresults in the filling up of a reactor until it is overflow-ing. Then, all things happen at the same time. Youhave to get the slip in operation and get all down-streams in operation all at once. However, with thepool reactor, because there are certain provisionsmade, we can, during filling up of that reactor, alreadyallow certain flows to go directly to the stripper. Wecan very slowly load the stripper and all downstreamequipment.F. Granelli, Snamprogetti: Looking at Figure 7 whichillustrates the pool reactor, I can clearly understandthat the fresh CO2 in the stripper enters this equipmentwith the sparger all along it. I can also understand thatin the first part of the figure and on the right part, youhave the tube bundle. Condensation is improved bythe turbulence. You have experience in your plant inBangladesh, but not with a full condensation. It is onlya partial condensation, but in the second part of thisreactor, I see some diaphragm to prevent that back-flow. The gas is divided into this sector with the head

of the liquid that is 1-3 meters, according to the size ofthe plant, and that means the diameter of the reactor. Ifyou compare with today's plants with 20 meters or 30meters, here you have no trays. You have a very lowhead of liquid. Maybe also the molar ratio is not favor-able and, finally, the residence time of the last part isvery short. I expect difficulty in condensation there.What is your feeling about that, unless you have avery strong recycle?Lemmen: Mr. Granelli, I appreciate your concern thatit won't work. However, we are convinced that by thesparger we will introduce gases in each and everycompartment. The movements in each compartment ofthe liquid will ensure that we will not have disastrouseffects on the urea reaction. Of course, this thing hasnever been built. It raises all kinds of questions. Yourquestion is one that has been brought up by DSMbefore they started to invest in this plant. A hundredtimes, our engineers have been looking at exactly thatsame issue. What you are bringing up is a criticalissue. We feel that with the provisions made for goodgas distribution and with the provisions made to avoidany backmixing, we will be able to reach at the end ofthat reactor the very high temperature of 186°C with apressure even a bit below the 138 bar pressure. So, Icannot tell you that it works because it doesn't at thismoment. I will be very happy to inform you about thatnext year, but that would probably be too early.Certainly, I will come back two years from now andtell you that it works.P. Orphanides, Consultant: Congratulations for yourexcellent presentation and these new ideas that youbrought up for the urea industry. Assuming your predic-tions that this urea horizontal reactor will work fine, thequestion I have concerns the inspection of the reactor inthe shell. You have in this equipment a rather voluminousbundle which is not removable, and you still have thesame lining where you need to make inspections andeventually repairs. Would that be a problem, or do youthink that you will never need to inspect and repair thelining.Lemmen: Of course, we will have to inspect it. Actually,the one in Bangladesh has been inspected once already.


Before DSM made the decision to go for the pool reactor,the inspection was done and we didn't find anythingunusual in there. There was no more or less corrosion thanin any other urea plant after a year of operation. To be ableto inspect it, we have designed it in such a way that youcan get around the bundle with at least a half a meterspace all through the reactor. The difficulty is if you gofor a very small plant. So, scaling down is a problem herebecause of what you're saying. Scaling up is no problem.If you are looking at liner repairs in half a meter space,they can be easily done. If you are looking at repairs on

tubes, then we indeed have a problem. However, becauseof the fact that we have only a very limited number ofmanufacturers, and because of the fact that the internalball welding process has been proven in several other car-bamate condensing environments for almost 20 yearsnow, we feel very confident that won't happen. The ques-tion which is raised every time we talk is if we have toplug a tube, what then? We have made tests at BSL onhow to plug tubes. It can all be done, and it can all bedone easier than one would expect.


Documents

1996: New Urea Process Designs - IFFCO : KANDLA