Vacuum Systems

Mr. Norman P. Lieberman – Why does an increase in the tower bottom’s level in my vacuum tower cause a loss in vacuum? Does this indicate an excessively high tower bottom’s level?

--George

            George – No. The level is not the problem. It’s that your boot temperature is too hot. The problem is excessive cracked gas formation in the boot. Assuming your level change is within the level span of the boot, increase the level from 30% to 70%. If the vacuum is adversely affected, try lowering the vacuum tower bottom’s temperature, using the circulating boot quench, by 10°F. You are also promoting coke formation in the boot which may coke-up the suction of your vacuum tower bottom’s pumps.

--Norm

Vacuum Systems

Mr. Norman Lieberman – What does it indicate if we have a gradual increase in CO₂ in the vacuum tower waste gas to the vacuum heater? A year ago the CO₂ concentration was one percent and now it is ten percent. Our vacuum is still quite acceptable. Also, the nitrogen (but not oxygen) concentration of the waste gas stream is also increasing, which I assume indicates an air leak.

--Carl

            Carl – You most likely have an air leak in the vacuum heater transfer line feeding the vacuum tower. The leak is getting progressively worse. Since the leaking line is under a slight vacuum, relatively small amounts of air are drawn into the vacuum tower. The oxygen then reacts with the hydrocarbons to form CO₂ (and perhaps some CO). Regardless, if you have a sudden loss of vacuum, the area of the leak will pressure-up. Hot vacuum tower feed will erupt from such a leak and auto-ignite. This is a very dangerous situation and I’ve been involved in such incidents three times in my career.

            I don’t really know how to advise you. Shutting down to repair the leak in the transfer line may very well result in a fire. One of my clients burned down their vacuum tower due to such a failure. Sorry for the bad news, but I believe my analysis is quite accurate.

--Norm

Vacuum Systems

Norm – I was reading your chapter on vacuum system operation. You talked about how raising the level in the seal drum can improve vacuum. I have observed this on our vacuum tower myself.  Certainly, this is not normal. We have to flood our seal drum to sustain a reasonable vacuum.

--Harvey C.

            Harvey – Unfortunately, you have developed a leak in the barometric drain line (i.e., your seal leg inside the seal drum). This is caused by biological corrosion. Raising the level to submerge this leak stops the seal drum gas from getting sucked-up into the leg and interfering with this drainage. I know it seems backwards, but flooding the drum will lower the condensate back-up level in the condenser. This lowers the vapor outlet temperature from the condenser and thus unloads the downstream ejectors.

            I suggest that you add some biocide to your seal drum to suppress the biological corrosion. Then, during the next turnaround, replace the carbon steel seal legs with 316 (L) S.S., or whatever allon is suitable for your service.

            I’ve discussed a similar incident in my book, Troubleshooting Process Operations, 4^th Edition, PennWell Publications.

--Norm

Vacuum Systems

Mr. Lieberman – I have a vacuum tower question. Reducing the ejector steam supply below its design (vendor specified) pressure improves vacuum. How could this happen? I suppose it’s best to operate at the lower seven bar pressure, rather than the design ten bar of steam pressure.

--Sriganesh

            Sriganesh – Likely one third of the jets I look at have the property that reducing the motive steam supply pressure below the design value improves vacuum or at least does not harm the vacuum.

            The two most common reasons for this are that the ejector steam nozzle is badly worn, or that the downstream condenser is overloaded. Either way, it’s an indication of poor system performance.

--Norm

Vacuum Systems

To Norman Lieberman: Norman. We met in India last year. I have a serious problem. Our vacuum on our asphalt vacuum tower is always bad. But sometimes it suddenly becomes much worse. It seems to correlate with increased naphtha product from the seal drum of the vacuum tower overhead system. As a result, we cannot produce proper grade of asphalt for sales. Any advice would be greatly appreciated.

--Kumar

            Kumar – If the naphtha production rate is very high, the naphtha liquid level in the first stage condenser (or precondenser) may start to back-up. This reduces the surface area exposed to the condensing vapor, which consequently will increase the vapor load to the downstream ejector. However, if the naphtha liquid level rises to the bottom edge of the air or vapor baffle inside the condenser shell, then the vapor is trapped inside the condenser. Pressure will build inside the condenser shell until the pressure is great enough to push the liquid level down below the bottom edge of the air baffle, so that the vapor can escape out of the condenser.

            On one occasion, I found this problem to be caused by a high liquid level in the seal drum, which was a consequence of an erroneous seal drum level indication. Also, the seal drum naphtha pump had lost capacity due to a worn pump impeller wear ring. Hope this helps.

--Norm

Vacuum Systems

Mr. Lieberman – Which should be hotter – the vapor or liquid outlet of my vacuum tower pre-condenser? Currently the liquid outlet is approximately 20°C and the vapor outlet is 40°C. That difference is increasing and our tower top vacuum is becoming progressively worse.

--L.T. Lee

            To Mr. L.T. Lee – By design, the vapor outlet of a normal surface condenser (where the vapor outlet is located on the side of the condenser shell) is cooler by about 20°F-30°F. This is accomplished by an internal baffle located inside the tube bundle. It’s called the air or vapor baffle. It forces the vapor to flow down and then up across the tubes.

            There are two possible reasons for your problem. First this air baffle may be leaking. This is always caused by defective air baffle seal strips (please see my book, Process Equipment Malfunctions, McGraw Hill, 2011.

            Secondly, there may be condensate back-up due to inadequate drainage from the seal leg. Your seal leg may be plugging or your seal drum may be filling with corrosion products or sludge. Blowing out the seal legs with steam will help in this case. Perhaps the seal leg is sucking in air, due to a seal leg leak. Such a leak will prevent proper drainage through the seal leg.

            Regardless of the cause, an increase of the vapor outlet temperature is certain to cause a loss of vacuum and could possibly result in the jets making a surging (i.e., erratic) sound, which is a certain indication of a loss in the sonic boost of the jets.

--Norm

Vacuum Systems

     Dear Dr. Lieberman – I have purchased many of your books, in which you have kindly invited questions. My question pertains to our new vacuum tower operation. Every morning about 10:00 AM, our vacuum ejectors start to make a regular sound. The operators say the ejectors start to breathe. At the same time our flash zone pressure increases from 10 to 12 mm of mercury to about 20. Any suggestions you may offer would be sincerely appreciated.

--Amed

     Amed – Your jets are surging. Most likely, because of higher discharge pressure. The cause of the higher discharge pressure is probably an increase in the plant’s cooling water temperature as the sun comes up. When I say your jets are surging, I mean that they are losing their sonic boost as described in my book, Process Equipment Malfunctions.

     To start with, try back-flushing the condenser water side on the discharge of the surging jet. Unfortunately, there are dozens of different things that can cause jets to surge. But anything you can do to improve water flow through the downstream condenser is sure to help to some degree.

--Mr. (not doctor) Norman Lieberman

Vacuum Systems

Norm – I was reading Troubleshooting Process Operations. In your chapter on vacuum towers, you talked about gas composition of the sea drum off-gas. I sampled this stream. It is 40% hydrogen sulfide. Is this possible? My boss said that I have done something wrong.

--Raymond

            Raymond – First, let me warn you to be careful. One breath of that off-gas will knock you out. Yes, I’ve measured up to 40% myself in Aruba, when the vacuum tower bottoms was about 4% sulfur. There is not much to be done to reduce the H₂S concentration. More typically, I see off-gas in the 10%-20% H₂S range. I hope this helps with your supervisor.

--Norm

Vacuum Systems

Mr. Norman Lieberman: I attended your Troubleshooting Seminar in 2001 in Ft. McMurray. I have a question relating to my vacuum tower (I’m now working in a refinery on the Gulf Coast). What does it mean if ice forms on the outside of the ejector body?

--Thank you for your help. I really enjoyed your seminar. Frank.

           

Frank – This is a sure indication of water in the motive steam supply. Not just a little water, but probably over 10%. If you can eliminate this moisture, I’m sure that you will see an immediate improvement in your vacuum tower performance.

--Norm

Vacuum Systems

Dear Mr. Norman – Our vacuum tower top pressure is 40-45 mm of mercury. We have two large vacuum jets on top of the tower. The front part of both jet bodies is around 80°F.  The steam supply is 150 psig and 360°F. The temperatures are obtained using my infrared gun. Is that 80°F normal, or good, or bad?

-- Henry R.

           

Henry – The 150 psig, 360°F motive steam may be saturated or it may be wet. When steam expands through the ejector nozzle, its temperature is converted to velocity. The more efficient the conversion, the higher the velocity of the steam. The greater the velocity of the steam, the more efficient the compression of the vacuum tower off-gas. Hence, in that sense the cooler the mixing chamber (i.e., the front end of the jet), the better.

            However, if the mixing chamber is cool, that could be due to moisture in the supply steam, flashing as it enters the mixing chamber. This conversion of latent heat to sensible heat just robs energy from the motive steam and thus slows down the motive steam as it enters the diffuser. And this is bad.

            If the part of the diffuser immediately downstream of the mixing chamber doesn’t get too hot to touch within one or two feet of the mixing chamber, that’s an indication of wet steam.

--Norm

Fired Heaters

Mr. Lieberman, According to your book, “Troubleshooting Process Operations,” the correct place to locate an O₂ analyzer is just below the bottom row of convective tubes. Is this correct?
--Tony, Louisiana

            Tony – Best to adjust excess air to max heater outlet °F, or to minimize fuel use. On this basis, an O₂ analyzer is not needed to control air. Then, if an analyzer is used, locate it above the convective box to help monitor tramp air leaks in the convective box.

--Norm

Steam Turbines

Dear Norm – From an operator’s point of view, what is the best way to operate a steam turbine? What are the two valve handles on either side of the speed governor for? How do we operators know what speed to set the turbine to operate at? Our engineer just tells us to operate at the turbine’s design speed. Is this right?

--James, Oklahoma

            James – Operate turbine speed to keep downstream process control valve in a mostly open but still controllable position. Alternately, connect the governor directly to the process variable to be controlled. That is, control an upstream level by directly varying turbine steam supply flow. The two valves you refer to are called horsepower or port valves. If the governor was a delta P of more than 5% of the motive steam pressure, close one of the port valves until the governor is mostly open.

-Norm

Steam Turbines

Mr. Lieberman – I read in your book, “A Working Guide to Process Equipment,” that the steam chest pressure of turbines should not be less than 10 or 20 psi of the supply pressure. But how to measure the steam chest pressure when a turbine is in operation?

--Prabu, India

            Prabu – If there is no connection on top of the steam chest, then you may use the small (½”) condensate drain to the sewer from the bottom of the steam chest. I just place my gauge on this drain. Closing a port valve will cause this pressure to increase which will save steam. Run at the maximum steam chest pressure by closing port valves or until governor is 100% open. Good to hear from you again.

--Your friend, Norm

Steam Turbines

To Norman Lieberman – Can we safely operate a topping steam turbine if the governor speed control valve is out of service and the over-speed trip is working correctly?

--Carl, Texas

            Carl – No. If you are controlling the turbine speed manually, then a reduced load to the turbine will cause the turbine to over-speed and trip. Sooner or later, operators will tire of the repeat trips and “wire-up” the over-speed trip mechanism (i.e., disable the over-speed trip protection). It’s an industry wide problem that has resulted in injuries and death. The only answer is that both the governor and trip must both be working. See my book, “Troubleshooting Process Plant Control,” Wiley Publications.

--Regards, Norm

Rankine Cycle

Hi Norm – I am an ME and working with an associate who is not technically inclined. I am trying to provide him a layman’s explanation as to why the waste heat in a condensing turbine is unrecoverable for process uses and why a topping turbine is so much less efficient at making electricity than a condensing turbine.

            Do you know of anything that explains these principles in simple terms or with an illustration?

--Kerry, Florida

            Kerry – When steam expands into a vacuum, its volume gets a lot bigger. At a bigger volume, the steam’s velocity is much greater. It’s the velocity of the steam striking the turbine wheels that spin the turbine. Like the wind causing a windmill to turn. The faster the steam or air, striking the turbine blades, or the air striking the windmill sails, the more work that can be extracted (or electric power, which is work) from the spinning turbine.

--Norm

Crude Unit Operation

Mr. Norman Lieberman – I have an old seminar manual from 1996. You have explained optimizing crude desalter operation – but do not explain optimizing the desalter temperature. Is there a single value? I’m trying to train a young process engineer assigned to our crude pipe still.

--Bob, Louisiana

            Bob – Anything I wrote in 1996 is outdated. I’m much smarter now. There is no single optimum desalter temperature. A desalter outlet temperature of 300°F may damage the desalter electrical components. I’m sure anything below 220°F is too cold.

            Raising the desalter temperature will cut viscosity and s.g., which will reduce height of emulsion and brine (BS & W) carry-over into deslated crude. However, the increased solubility of water in hotter crude will increase desalter amps and water content of desalted crude. The extra water in fractionator feed will promote corrosion in top few trays of the tower due to H₂O condensation.

            Another big problem is vaporization in the desalter. I have one client in Indiana who runs their desalter at 45 psig and thus cannot increase desalter temperature much over 220°F without stirring up the contents with vaporization of steam and light hydrocarbons. Keep the desalter inlet at least 10°F below the calculated bubble point of the crude taking into account that it is saturated with water.

            A typical desalter outlet temperature for heavy (18°API) Venezuelan crude is 285°F. A typical desalter temperature for a light, sweet crude (35°API) is 245°F. If you have excessive growth in the emulsion layer, a warmer desalter will be optimum. If BS & W is less than 0.1 wt%, a lower desalter temperature may be optimum. Hope this longish answer helps.

--Regards, Norm

Crude Unit Operation

Dr. Lieberman – Here is Nehi. Should we add NaOH to desalt effluent?

--Best Greetings – Nehi, India

            Nehi – Add caustic to crude effluent from desalter is normal. You should minimize caustic addition. I use two methods. One way is to add NaOH based on ⅓ MgCl₂ content of raw crude. That is, about 65% of MgCl₂ is extracted with brine and the remainder hydrolyzes to HCl. Second method is to reduce chloride content of crude tower overhead drum boot to 10 to 15 ppm. Too much caustic is bad. Regardless never use more than 3-5 ppm of NaOH of crude. Spent caustic from a mercaptan extraction unit can be used as well as fresh 10-15 Be NaHOH.

--I’m not a Doctor – just Norm

Crude Unit Operation

Hi Norm – For a crude desalter, what is optimum mix valve delta P? Our crude is 25°API, we use 6 wt% water and desalter temperature is 250-270°F.

--Carl, California

            Carl – I used to think 15-25 psi. But, now I know better. Let’s say you have NaOH in the wash water and your brine pH is nine, rather than an optimum 6½. Thus, you have the desalter interface between H₂O and crude filled with several feet of emulsion so that you’re getting water carrying over into your crude heater. You then need to back way off on your mix valve delta P to reduce emulsion carry-over. It’s senseless to have a 20 psi mix valve delta P if it promotes excessive emulsion carry-over. In my experience, the most important controllable variable in desalter operation is pH control of the brine. Temperature is also critical but not often controllable. You can call me at 1-504-887-7714 for an in-depth discussion.

--Norm

Crude Unit Operation

Dear Norman – I enjoyed working with you last month. A follow-up question. How come our crude tower floods with such a tiny increase in delta P?

--Sam, Washington State

            Sam – Fe(HS)₂ deposits accumulate on the upper few trays of the crude tower, between the jet draw and the second to top tray. If fouling causes flooding starting at the top few trays, jet will entrain overhead. However, flooding progresses up a tower and not down. Thus, in your 40 tray tower, only the few trays flood, but delta P remains low. I recall explaining all this to you before, but I guess you were distracted by a text message from your girlfriend. Incidentally, iron sulfide deposits are insoluble in water and hence cannot be water off on-stream.

--Norm

            P.S. You’re running your tower top temperature 20°F too cold.

Crude Unit Operation

Norman Lieberman – What is a reasonable salt content in the desalter effluent? Our chemical vendor says 90% salt removal in our single stage desalter is good. Are desalting chemicals really helpful?

--Pranab, India

            Pranab – Desalting dispersant chemical is truly helpful. However, it is best added not to the desalter but to the crude charge tank to remove brine upstream of the desalter. I’m referring only to the oil soluble chemicals, not water soluble dispersants.

            The 90% salt removal is not the best measure of desalter efficiency. Most of your salts (90%) are NaCl which is not subject to hydrolysis to HCl. Best just to measure the MgCl₂ cone in desalter effluent which is the only salt in crude that will mostly hydrolyze at crude unit heater outlets of minus 700°F. CaCl₂ also hydrolyzes, but typically in the hotter vacuum tower heater outlet.

--Norm Lieberman

Sat and Unsat LPG

Dear Norm, Why have two different LPG merox units – sat and unsat in refineries?

(1) In our configuration the straight run LPG from the crude unit is going to SAT MEROX UNIT – and then to product tanks.

(2) LPG from FCCU and COKER is going to UNSAT MEROX UNIT – and then to C3/C4 splitter – then to propylene/propane splitter.

Reason is that only the LPG from FCCU and COKER is having propylene in it (which is formed during cracking).

Is this right?

--Prabhakar, India

            Dear Prabhakar, I assume you are talking about a Merox Extraction plant, not a sweetening plant. You can split cracked C3-C4 first and then mix sat and un-sat C3’s to extract mercaptans. Propylene is irrelevant. Obviously, cracked P-P’s are 50% propane anyway.

            I’m coming to India later in the year. I’ll let you know when and where.

--Norm

Distillation

Dear Mr. Lieberman – Which is best tray for good turn-downs? Our depropanizer has to operate at very variable rates.

---Roja, Venezuela

Roja – I assume this is a new tower you’re designing. Bubble cap trays are best. Your tower I.D. will be about 10% bigger than if you used a modern grid tray or valve tray. Bubble caps will operate efficiently at a far larger range of vapor flow than any other sort of tray.

---Norm

Coker Heater

Dear Mr. Lieberman, What should be the maximum oxygen content we can maintain coker heater stack? I heard that some of the shell cokers even operate more than 8%. What is the reason?

---Prabhakar, India

Prabhakar, the higher the excess O₂, the cooler the radiant firebox. The reduction in radiant heat density is offset by the increased convective section duty. The lower radiant heat transfer rate results in reduced radiant tube coking and longer heater run lengths. Of course, energy efficiency suffers.

---Norm

Centrifugal Pumps

Hello Norman Lieberman. We have a big flashed crude oil heater charge pump. It pumps flashed crude from the flash drum. When we start-up this pump (sometimes we bypass the flash drum and pump), the pump cavitates. Meaning, it’s hard to get the pump to take suction at first. I throttle way back on the discharge valve. I raise the level in the flash drum to maximum. Still the pump cavitates when we first start-up. Eventually we do get it running. Then it runs find, if we have not damaged the pump seal during all the start-up tries. Any ideas?

---Henry, Los Angeles

Henry – Your problem is lack of “Starting NPSH.” When a pump is started, the liquid in the suction line has to be accelerated from stagnant conditions to several feet per second. That energy required comes from the available NPSH needed to satisfy the pump’s steady state NPSH requirement. Hence the observed cavitation. To provide the starting NPSH, quickly raise the flash drum pressure. This will instantly increase the pressure at the pump’s suction. The vapor pressure of the liquid will also increase by the same amount. But it will take time for the more volatile liquid to flow down into the pump suction. During this short period, you will have a surge of available NPSH.

---Norm

Fired Heaters

Norm. I have attended your seminar many years ago. My question is why do my operators say that they don’t have enough air, yet the air registers at the bottom of the heater are halfway closed. The stack damper is always left open. We could run more crude if we had more air to the heater. Hope you remember me.

--Robert, Indiana

Robert. The problem is not lack of air, but lack of draft. If you open the air registers further, the delta P of the flue gas thru the convective section tube banks will increase and you will develop a positive pressure below the lowest rows of convective tubes. This is dangerous. Hot flue gas with SO₂ can blow-out of the upper part of the firebox. Your operators are right to restrict air flow thru the air registers. Is your stack damper truly wide open? Have you tried cleaning the outside of your convective tubes? Have you used your steam soot-blowers regularly? Make sure all tramp air leaks in the convective section have been sealed. Unfortunately, I don’t remember you at all.

--Norm

Fired Heaters

Mr. Lieberman – I have read that draft at the top of the firebox should be 0.05 to 0.10 inches of water to minimize tramp air leaks into convective tube banks. We cannot seem to accomplish this. It seems like the draft is too erratic. It jumps around by itself. My boss wants to install auto stack damper controls to stabilize draft. Is this going to help?

--Dan, Wyoming

Dan – Erratic draft is caused by the wind blowing across the top of the stack. Of course draft will also vary with stack temperature, but not very fast. Gusts of wind will create very rapid changes in draft of 0.1” H₂O to 0.2” H₂O over just a few seconds. I can assure you, based on personal experience, that automated stack dampers will not work.

--Norm

Fired Heaters

Mr. Norman P. Lieberman. Dear Sir, what causes fins in the convective section to break off? This is a big problem for us. I have watched your video on fired heaters, but you did not explain this problem.

--George, Texas


George, when the fins break off, are they rusty or covered with a sticky, gray paste type deposit? The sticky paste is iron sulfates. That’s caused by a combination of high sulfur fuel and low flue gas temperature. Sulfuric acid is formed when sulfur is burned and precipitates out of the flue gas as follows:

• 4% sulfur in fuel oil --------- 450°F
• 100 ppm H₂S in fuel gas ---- 300°F

The acid attacks the fins and forms the sticky iron sulfates.

The rust is caused by afterburn. That is, the fins are burnt-up. The way to avoid this is to add sufficient combustion air to avoid CO, alcohols, ketones, etc. from escaping from the firebox, mixing with convective tramp air leaks, and reigniting. Hope this helps.

--Norm

Fired Heaters

Mr. Lieberman, Process Improvement Engineering. Thank you for your books. I write to ask you a question. What is best in a heater to increase heater run length – very much air or very little air?

--Garcia, Colombia


Garcia – More air will certainly increase your heater run length. It will keep your fire box cooler, but also wastes energy. Optimizing the air to fuel ratio must always minimize heater run length and maximizes tube coking rates. It’s sad, but that’s just the way it is.

--Norm

Fired Heaters

Hello Dr. Lieberman. I am Pabu from India. We blew steam up our stack to get more draft. Does this not waste steam? Our heater is small and our stack is short. The structure may not support a taller stack, which is costly. Be kind and advise us.

--Pabu, India


Pabu. My cost effective method to develop more draft is certainly not to use steam. That produces more draft because the molecular weight of steam is 18, and the molecular weight of flue is 30. Best approach is to locate tramp air leaks in the convective box and ducts. Seal them with aluminum duct tape. Eliminating cool air will increase the temperature of the stack, reduce the flue gas density, and thus increase the draft. I have increased draft by as much as 0.1” H₂O by sealing off such leaks. This also may increase heat recovery in the convective section and save energy.

--Regards, Norm

Fired Heaters

To Elizabeth and Norman. This is Earl from Bakersfield. We blew-up a heater and injured seven operators after a tube leak on a vacuum heater. How could we have prevented this accident? We had tried to use snuffing steam on a previous tube leak incident, but it blew the burning oil out of the box.

--Best Regards. Earl, California


Earl. Snuffing steam is absolutely necessary to preclude an explosive mix from forming when the leaking tube passes are blocked-in. It reduces the air that could cause the explosive mixture. Your problem is that you are not maintaining the steam tracing, insulation, and steam traps on the snuffing steam supply lines. Blowing water into the fire box will blow the fire out of the box and is very dangerous. I’m speaking from personal experience at the Good Hope Refinery in Norco, Louisiana. As you block in the leaking tube, add steam to the fire box, but it has to be reasonably dry and free of water.

--Norm. Regards from Liz.

Distillation

Hello Mr. Lieberman – I met you at your course in Singapore in 2000. How can this be? I have a deprop. that has sieve trays. It worked very well. All we did was to change from 2 mm to 4 mm tray deck thickness because the trays were getting thin. Nothing else. Now, the tower will not produce on spec. LPG propane at the reflux rate used before turnaround. We must use 30% more reflux for same feed.

--Rao, Singapore


Rao – Your problem is low vapor delta P thru sieve holes. While the vapor velocity thru sieve holes is constant, the orifice factor decreases as the tray thickness increases. The effect is surprisingly large. Vapor delta P varies linearly with the orifice factor.

--Norm

Distillation

Dear Mr. Lieberman. I have an MDEA fuel gas scrubber to remove H₂S. What can I do to stop the amine from carrying-over as I increase the feed gas rate? I have reduced the amine circulation rate, but this has not helped. It even made the problem worse and increased MDEA losses.

--Linda, Alberta

Linda. Reducing amine circulation rate increases acid gas loading and makes the amine foam worse. Reducing amine strength would help, but this also makes amine foam-up. Best to switch from MDEA to DEA, which foams less at the same molal concentration. Improve filtration to reduce particulates. Proper oil skimming and control of heat stable salts is vital. Maintain lean amine 15°F hotter than sour fuel gas. However, never add NaOH to reduce heat stable salts. This will only make your problem worse. If you have ⅜” sieve trays, increase size of sieve holes to ½”. Perhaps your downcomer open area is too small. Increasing tray spacing will always help. Most absorbers have two to five times the number of trays required.

--Norm

Distillation

Norm. What does this mean? I have a distillation tower that I think is flooded. I raise the reflux, yet fractionation does not improve. Does this not indicate flood? But the overall delta P, which is constant and really low, does not increase as I increase the reflux flow.

--Neal, UK

Neal – Flooding progresses up a tower, not down. If the top trays are fouled, then only they will flood. The reflux flow will just recirculate back to the reflux drum and not increase vapor flow at all. Fouling of the top trays is common due to corrosion of the overhead condensers and the sublimation of volatile salts (NH₄CL or NH₄(HS)₂.

--Norman

Distillation

Mr. Lieberman. I am an operator in a refinery. My tower must operate at the same reflux flow at the small and big feed rates. Why can’t I reduce the reflux by half when I reduce the feed rate by half?

--Vaidus, Lithuania

Vaidus – I imagine that you have valve, sieve, or grid trays rather than bubble cap tray decks. Likely your trays are not level. You then have to keep a high vapor rate to generate enough vapor delta P to keep enough pressure below the tray decks to prevent liquid leakage at the sagging portions of the trays. Once liquid leaks through such out-of-level trays, fractionation efficiency will be ruined. The high vapor flow is generated by high reflux flows.

--Norm

Distillation

Dear Norman – I attended your seminar in Toledo in 2008. Is it okay to have the outlet weir of a tray lower than the bottom edge of the downcomer from the tray above?

---Carl

Carl, if the crest height plus the weir height exceeds the downcomer clearance, this should be okay. It’s called a dynamic seal. To calculate the crest height in inches of clear liquid:

0.4 (GPM per inch of weir)^0.67

GPM is hot U.S. gallons per minute

---Norm

Crude

Norman – Can crude LPG contain RSH or are mercaptans only found in cracked materials. Our crude has a lot of H₂S, but can it have RSH also?

--Probakar, India

Probakar, RSH exists in virgin materials – often in very high concentrations. If you remove all H₂S, then the RSH can break down in the KOH dryer to form traces of H₂S and cause LPG to fail copper strip corrosion test. Methy mercaptans are easily extracted from LPG with caustic. Do not pay licensing fee to UOP, etc. for NaOH regeneration. This is all open-art and is no longer covered by U.S. patents.

In cracked materials, you may also have COS and disulfides. This is a different problem.

Do you remember when we did our pressure survey on your vacuum tower in a (-25° wind)?

--Norm

Condensers

            Mr. Lieberman – Every summer we run out of condenser capacity. My problem is hardness

            fouling deposits in the tubes. But I cannot shut-down. What do you recommend?

 ---Harry, England

Harry – Step 1 is to blow air or N₂ into the cooling water inlet. Do this every day for ten minutes. It helps prevent accumulations of hardness deposits. Step 2 is to back-flush the exchanger. The bigger the back-flush valve the better. Step 3 is acid cleaning. This is done on-stream using inhibited HCl (10%). Carefully neutralize acidified water with NaOH as it leaves the condenser.

---Norm

Reboilers

To Norman Lieberman – My alky DIB tower reboiler stalls-out. It just stops working. Meaning, it stops adding heat to the DIB. What is my problem?

--Rob, pane operator, Texas

Rob – The bottom tray of your tower is leaking. If you have a line connecting the bottom product line and the reboiler inlet line, raise the bottoms level to max and open the gate valve to this line. During your next turn-around, replace the bottom tray with a welded total trap-out chimney tray. The other, not so good way to restore thermosyphon liquid circulation, is to back liquid to the reboiler vapor return nozzle.

---Norm

Fired Heaters

Mr. Norm Lieberman. At the seminar you taught us last year, you talked about adjusting excess O₂. My boss says 3%-4% should always be our target. Is this right?

---James, Ohio

Jim – I remember that you asked this same silly question at the seminar. Excess O₂ must be adjusted without reference to an arbitrary target. Adjust your excess O₂ (assuming you’re on TRC) to minimize fuel gas consumption to reach the desired heater outlet temperature. This is best done on the stack damper unless your draft in the fire box is less than 0.05”-0.10” H₂O. If less than this range, pinch on your air registers.

---Norm

Compressors

Norm, why is it normally best to suction throttle, rather than to spill-back, to control my FCU wet gas compressor suction pressure in the fractionator reflux drum?

--Kumar, India

Kumar, I assume you have a motor driven, fixed speed compressor. If you had a variable speed compressor, you would simply slow the compressor. When we suction throttle to hold back-pressure on the reflux drum it will start to force the spill-back valve to close. This does two things:

1.      Total number of moles of wet gas compressed will drop.

2.      Compression ratio (due to the suction throttle valve delta P) will increase.

As long as you are on the flat portion of the flow vs. head compressor performance curve, the motor amps will decline.

--Norm Lieberman

Distillation

Dear Norm – How can I tell for sure my debutanizer trays are flooding without running an expensive TRU-TEC radiation scan?

--Fred, Texas

Fred – A delta P survey is not always meaningful unless flooding is starting at the lower trays. Thus, with the reboiler on auto, increase the reflux. The tower bottom minus top temperature should increase. If this doesn’t happen, then the tower is losing tray efficiency due to entrainment. A loss of tray efficiency with increasing vapor-liquid traffic is due to jet flood.

---Norm