reboiler in the column 3

[skvishwa]

hi all,
This question referes to a thread i saw some time back about the reboiler "slip flow",because of the low velocity and all.....
I am facing a similar problem in a column reboiler.The system includes a vertical thermosyphen reboiler in a column .The column is mainly a stripper that is being for stripping the Hydrogen Sulphide and light hydrocarbon from the Wild naphtha generatd from the Diesel Hydrotreater.
The problem is very peculiar.The comlumn reboiler does not function when the feed flow is below a particular value and starts to function after reaching a particular column feed flow,but after running normal for some time again it slowly starts to behave abnormally.
The column top temperature, reboiler inlet temp,bottom temp,all come down.But the reboiler outlet tempearture remains same.this indicates that the thermosyphen is not getting stabilsed and the flow thru the reboiler is decreasing .After the stable operation for some time the reboiler thermosyphen gets broken.
Any solution?
Santosh

 

[epoisses]

Hi Santosh,
Assuming you're operating this column within its design window, my first guess would be that something's damaged around the reboiler. You may want to carry out a gammascan to see if that gives you a clue of what's going on inside.

 

[hacksaw]

are you running a flooded reboiler?

 

[25362]

The reboiler is but a small part of the system. Stable operating conditions and types of control are inherently linked with a successful reboiler operation.

Thus, adding to hacksaw question: Is it a "once-through" unit, fed from the lowest tray downcomer or a circulation reboiler taking liquid from the tower's bottom ? What steam is being used, how much superheat does it have ? Is the amount of naphtha + vapors of the top of the tower about constant ? Is the top condenser working at a constant temperature and pressure ? Can trays and reboiler foul ? How are the steam supply and condensate removal controlled ? How is reflux controlled ? Is the stripper operated at design conditions, otherwise has it ever been revamped to handle larger V/L loads ?

Kindly comment.

Note: Elements of Petroleum Processing by D.S.J. Jones (Wyley & Sons) ISBN 0-471-95254-0, brings an example of a heat and material balance on a hydrotreated diesel stabilizer (stripper) that may be of help.

 

[sshep]

The "liquid slip" type reboiler failure is associated with a build-up of heavy residue in a recirculating system. Other process-side heat transfer problems include film boiling and foaming. Before considering any of these type phenomenon, field check for proper liquid level since this drives the circulation (the circulation can be estimated from geometry and heat input, but unfortunately can rarely be measured directly). It sounds like you are already satisfied that the problems are not related to the steam side. A bit more description of the configuration could be useful.

It is not immediately obvious to me that general conclusions can be made based on the reboiler return temperature since a very small amount of lights can cause a significant temperature temp decrease without changing the shape of the vaporization curve in the region of 10% to 40% vaporization.

 

[skvishwa]

Thank you guys.Thanks for your prompt replies.
The information about the reboiler/column is as follows.
1.This is a flooded reboiler with the hydrocarbon in the tube side and steam(14.5 kg/cm2g,215 deg c temp.) in the shell side.
2.Column has not been revamped.
3.There is no reflux to this column.The feed enters the column at the top tray(feed is typically 90-180 cut range naphtha and 2000 ppm H2S).
4.Trays are valve trays.
5.The chances of the chockage of the column tray are negligible as it is a very clean service.
6.There is separation baffle in the bottom of the column with the top portion of it having a cut so as to facilitate the overflowing of the liquid to the other part which is the rundown pump suction side.The liquid flows from the lowest tray to one side of this separation baffle,from here the liquid goes to the reboiler-tube side.The reboiler return goes to the other side of the baffle.Here vapour goes up and liquid gets collected at the bottom.From here the pump takes suction and sends the liquid to rundown after cooling.
The steam inlet system to the reboiler is lilke this.The flowmeter is given in the inlet to the reboiler and the control valve has been provided in the condensate outlet side.Thus the control is there by covering/uncovering the area available for the condensation of the steam.We have operated the control valve with the minimum opening to the maximum opening without any success.First indication of the start of the malfunction of the reboiler is seen in the reduction in the steam flow to the reboiler.The steam was checked for the condensate, it was found ok.Also whenever the thermosyphen broke the reboiler condensate outlet which is liquid condensate during normal operton of the reboiler turned to steam as in the inlet.So,we ruled out the problem with the steam part.
Then the damage with the separation baffle was suspected but since the system worked nicely for some time (approx.12-18 hrs) without any problem at higher feed flow, it was ruled out.Also the system worked fine even at the lower levels(level indication is on the pump suction side)also so it was not expected to be a problem.
There is on more observation ,whenever the level of the liquid at the bottom is increased and it becomes full(as seen physically in the level guage)the reboiler start to funtion.We understand that it is because of the liquid back up in the reboiler thru the vapour return nozzle.As the level is decreased it turns back to its abnormal ways.
The normal feed to the column is very small (in the range of 2-3 m3/hr).
Best regards,
santosh

 

[sshep]

I had a thermosyphon reboiler problem with symptoms similar to yours (thread124-66970) We lost heat transfer for no apparent reason, with the consequence that no steam will condense (symptom of process side problem).

I can give you some ideas to try, but please clarify items of your configuration that seem strange for a stripping column of such low feed rate.

1.) There are two compartments in the bottom and the reboiler is once through into the rundown side (yes/no)?
2.) The reboiler compartment can overflow onto the rundown side and was probably designed with the intent of a continuous overflow (yes/no)?
3.) Are the compartments are side by side seperated by a baffle, or are they at different elevations? If side by side is there an equalization line?
4.) Do you have any level instrument or gage on the reboiler feed compartment?

This is where I am going:
Since this is a stripping only column, your reboiler circulation rate cannot excede the feed rate which you state to be very low (2-3m3/hr). Steam flow sets both the tower vapor rate (desired response) and the reboiler circulation rate (unfortunate consequence). I think a recycled thermosyphon action might be more stable because in that case you don't care what the circulation rate is.

Your system may be drawing the reboiler compartment down to meet the feed rate constraint, resulting in unstable operation. Often since the designer expects the compartment to be full, he figures any level indication to be unneeded(question #4). This theory could explain why high level gets such good results. Depending on answer to #3, perhaps you can make the reboiler work better by equalizing the two compartments and running at a high (but controllable) level.

just a thought- sshep

 

[25362]

You haven't told us whether the dissolved hydrogen and other lights' content, i.e., feed composition, besides the 2000 ppm H2S, pressure and temperature are constant.

Anyhow, this is just a thought: if there are no mechanical problems, such as holes in the bottom baffle, by having no reflux, and just the feed flashing as it enters the stripper, one wouldn't expect much fractionation.

It appears that as time progresses somehow the wild naphtha heavier fractions with NB +160[sup]o[/sup]C would tend to concentrate in the reboiler-side compartment raising the bubble point until the 200[sup]o[/sup]C (14.5 kg/cm[sup]2[/sup] ga.) steam is unable to maintain the thermosiphon effect.

The feed then overflows through the reboiler into the other compartment. That may be the reason why, after the thermosyphon effect is broken, upon rising levels, lighter material re-enters the reboiler, which starts operating again.

If the malfunction reason is as depicted above, one probable step would be to adjust the tower pressure (not given) to bring down the bubble point of the naphtha entering the reboiler.

Please comment on the above to enable more ideas to pop up.

 

[katmar]

Hi Santosh,

You have made a good job of describing your problem, but there is one missing item that I always find helps me to get my thoughts on the right track. That is the definition of whether this is a "Day 1" problem (i.e. it was always there) or if it is a problem that only started occurring after some time of operating normally. This allows you to determine whether it was a design/construction fault (if it is a "Day 1" problem) or whether something has broken, gone wrong or changed in some way.

Your simple, practical check that there is steam in the condensate line tells us that there is no problem with the steam system, and that there is no heat transfer taking place.

Now we combine the practical work with a little bit of theory. We all know
Heat transferred = Heat transfer coeff x Area x Temp diff

The area cannot change, and you have told us that the temperature difference has not changed so we can conclude that the heat transfer coefficient has decreased dramatically.

What would cause the HTC to drop? Normally it is high because of the boiling. The decrease means there is no boiling. Either the liquid in the base of the column is below its boiling point, or the tubes are full of vapor.

It is very unlikely (given all the info you have supplied) that the liquid is below its boiling point, so I would conclude that the tubes are filled with vapor. At start-up the base of your column will be filled with liquid, there will be liquid in the reboiler tubes, and the reboiler will work properly.

You say that the "return from the reboiler goes to the other side of the baffle". I suspect that this is the problem. With the small feed rate you have, the stream from the bottom tray to the side of the compartment that feeds the reboiler will be less than the liquid thrown from the reboiler return to the other side of the baffle, thus gradually depleting the liquid level on the side feeding the reboiler. After some time this compartment will run dry, as you have experienced. At higher feed rates this will not happen.

I would ask why there are two separate compartments in the boot of the column at all. If you are going to have two compartments, then the reboiler return should go to the same side where the reboiler feed is taken from. Only the excess should overflow to the other side, where the level controller will remove it.

This is where we need to know if it is a "Day 1" problem. Should there have been a deflector plate installed to keep the reboiler return on the same side of the separator plate? If it is not a "Day 1" problem - was there ever a deflector plate that has now broken away?

I prefer to have no separate compartments in the bottom of the column, and set the level controller to ensure that the reboiler always has the correct head on it. Sshep has made very much the same analysis as me, and because he has seen this problem in real life before, I would have confidence in his opinion.

regards
Katmar

 

[skvishwa]

Thanks for your great response.
I would add/correct some information to my last post.
1.I rechecked the column drawing and found that the Reboiler is not the once through type, there is a deflector plate below the reboiler return nozzle which sends the total Reboiler outlet material –liquid back to the Reboiler inlet side. And from there the liquid overflows to the pump suction side. That way it is ensured that the Reboiler side level is available.
2.The compartments are the part of the same column bottom and there is no provision for the equalization line.
3.There is no level measuring instrument on the Reboiler inlet side.
4. The feed composition to the column is practically constant. It is dependant upon the hydrotreater reactor severity in the upstream, which is not changed much during normal operation.
5.We have tried to operate the column at the lower pressures, and saw the increased Reboiler circulation during the course of reduction and things started to awry after some time. We operated at the pressures 1 kg/cm2 below the normal value (5.8 kg/cm2g).But the Reboiler did operate for some time at this reduced pressure. The duration of such operation was approx 18-20 hrs.
6.This is a Day-1 problem. It did not crop up some time of operation.
One more thing i want to add is once we had the opportunity to operate the column at feed flow 3-3.5 m3/hr as against the 1-1.8 in the normal operation.The column operated satisfactorily for days.
Best regards,
Santosh

 

[DickRussell]

Having a baffle in the bottom of a column can be useful from a process point of view, although not necessarily in this case. When the bottom tray liquid contains a significant amount of light material that ought to pass through the reboiler at least once, a correctly arranged baffle can do that. In that case, the bottom tray seal pan empties into the side that feeds the reboiler, and the reboiler return liquid dumps into the other side, from which the bottom product is drawn. The reboiler return side overflows the excess to the reboiler feed side of the baffle. Flow of net bottoms is controlled to maintain level in the reboiler feed side.

Of course, having the plumbing around the baffle arranged as Katmar suggests, with the reboiler returning to the feed side and only the overflow going to product, can be used to fixed the head of liquid supplied to the reboiler, regardless of level in the other side, from which the bottoms is drawn. If your column is set up this way, then I suppose one possibility for explaining the reboiler problem is that there is a severe leak from the reboiler feed side to the bottom product side, resulting in insufficient head of liquid to the reboiler. If that is a possibility, then it might be possible to verify it without shutting down by increasing the level in the bottom product side.

However, the Aug 04 post seems to say that some of the bottom tray liquid goes to the reboiler, while the excess is supposed to overflow the baffle to the bottom product side. The Aug 05 questions by Sshep are asking for the exact physical arrangement and intent of the design, and I haven't seen the answers posted yet. I think he has pointed to a likely problem - the incompatibility of original design and low feed flow that won't support operation of that design.

Without any baffle, the bottom tray liquid just mixes with reboiler return liquid and thus, by dilution, some of it passes directly to the bottom product. This is fine in a normal case, where the stripping trays are doing the required separation and the reboiler is just there to provide energy. In some cases, however, there is a significant temperature difference between bottom tray liquid and reboiler outlet, indicating that the reboiler is doing a significant piece of lights removal. In that case, having the baffle to force bottom tray liquid through the reboiler at least once allows the reboiler outlet to be cooler than it would otherwise in order to get the mixed liquid leaving the bottom as product down to the proper level of lights. This can be important if the column pressure is such that the reboiler is designed for a close approach to steam temperature. Of course, the other way to address the situation, notably when the bottom product is very heavy and a small fraction of the feed, is to use a pumped-through, suppressed vaporization reboiler, with high circulation rate.

 

[25362]

Just a thought: The lowest valve-type tray may leak either from a mechanical failure or because it operates below the liquid flowrate window.

In either case, it may be pouring part of the feed straight into the next-to-the-reboiler compartment at low flow rates. When feed is increased the tray works better allowing its DC to dump enough liquid into the reboiler compartment, facilitating its normal operation.

Kindly comment.

 

[katmar]

Based on the extra info, I would bet that either the deflector plate on the reboiler outlet is missing/damaged, or (as DickRussell suggests) there is a leak from the one compartment to the other.

If you cannot stop and inspect the column now, DickRussell's proposal of lifting the set-point on the level controller is about the best you can do while still on-line. If possible, raise it slowly to approximately the level of the reboiler's top tubesheet, checking for problems as you increase it.

Good luck
Katmar

 

[JoMM]

Hi Skvishwa,

I once worked on a unit with an almost similar arrangement having no level indicator on the reboiler side and it was sometimes a nightmare because of operator's lack of understanding mass & energy balance.

Successfull operation at higher feed rates is fascinating and it does suggest a dry-out problem on the reboiler side. This type of installation, no level measurement on the reboiler side, assumes that you will always have the correct mass balance at the column bottoms but that can be easily changed by inappropriate energy input into the column espcially at low feed rates. Basically part of the question is: Is the feed to heat input ratio right (energy balance)? What happens to the level and the flow rate on the pump side? Check the operating data to see if you have a drop in level/flowrate on the pump side sometime before you realise a reboiler failure (Time between the first time the level valve cuts back and reboiler failure will be higher for bigger reboiler volume). ??? Hope to here from you.

 

[skvishwa]

Hi JoMM,
The feed to heat input can be varied by varying the steam flow to reboiler.I have done that too,but that did not give any result.I have already mentioned it in the my earliar posts.
Whenever the thermosyphon breaks, the level in the pump suction side increases ,since it is in cascade with the flow to rundown ,the flow also increase.This also manifests that the total vaporisation has come down.

santosh

 

[katmar]

Hi Santosh,

I'm sorry to hear that you still have this problem. Please can you clarify the following situation:

On Aug 4 you wrote:
"Whenever the level of the liquid at the bottom is increased and it becomes full (as seen physically in the level gauge) the reboiler start to funtion. We understand that it is because the liquid backs up in the reboiler thru the vapour return nozzle. As the level is decreased it turns back to its abnormal ways."

And on Aug 21 you wrote:
"Whenever the thermosyphon breaks, the level in the pump suction side increases, since it is in cascade with the flow to rundown, the flow also increase. This also manifests that the total vaporisation has come down."

Does this mean that the problem is self correcting? When the thermosyphon action fails and the boot level starts to increase (as described on Aug 21) does the level get to a stage where the reboiler starts to operate correctly (as described on Aug 4) without any intervention from the operating staff?

If the problem is not self correcting, please describe the action taken by the operators to get it back to normal operation after the thermosyphon breaks.

regards
Katmar

 

[hacksaw]

some depending on the composition of your feed, some reboilers work best flooded.

 

[JoMM]

Hi Santosh;

What you are not saying is what your operating data says happens BEFORE the thermosyphon breaks and that is what I am trying to get to. What I am suggesting to you is that you should see the following if you have a dryout problem: The level on the pump side should drop for the same rundown flowrate. Now, because you are having flow cascaded to level, an even better indication is the rundown flowrate which I expect to drop.

I am also interested in clarification of the comments made on Aug 4th (See Katmar's reply Aug 22) and most important the question about the restart-up of this reboiler.

Have you checked the vessel drawings to ensure that the level indicator, pump suction and reboiler suction on the vessel are orientated as shown on the as built drawings. [Just because the drawings indicate that the level indicator is on the pump side does not mean that it is like that in the field. These things does sometimes get missed during check-outs. So an actual field check must be done during troubleshooting].

Regards

Jo

 

[22082002]

Request vishwa to refer the distillation design and Operation by henri kister. In the end of these books he has given several operation problems solved and design book gives guidelines.Other book -reliable operation for petrchemical and refinery(check the title). he has this problem explained.

Just to the pressure drop analysis, vis-a-vis level of reboiler, and liquid level in column. It's crucial that the reboiler is placed at appropriate height w.r.t.to column liquid leve(min and max), and variuos pressure drops across inlets and outlets and vapor return line. This may be wrong for the case. This may be the cause.

The partition plate ensures that the reboiler is always flooded, i feel it is not contributing to the problem. Above para is what I suspect.

In the wake of inadequate information(para 2), it is quite beating around the bush.

Request vishwa to update on this.
All the best...

 

[22082002]

By saying beating around the bush, I no way wish to disrespect etseemed gentlemen gave valid judgements. Please excuse me if I offended.

It's just that the fundamental information/analasys may be missing on poster's side.