INCREASING COOLING WATER FLOW THRU AN ELEVATED CONDENSER OR COOLER

Clare! Let’s open the cooling water outlet valve to get more water flow.

No, Carl! The Condenser is 60 feet above grade. The pressure at P₁, is under vacuum! Opening that valve will give us less cooling water flow!

NO! Opening a valve will always increase flow!

Sorry, Carl! Opening that valve reduces the pressure at P₁, further below the atmospheric pressure. This causes the air to flash-out of the cooling water, which chokes back water flow!

Clare! WRONG! I’m really smart! Anyway, where’s the test to prove you’re right?

OK. I’ll close the valve and you’ll see the temperature at T₁ will go down. But don’t close it too much! Otherwise, you will throttle the water flow. Then, T₁ will get hotter!

But Clare! How do I know how to adjust that stupid valve?

Carl, dear! Set the valve to hold a backpressure of about 3″ Hg. That’s minus 0.10 atmosphere. At 100 °F, that will stop air evolution from the water, but not throttle the water flow too much!

HOT VAPOR BY-PASS PRESSURE CONTROL

Clare! Close the hot vapor by-pass valve! We need to lower the tower pressure. Do it now!

Sorry, Carl! When I closed the valve the tower pressure went up … not down!

No, Clare! Closing the valve will cool off the reflux drum! The pressure at P2 will drop, and draw down the pressure at P1. Understand?

But Carl! How about the pressure drop across the air cooler? It increases as more flow is forced through it. True, the pressure at P2 will always fall! But the pressure at P1 may go up or down—depending on the air cooler DP!

But, but …? Closing the hot vapor by-pass is supposed to lower the tower pressure, according to my design manual!

But suppose the tubes get full of salts and scale? Then what? Also, Carl, we now have a vacuum in the reflux drum, which can be quite dangerous! Air could be sucked into the drum and an explosive mixture could form! Don’t forget there’s pyrophoric iron sulfide deposits (Fe(HS)₂) in the drum! They’ll auto-ignite at ambient temperatures!

STALLING A THERMOSYPHON REBOILER

Clare! Open the steam supply valve! Quick! We need more reboiler heat. The reflux drum is going empty!

Sorry, Boss! That won’t help! The Once-Thru Thermosyphon Reboiler is STALLED OUT!

Clare! More steam flow will have to give us more heat to the reboiler! Open that valve!

Opening the steam valve will not increase steam flow when the reboiler is STALLED-OUT!

STALLED-OUT? What does that mean?

Stalled-out means heat duty is limited by the process flow to the tube-side of the reboiler! The process flow rate to the reboiler is real low now and limiting the steam condensation rate!

How do you know that, Clare? Do you have X-ray vision?

Carl! Look at the reboiler outlet. It’s 450°F! The tower bottoms are only 330°F. Most of the 300°F liquid from tray #1 is leaking past the draw pan, and dumping into the bottoms’ product!

OK, Clare, OK! But still, the steam inlet valve is only 50% open! Won’t opening it 100% help some?

No, Sir! The pressure at P1 on the steam inlet line is 500 PSIG! The same as the steam supply pressure. There is zero DP across the steam supply valve. The valve position, with no DP, is IRRELEVANT!

I guess we should have used a total trap-out chimney tray for tray #1! I remember you suggested that last year, Clare. Perhaps you’d like a transfer to the Process Design Division? They would probably love to have you! I remember that in the old days we had bubble cap trays, which could never leak and cause this loss in thermosyphon circulation, or stalling-out.

OPTIMIZING FRACTIONATOR PRESSURE

Clare! The best way to optimize tower pressure is to target for the lowest pressure!

Why is that, Carl?

Because, Clare, as we learned at university, the lower the pressure, the greater the RELATIVE VOLITILITY between propane and butane!

But Carl! Suppose the lower tower pressure causes entrainment? Then, a lower pressure will reduce tray separation efficiency and make fractionation worse!

Well! What do you suggest? It takes too long to wait for lab sample results.

Carl, I suggest:

1. At a constant reflux rate, start lowering the fractionator pressure.
2. Now, watch the delta T (T1 - T2).
3. That tower pressure, that maximizes delta T, will give the best split between butane and propane. But make the moves slowly!

Clare! You really should take a more positive attitude towards your engineering degree, and show more respect for the principle of relative volatility!

ADJUSTING STEAM TURBINE SPEED TO MINIMIZE STEAM CONSUMPTION

Clare! Always run steam turbines at a constant speed! In the U.S., 3,600 rpm, Europe 3,000 rpm!

I’m sorry, Carl! But I don’t agree!

Exactly what’s your problem? All process plants run their turbines at 5% below their maximum rated speed. That’s best!

Carl! The best way to set the speed of the turbine is to slow the turbine down until the control valve on the discharge of the pump it’s driving is in a mostly wide-open but still controllable position!

All to what purpose?

Well, Carl, for each 3% speed reduction, the steam required to drive the turbine will fall by 10%. Work varies with speed cubed: W ~ (speed)³ … That’s the Affinity Law!

Thanks, Clare! I’ll write new instructions for the operators!

Carl! Actually, you can get rid of the control valve on the discharge of the pump, and run just on governor speed control, to adjust the upstream level or the pump discharge flow and pressure. That will save even more of the motive steam.

STEAM CONDENSATE DRAINAGE FROM REBOILERS BLOWING CONDENSATE SEAL

Clare, listen up! Open that condensate drain valve more. We need more heat to the reboiler!

Sorry, boss! I think opening the condensate drain valve more will reduce reboiler duty! I’m not too sure!

Not sure? Can’t you just follow my instructions for once? It’s getting late!

Carl, here’s the problem! If I open the drain valve too much, we’ll blow the condensate seal. Steam will blow through the reboiler tubes, without condensing. If I open the valve too little, we will suffer from condensate back-up!

I’m confused! Then how do we know whether to open or close that stupid condensate drain valve?

Well, if we close the valve, and T1 goes up, it means we were previously blowing the condensate seal! If we close the valve and T1 goes down, it means we were previously suffering from steam condensate back-up!

OK! Let’s optimize the drain valve position to maximize the reboiler outlet temperature, and go to lunch. I’m hungry! Seems like it’s all a balance between condensate back-up and blowing the condensate seal. Kinda like eating too much or too little.

EFFECT OF REFLUX ON FRACTIONATOR TOP TEMPERATURE

Clare, dear! To lower the tower top temperature, one should always raise the reflux rate. It’s a basic idea of Process Control!

Sorry to disagree again, Carl! It ain’t necessarily true!

Now Clare! Read any distillation textbook. They all agree with me! What’s your problem?

It’s this: If the top tray is at its flood point, raising top reflux must increase reboiler duty, because the reflux comes from the reboiler. This will increase vapor flow to the top tray!

So what? The reflux will cool off that extra vapor! The reflux will knock back the heavier components in the vapor. What’s your problem, Clare?

What you say is true, boss. Up to a point. The INCIPIENT FLOOD POINT! Above that point, extra vapor promotes ENTRAINMENT. The result is droplets of heavy liquid blow up through the trays, increasing the boiling range of the top product and temperature at the top of the tower.

That’s BAD! Because then the tower top temperature will go up, as the reflux rate is increased. Worse, the top reflux rate, and reboiler duty will then increase more, and make the problem even worse!

Yes, Carl! I call this getting caught up in a POSITIVE FEED-BACK LOOP! The reflux needs to be switched to manual and partly closed to break the feed-back loop.

CENTRIFUGAL PUMP HEAD VS. FLOW PERFORMANCE CURVES

Clare! Where did you get that pump curve from? It’s wrong!

From plant data on the new giant vacuum tower residual pump. I plotted observed flow vs. pump discharge pressure!

Well! It’s contrary to the manufacturer’s curve! The head and flow are both going down, at a low flow. Impossible, I’d say!

The pump is never supposed to operate at such a low flow. But, Carl, you purchased this oversized pump yourself!

Oh! Well then, I guess it’s OK to run it below point “A”, as we have lots of excess head anyway!

No, Carl! It’s not OK. Below point “A”, the pump vibrates in a most alarming manner!

No need to mention this to Carl … but Norm first saw this in 1991, at the Coastal Refinery in Aruba! Norm still complains to this day that those pump vibrations loosened the fillings in his teeth! Imagine what that did to the pump’s mechanical seal? The pump had a minimum flow spill-back, but when Norm opened it, the pump lost suction and cavitated is a most alarming manner.

CONDENSATE BACK-UP IN CONDENSERS-THE EFFECT OF SUB-COOLING

Clare! Let’s get condenser “B” cleaned. Look how high its outlet temperature is!

Actually Carl, it’s “A” that’s really underperforming!

But the outlet flow of “B” is hotter than “A”?

Well, that’s true! But only because “A” is suffering from CONDENSATE BACK-UP and sub-cooling! About 40% of the tubes in “A” are submerged in liquid, but only 10% of the tubes in “B” are covered in liquid!

So, Clare, I suppose that now you have X-ray vision too? How can you know that “A” is suffering from condensate back-up?

Just run your hand along that channel head cover! You can feel the temperature gradient; where the channel head feels cooler, is the liquid level. Tubes covered in liquid, can’t condense any vapors! Because the vapor is not even in contact with any of the tubes.

You know Clare, this kind of reminds me about the condensate back-up problem we had in the channel head of our steam reboilers! Kinda the same process problem!

DISTILLATION TRAY DOWNCOMER BACK-UP AND LIQUID FLOODING

Clare, let me explain what causes downcomers to flood! It’s because the downcomers are too small! That is, downcomer loading exceeds 175 GPM per square foot of downcomer cross sectional area!

I certainly agree, Carl! But there are many other reasons for downcomer flooding. Shall I explain?

Such as?

Well, Carl! Such as loss of the downcomer seal! If the weir height is adjusted wrong, it may be lower than the downcomer clearance! Then the downcomer seal would be lost! Vapor would blow up the downcomer and prevent liquid from draining out of the downcomer!

Easy enough to prevent! Just keep the bottom of the downcomer real close to the tray deck below. No problem!

No, Carl! Restricting the downcomer clearance will cause too much head loss under the downcomer, and also increase downcomer back-up! Also, if tray #1 gets too dirty, the flowing vapor pressure drop through tray #1 will go up. This will push up the liquid level in downcomer “A”! And Carl, if tray #2 leaks really bad, the bottom of downcomer “A” will become UNSEALED! This would also cause vapor to blow up downcomer “A” and retard liquid drainage from tray #1!

I guess this would cause tray #1 to flood. That’s bad!

Not only tray #1! With time, all the trays above would also flood!

OK, Clare! Just be careful, when you inspect the tray installations, that you have a ″ overlap, between the top edge of the weir and the lower edge of the downcomer from the tray above! And make real sure the weir and bottom edge of the downcomer are LEVEL!

EFFECT OF FOAM ON LEVEL INDICATION IN DISTILLATION TOWERS

Clare, I’ve just checked the level in the tower. It’s 2 ft. below the reboiler return nozzle. It’s fine!

Actually Carl, it’s too high! I believe the high tower bottoms level is causing the tower to flood!

But can’t you see, Clare, that the level in the gauge glass is quite low? Look at the level, woman! Use your eyes for a change!

But Carl, the level in the tower is higher than the level in the glass because the specific gravity of the liquid in the glass is greater than the specific gravity of the FOAM inside the tower!

How can you tell that? I don’t see any foam in the glass!

It’s simple, Carl. I lowered the level in the glass by 8″ and the tower Delta P dropped by 20%! Understand now?

No! I’m totally confused!

Lowering the external liquid level by 8″ may have reduced the internal tower level

(i.e., the foam level), by 16″! That is, below the reboiler return nozzle. This then stopped flooding of the trays, due to a high foam level! Norm calls this: FOAM INDUCED FLOODING! Do you remember Norm? He bought us lunch last month!

What we really need, then, in foaming services, is radiation type level measurement, which we could calibrate, for whatever foam density we wanted! Then we could know the real level for certain! I think the k-ray company sells a neutron back-scattering device to measure levels and foam densities.

SPLIT LIQUID LEVELS IN VERTICAL VAPOR - LIQUID SEPARATORS

Look, Clare! Split liquid levels indicate that we have a vapor and liquid sandwiched together! Liquid-vapor-liquid-vapor!

Not really, Carl! It’s a sign of foam!

Foam? But why, dear woman, do we see three different levels, in the three glasses?

Because, Mr. Carl, the lighter, frothy foam, floats above the denser foam, which floats above the clear liquid. You can observe the effect in a glass of beer!

Then what does the level in the gauge glass represent? I’m confused!

The density of the foam between each level tap. If the gauge glass is 1/3 full, the foam density is 33% of the density of the clear liquid in the gauge glass!

But what happens Clare, if the foam rises above the vapor-liquid inlet nozzle?

MASSIVE ENTRAINMENT! TRAY DECK FLOODING!

OPTIMIZING EXCESS AIR IN A FIRED HEATER TO MINIMIZE FUEL CONSUMPTION

Oh, Clare! Please cut back on the heater air register! We have 4% O₂ in the flue gas, and our target is 2-1/2%!

Not a good idea, Carl! The 4% is a minimum for safe operations for this heater, at this time!

But Clare, our heater expert says the optimum excess of O₂ is 2%-3%. Do you think you know more than the heater expert too? Don’t be such a know-it-all.

The optimum O₂ is a function of burner air-fuel mixing efficiency! The burners are in bad condition, and their mixing efficiency is bad!

But still, we’ll save some fuel gas, at the lower excess air?

Carl! If we fall below the optimum O₂, then the heater outlet temperature will go down, as fuel gas oxidation is suppressed, and fuel gas reduction and reactions are favored!

Oh, yes! Please elaborate, Clare!

Oxidation is an exothermic reaction, which releases heat. Reduction is an endothermic reaction, which absorbs heat. If we start to favor a reduction reaction because, in this case, we have too little air, the heater outlet temperature will be driven down!

But then, the fuel gas flow will increase, which will favor the endothermic reduction reaction more and more!

Quite true, boss! This will drive the fuel gas regulator valve open more and more! The heater will be trapped in a positive feed-back loop!

If that happens, why, we’ll just open the burner air registers! We’ll open the stack damper! You know, to get more air.

This may possibly blow the heater up! At Norm’s former employer, two operators were killed by this mistake. It happened at their plant in Norco, Louisiana.

Hmm! I guess the optimum O₂ can only be found by trial and error. The optimum excess O₂, is that air flow that minimizes the fuel gas rate!

DISTILLATION TRAY DUMPING OR WEEPING WITH VALVE CAP TRAY DECKS

Clare! Thank God for the Koch brothers! They invented the valve tray cap, which prevents tray deck leakage and weeping!

In reality, valve caps are a fraud! They don’t work!

Watch your mouth, woman! That’s the Koch family you’re insulting!

Carl! I know you believe that at low vapor rates, the valve caps are supposed to act like little liquid check valves, and stop tray deck dumping.

Yes! That’s the idea. Isn’t this true?

No! FRI data shows that even when the tray decks are very level, valve trays are only marginally better than 3/8″ hole sieve trays!

Then, Clare, what does keep valve trays from dumping, if not the valve caps? Don’t forget trays have 2″ - 3″ high outlet weirs! There’s a couple of inches of liquid on each tray deck!

Yes, Carl! Also, the CREST height contributes to the weight of the liquid on the tray decks! It’s the Delta P DRY that prevents tray deck dumping and weeping!

Delta P dry? Clare, what does that mean? I’m getting confused!

It’s the pressure drop of the vapor, as the vapor accelerates through the perforations on the tray deck! Delta P dry should be about equal to the weir height plus the crest height, taking into account tray deck levelness, plus the tray aeration factor (typically 50%).

FIRED HEATER - TUBE FAILURES

In general, Clare, furnace heater tubes fail due to localized overheating. This melts the tubes!

That’s not quite right, Carl. The tubes don’t melt. They undergo HIGH TEMPERATURE CREEP!

Creep? What’s that?

Oh Carl! At 1380° F, high chrome tube walls become deformable. Then, the internal pressure causes the tube wall to bulge out!

I see! It’s like blowing up a balloon! The wall of the balloon gets thinner and thinner, as the balloon expands!

That’s right! The tube wall thickness at the bulge becomes too thin to constrain the tube internal pressure of the process fluid!

But what causes the local hot spot? Localized overheating? Flame impingement?

Not usually! Most often, it’s coke formation inside the tube, due to low mass velocity (less than 100 lbs / ft² / second), or feed interruptions, or left over coke from prior decoking operations during a turnaround! Velocity steam in the heater passes also helps protect the tubes from failure!

You’re right, Clare! But for Delayed Cokers, it’s best to keep a minimum velocity of 200 lbs / ft² / second! But you’re talking about coil steam, not snuffing steam.

LOW AIR FLOW IN A FIN FAN FORCED DRAFT AIR COOLER

Clare! Check the pressure drop of the air flowing across the fin fan air cooler bundle! I think the bundle is fouled. The fins are full of dead moths!

Carl! That won’t serve any purpose! The fan discharge air pressure and discharge flow remains relatively the same, even when the outside of the fins get fouled!

But look how hot, 140°F, the air blowing out of the top of the fins is running! Certainly we’re getting less air flow and more delta P?

The air flow is less through the bundle, as you’ve said! But the flow leaving the fan is constant! The air flow through the bundle just drops off to keep the DP constant!

Well I’m confused. It seems like some air has gone missing, if the flow from the fan is constant, but the flow through the tube bundle drops off?

Your “missing” air flow is recirculated around the periphery of the screen, underneath the fan. A small piece of paper will be blown off the screen around its edge.

That’s true! But I always see this recirculation flow!

Yes! But as the fins foul, the amount of recirculated air increases, and the amount of air flowing through the fins decreases, but the total air flow is approximately constant!

MEASURING AIR FLOW FOR AN AERIAL FIN FAN AERIAL COOLER

Go get an anemometer! That’s used to measure air velocity. Clare – wind speed. We need to calculate the air flow through the aerial cooler!

Actually, Carl, that’s not the way it’s done!

What is the way? I think the fan is not spinning fast enough and air flow is too low! Maybe the belts are slipping?

Well, Carl, I’ve measured the air inlet and outlet temperature (130°F - 90°F = 40°F) to get a temperature rise of 40°F. The specific heat of air is 0.25 BTU/LB/°F. Thus, each one pound of air is picking up 40°F × 0.25 = 10 BTU/LB.

Okay! But how do we calculate the air flow?

I need to calculate the process duty. Let’s say that’s 10,000,000 BTU/HR.

The air flow is then: 10,000,000 BTU/HR ÷ 10 BTU/LB = 1,000,000 LB/HR of Air Flow

Thanks, Clare! That seems simple enough! We can compare the calculated result to the vendor’s performance data sheet!

Yes! But the vendors never take into account air recirculation, which happens even when an air cooler is new and clean. Therefore, the observed air flow through the bundle is always too low by maybe 20%.

MEASURING COOLING TOWER EFFICIENCY APPROACH TO WET BULB TEMPERATURE

Clare! There’s something wrong with the cooling tower!

Mr. Carl! Why do you think the cooling tower performance is bad?

Use your brain, woman! The ambient air temperature is 80°F, and the effluent water from the cooling tower is 100°F. That’s a 20°F delta T. The unit was designed for a 12°F.

True enough, boss! But the cooling water tower was designed assuming very low relative humidity. This is New Orleans, and the humidity is really high!

Relative humidity? I’ve heard about that! How does that affect the cooling water temperature?

First, I’ll measure the WET BULB TEMPERATURE! I’ll tie a piece of cloth around the bulb of a lab thermometer. Soak it with water and then swing it around for a minute. This gives me the wet bulb temperature.

That sounds familiar! Then what?

Next, subtract the wet bulb temperature (78°F) from the cooling water return temperature (100°F), to obtain the approach to the wet bulb (100°F - 78°F) = 22°F. An approach temperature of:

• Less than 10°F – excellent
• Above 20°F – terrible

So, admit it, Clare! I was right! The cooling tower is defective!

Not exactly, sir! It looks like the belt is slipping on the cooling tower’s induced draft fan! That’s pretty easy to fix!

Clare! It could also be that the angle of the cooling tower’s fan blades are not set steep enough. I read that 22 degrees is pretty much the optimum.

ADJUSTING HEATER STACK DAMPER FOR OPTIMUM ENERGY EFFICIENCY

Clare! Make sure that the heater stack damper is half closed, then you may adjust excess air using the air registers, at the base of the heater!

Perhaps that’s not a good idea, Carl! We could develop a positive pressure at P₁, below the bottom row of convective section tubes!

Okay! That could damage the heater’s roof arch supports. Better to open the stack damper 100%!

Perhaps, Carl, that’s not a good idea either! Too low a pressure at P₁, will draw cold tramp air into the convective section, cool the flue gasses, and reduce convective section heat recovery!

Hmm? Then just partly close that stack damper until the pressure at P₁ is about 0.05″ to 0.10″ of water draft. I read in Norm’s books that’s the optimum draft.

What Norm forgot about was wind! If the wind dies off suddenly, the draft may suddenly decrease, and the pressure at P₁ would go positive. Then, if you’re burning high sulfur fuel, the SO₂ will blow out of the fire box and could easily kill someone! In New Orleans, it’s not so windy, and Norm forgot about this danger!

Okay, Clare! Do your best! I’ve got to go to an important meeting now! I agree that the air registers and the stack damper need to be used as a team to optimize excess air and draft simultaneously.

PREVENTING TRAY DUMPING BY USE OF BUBBLE CAP TRAYS

Clare! Bubble cap trays should never be used in a modern design of a distillation tower! Use modern grid trays!

Sorry, Carl! But I tend to disagree. Bubble cap trays do have distinct advantages!

No, Clare! Don’t you realize that bubble cap trays have 10-15% less vapor handling capacity, before jet flooding, than do valve, or grid, or even sieve trays?

That’s very true, Carl! But bubble cap trays are not subject to tray deck dumping, or leaking, or weeping! As long as the height of the riser is above the weir!

Hmm? But why are bubble cap trays not widely used any longer? Even a modern grid tray will lose tray efficiency when the vapor pressure drop gets too low, and the tray deck dumps liquid. That promotes vapor-liquid channeling!

Mainly, boss, because during turnarounds, it’s a lot of trouble to remove each bubble cap, and clean between the riser and the cap!

You’re right, Clare! Bubble cap trays will have a better turndown efficiency than do perforated tray decks. Let’s use some! Especially in distillation services where the vapor flow rate can vary a lot with feed and reflux rates.

DEMISTER FOULING IN VAPOR-LIQUID SEPERATOR VESSELS

Clare! Use of a 4″ thick stainless mesh demister pad in a vapor-liquid separator is always good design practice.

Not always, Carl! Sometimes a demister will promote entrainment!

Don’t be a fool! Demisters cause entrained droplets to coalesce into larger droplets and settle out more readily! Haven’t you even heard of Stoke’s Law?

But Carl, in refineries, demister pads will foul and partly plug. Then delta P across the demister will cause it to be torn from the vessel wall. Vapor channeling and locally high velocity will then cause worse entrainment!

Okay then! We’ll just use demisters in clean services – which I guess precludes their use in most refinery services.

Actually Carl, what works better than demisters is an IMPINGEMENT PLATE! If designed correctly, it converts one radial entry point into dual tangential entries. Norm tells me he uses this design, especially in towers with high inlet velocities. The idea is to split the flow as it enters the tower in two opposite directions around the inner circumference of the vessel.

For real bad entrainment, a “Vortex Tube Cluster,” will do a super good job of separating vapor from the droplets of liquid.

EFFECT OF TEMPERATURE ON LIQUID LEVEL INDICATION

Mr. Carl! You’re running the tower bottom level too high!

Clare! I know what I’m doing! The level is only 60%. Use your eyes, woman! Look at the level glass!

But Carl, the glass is 500°F colder than the tower. The s.g. of hydrocarbon liquids, increases 5%, for each reduction of 100°F:

What’s your point, Clare? We’re talking about levels, not density.

Dear man! Do not yell at me! The liquid in the glass is measuring the pressure difference between points A and B, in terms of the s.g. of the fluid in the glass! As the s.g. of the hotter liquid is 25% less than the 60°F liquid, the tower level will be 25% higher than the level in the glass!

Hmm …? I guess that means that if the level in the glass is like 60%, then the level in tower might be 80%! Is that right, Clare?

At least, boss! If there is FOAM inside the tower, the discrepancy may be much worse! If the bottom level rises above the vapor inlet, the liquid or foam in the bottom of the tower will be pushed up against the bottom tray! Then the entire tower is going to flood!

DRAW-OFF NOZZLE CAPACITY LIMITS

Clare! Forget about your project about adding a new nozzle on the crude tower!

But Carl, the existing 4″ nozzle is too small. The hydraulic head loss through the nozzle is 12″:

And the height of the sump above the center line of the nozzle is only 13″! Without a new 6″ nozzle, the sump will overflow and dump kerosene into diesel!

Pay attention! The code stamp on the vessel says “Stress Relieved – Do Not Weld”! What part of that, Clare, don’t you understand?

Sorry, Carl! That just means that if we do weld a new nozzle on the tower, that the weld has to be Post Weld Heat Treated or stress relieved, to stay within ASME boiler code requirements! It’s no big deal – maybe $10,000?

Sounds like a lot of trouble. But as an American, I always support my local ASME!

ON-STREAM REPAIR OF TUBE LEAKS IN SURFACE CONDENSER

Oh Clare! Please tell operations to shut down the main air blower on the FCU! We’ve developed a giant tube leak in the surface condenser! There go our profits for the 3^rd quarter!

How do we know that, Carl? That’s terrible news!

Dear girl! Don’t you ever look at the lab results? The conductivity of the steam condensate is through the roof!

You’re right, Carl! But we can maybe fix the leak on-stream!

On-stream? Like with magic? Wake up to the real world!

No, sir! We can inject shredded fiber glass into the cooling water supply, provided the leaks are not too gigantic and have developed gradually. Norm used to use asbestos fibers in the 1960s! It’s a way of buying time.

Sounds simple! But is there a downside? Don’t forget what that French woman said, “You can’t have your cake, and eat it too!” Ha-ha! Personally, I don’t like cake!

Well, yes! The problem could be the fibers could plug up the pump’s suction screen. Our vacuum condensate pump has external seal flush, so the seal is not a problem! Also, we would need to put the condensate to the sewer, to avoid returning hardness deposits to boiler feed water! Also, Marie Antoinette was Austrian, not French!

Okay! Let’s give her a try! If it works, we won’t have to shut down the CAT!

WHEN ARE VORTEX BREAKERS REQUIRED?

It’s good engineering practice to use a VORTEX BREAKER in all draw-off nozzles! The vortex breaker should be shaped like a cross! Can you remember to follow this accepted standard, Clare, in all your designs?

I don’t totally agree with you on this, Carl!

What now, woman? Why do you always want to debate everything?

Well, for one thing, at velocities of less than 2 ft/sec, vortexing in nozzles is not a problem! At draw-off nozzle velocities above 4 ft/sec, vortex is sure to be a problem!

And how do you happen to know that, Clare?

I performed some experiments in my bathtub! I like to bathe!

Still, it seems like we should use vortex breakers all the time, as they’re cheap and don’t hurt anything. Better safe than sorry!

Actually, vortex breakers can cause a great deal of trouble. Stray pieces of tower internals and turnaround trash may be caught up in the vortex breaker vanes and restrict flow!

Hmm! I hadn’t thought about that. But then, I take showers! Ha! Ha! It’s quicker than a bath! But what if nozzle exit velocities are above 4 ft/sec? Then, you have to use my idea of a vortex breaker!

You’re quite correct, Carl! But then Norm suggests a 1″ × 1″ mesh, 316SS screen be welded across the top of the sump, to protect the vortex breaker from plugging!

Well, Clare, okay! I don’t have time to worry about your little details!

NAPHTHA INJECTION TO CENTRIFUGAL COMPRESSOR

Look here, Clare! Have you utterly lost your brains? You can’t inject liquid into a compressor! You’ll tear up the rotor and break the valves!

Correct, boss! For a recip, one should never permit liquids to enter the compression cylinders! However, for a centrifugal compressor, a finely divided mist or spray can be helpful!

And why is that, dear woman?

Well, the gas enters the centrifugal compressor at its dew point temperature and pressure. And then, as it’s compressed, the heat of compression superheats the gas. The gas dries out, and any solid fouling deposits, like salts, precipitate on the rotor’s wheels! It’s a slow, but steady process!

Hmmm? I was just wondering, if you’re so smart today, why are the amps on the motor driver of the coker wet gas compressor going down, as the rotor fouls? You’d think the amps would go up?

Fouling causes a large reduction in compressor capacity. The efficiency drops too, but the capacity declines faster.

Well, Miss Brilliance! What are we going to do about the rotor fouling on the coker, hydrocracker, and reformer centrifugal compressors?

Carl, as I tried to explain a moment ago, spraying a finely divided mist through a spray nozzle will prevent the gas from drying out and keep the rotor clean and will …!

… and probably wreck the machine!

No, Carl! Not if the spray injection rate is a finely divided naphtha mist calculated to absorb the heat of compression by evaporation! Also, the liquid spray has to be tripped off when the compressor shuts down!

Well! Go ahead with your calculations! I’ve got an important meeting to attend!

Hmmm? I suppose it would be really, really bad to spray a mist into the suction of a reciprocating compression? Likely wind up wreaking the discharge valve plates!

INTERNAL OVERFLOW FROM TOTAL TRAP-OUT CHIMNEY TRAY

Oh, Clare! That FCU Debutanizer is flooding! We’re going to have to cut cat feed! I guess the trays are fouled again. Let’s plan a shutdown to chemically clean that rotten tower again!

Sorry, Carl! That’s not our problem! We cleaned that tower six months ago – no help! It’s not dirty trays!

Look, Clare! Don’t argue! Just for once, can’t you follow my instructions? UQP reviewed the tray capacity! They’re only 68% of jet flood and 45% of downcomer back-up! Doesn’t that prove that the trays are fouled? Think, woman! Look at the Aspen simulation!

Actually, Mr. Bossman, it’s a UQP design error that caused the problem! The reboiler shell side has become fouled, and its shell side DP has increased a few psi. That pushes up the liquid level in the chimney tray feeding the reboiler! As there’s no adequate provision for internal overflow, the high liquid level in the chimney tray causes the bottom downcomer to back-up and flood!

So now, Clare, you think you’re also smarter than UQP! Well, you’re not! UQP included a 6″ slot for internal overflow in the chimney! I’ve seen that slot myself!

Yes, Carl! But did you happen to also see that the bottom of the slot is 28″ above the seal pan overflow lip? This will lead to excessive downcomer back-up and flooding! We need an 8″ overflow pipe, with its top edge in-line or below the seal pan!

That’s an awful way, Clare, to talk about UQP after that party they made for us last Christmas! How many crab cakes did you stuff yourself with?

VACUUM EJECTOR – LOOSE STEAM NOZZLE

Clare! Exactly why is it, that you have partly closed the motive steam nozzle to the first stage ejector? Is this one of your misguided attempts to save steam? We need to be maximizing vacuum, not saving steam!

Yes! I’m quite aware we should be maximizing vacuum! That’s why I have optimized the jet’s motive steam pressure!

But, woman! The manufacturer’s design motive steam pressure is 150 psig! Look at the name plate! Can’t you see it says “150 psig”? I suppose you’re smarter than the ejector manufacturer now? Well?! I’m waiting for an explanation!

All I can say, Carl, is that I reduced the motive steam pressure to 118 psig, and the vacuum improved from 42 to 35 mm Hg. The usual reason for this is that the 316 SS nozzle has become loose where it screws into the carbon steel body of the jet! And the motive steam is partly blowing into the jet mixing chamber, but not through the nozzle!

Hmm …? I wonder why the manufacturer didn’t make the back part of the ejector out of 316 stainless also? You know, Clare, to avoid galvanic corrosion in the presence of wet steam? I guess they have a really good reason?

Or maybe, the nozzle is partly eroded? Or maybe, the downstream condenser is fouled and can’t condense the motive steam completely? It’s kinda hard to know, Carl!

Okay, I get your point! Let’s throttle-back a bit more on the motive steam and …! Oh no! The jet’s surging! The vacuum’s breaking! Look, the pressure’s jumping up! Help, Clare! Help! Do something, quick!

Carl! You’ve forced the jet out of its “CRITICAL MODE,” of operation! The jet has lost its sonic boost! You’ve got to be more careful when you try to optimize the motive steam pressure! Take your time! Reduce the steam pressure slowly and cautiously! Don’t rush!

All right, Clare! A little less criticism would be appreciated! I’ve got it!

EFFECT OF TRAMP AIR LEAKS ON HEATER EFFICIENCY

Next time you’re outside, Clare, pinch-back on the heater air! We’ve got too much O₂! Either on the stack damper or air register! It doesn’t make any difference! But don’t forget!

It surely does make a difference! If I close back on the air register, draft will go up at Point “A”! If I close back on the stack damper, then draft will go down. Too much draft is bad. And too little draft can result in a positive box pressure, if the wind slows later on!

Okay, okay! I sure don’t want a positive box pressure! That’s dangerous! So just close those burner secondary air registers halfway, Clare! Understand? Okay?

Not okay! If we develop too much draft, then cold tramp air will be sucked into the leaky convective section! Then, Delta O₂ will go up! That’s the difference between the O₂ at Point “B” minus Point “A”. Then, the percent of the heater fuel wasted, due to the cold tramp air, quenching the convective section hot flue gas, will be:

Where Delta T is the stack temperature, minus ambient air temperature, in °F! (Use 300 if working in °C)

So, if I want to cut back on combustion air, if I close the air register, I’ll suck in too much tramp air leaks! But if I close the stack damper, then the radiant firebox pressure might go positive! And, blow-out hot flue gasses with deadly SO₂ out the sight ports! My, oh my! Life sure is complicated!

EFFECT OF A SINGLE FOULED TRAY

I just inspected the crude unit debutanizer, Carl! Big problem! We’ll have to get Mr. Thompson in maintenance to hydro blast tray #20. It’s fouled, and those tray caps are sticking to the tray deck!

No! I’ve approved the debutanizer for closure yesterday! Yeah! I saw that feed tray – tray #20 you say? – was pretty fouled with gunk! But the trays above and below that feed tray were clean! It’s just that single tray #20 was missed, and didn’t get hydro blasted clean. I’m not gonna get those maintenance guys mad at me for just one tray! You gotta learn to be diplomatic, Clare! Don’t wanna get Tommy mad at us!

Carl! If the valve caps are sticking on one tray deck, then that tray will develop a high pressure drop! It would be better to break off the valve caps, then start up with those caps sticking to the floor of the tray!

Listen to me, woman! I don’t care ’bout a little extra delta P on the debut! It ain’t no big problem! How ’bout makin’ us a new pot of coffee, and stop worrying ’bout nothin’? Stop complaining!

Make your own damn coffee! It’s not the pressure drop! It’s that a high vapor delta P on one tray will cause that tray to flood due to downcomer back-up. And then, all the trays above will flood with time! And all the trays below will dry out! And flooded trays and dried out trays don’t fractionate!

Clare – you been readin’ that Lieberman book again – A Working Guide to Process Equipment? Is that where you’re gettin’ these dumb ideas from? Okay. I don’t like to argue with women! Call Tommy on your radio. Channel 6. And get maintenance back down here to clean that tray #20. Will that make you happy?

STEAM TURBINE – SURFACE CONDENSER OUTLET TEMPERATURE

We’ve just received instructions from Tech Service to reduce the surface condenser outlet temperature! Their idea, Clare, is that a lower surface condenser outlet temperature, will reduce the turbine exhaust steam pressure! This allows more horsepower to be generated from each pound of steam. We might be able to save 5% of the steam to the K-805 steam turbine, if we can knock 5°F-10°F off the surface condenser outlet!

Yes! Quite so! I believe Norm calls this an improved “Isentropic Steam Expansion.” The lower the condensing steam temperature, the lower the surface condenser pressure. I’ll try to get the surface condenser temperature cooler by turning on another cooling water pump and also back-flushing the condenser’s tubes!

Why is it you always want to do things the hard way? All you got to do, Clare, is just raise the level in the Hot Well! Then you’ll see the boot temperature will go down! Look at the dial thermometer at T-1! See, that’s the reason we got experienced men like me out here! Understand now, Clare? I got me 22 years of experience out here!

Yeah! You’ve got one year of experience 22 times! Raising the Hot Well level too much will back-up condensate over the tubes! The condensate is gonna get sub-cooled, and lots colder! But the number of tubes exposed to the condensing steam will be reduced. Then the vapor outlet temperature at T-2 will increase!

Who cares? We monitor the temperature in the Hot Well, Not the vapor outlet! The vapor temperature is not important! Don’t you read the operating manual? It says to keep the Hot Well cool! Can’t you read, Clare?

The turbine exhaust pressure is controlled by the vapor pressure of water at the condenser VAPOR outlet! That means T₂ – not T₁! If you back-up the hot well, the pressure in the surface condenser will increase and the turbine is going to slow down with the same amount of driver steam flow! Try it and you’ll see!

Later that Day

Oh, Clare – would you mind starting P-303-C next time you go out! Maybe a bit more cooling water flow to the turbine might help us a bit! Can I get you a coffee?

WATER ACCUMULATION IN TURBINE CASE

Uh-oh! Clare! That turbine is starting to vibrate again! It’s water accumulating in the turbine case, hitting the wheels. If only the exhaust steam outlet was on the bottom of the turbine case! I suppose the drain on the bottom of the turbine is still plugged!

Carl, that drain won’t work anyway! That end of the turbine is under vacuum. We’ll just have to dry out the exhaust steam! I’ll start slowing down the turbine by closing off on the governor valve! That’ll heat up the exhaust steam by about 10°C!!

No, woman! When you close off on the steam inlet valve, meaning the governor speed controller, the supply steam gets colder! Even you ought to know that much! Understand, Clare?

Do not use that tone of voice when you speak to me! I’m not your wife! Anyway, you’re right, reducing the steam pressure will cause the steam leaving the governor valve to cool off. But throttling on the governor will heat up the turbine exhaust steam and make it drier!

Hmm … Yeah! I kind of noticed that the turbine exhaust steam temperature to our surface condenser does heat up when I slow that turbine down 50 or 60 rpm’s. Seemed kinda strange! I don’t know if …

If you look at the Mollier Diagram Norm gave us in his seminar last May, you’ll understand! Throttling on the motive steam supply to the turbine is reducing the amount of horsepower, or work, that the steam supplies to the turbine wheels. That leaves more heat in the steam exhaust!

Yeah! It makes the turbine work less efficient. Which then heats up the turbine exhaust temperature! Seems backwards though to cool off the motive steam to heat up the exhaust steam! But I guess anything that works less efficiently starts to heat up! I noticed, Clare, that you heat up too, once in a while! Ha-ha!

Save those sexist comments for that poor woman who had the misfortune to marry you! It’s just not a good design to have the turbine exhaust nozzle below the vacuum surface condenser, in a condensing steam turbine.

Yeah! Those dumb engineers should have elevated that there turbine high enough, so that we could drain right into that surface condenser. Uh … you got my coffee ready yet?