Critical Heat Flux - Kettle reboilers 1

[NiravK]

hello to reader,

We are carrying out thermal design of De-ethaniser and De-butaniser reboilers (kettle)for one of our gas processing units using Hot oil circulation(diesel) on tube side as heating medium.

I have questions about the same.

What should be the maximum heat flux which i should limit?
What should be the maximum film boiling (shell side) co-efficient?

HTRI calculates the critical heat flux as well as boiling co-efficient. However, client is asking to keep both the above values on lower side (may be based on practical exp.)
FLUX < 12000 Btu/hr.ft2
CO-EFF < 300 BTU/hr.ft2.F
which calls for very high heat transfer area than actually required.

while meeting the defined criteria by client and ensuring there is no 'film' boiling in any part of reboiler, i'm not able to keep overdesign below 30% and 40% in two reboilers.

what should be the approach for economical design?
Is there any other guideline or experience available?

thanks in advance.

 

[steveen]

Client is specifying overdesign, just advise him of incremental cost. Bigger bundle = bigger shell = bigger foundation.... and lower velocities on the tube side.

Ask Koch for a budgetdary quote on a twisted tube exchanger. They allow even higher heat fluxes due to more vapor escape paths. If you have never seen the technology, take this opportunity to look at it.

Is the heating medium off a sour stripper? Or is it possible to get sour wet material during startup or shutdown. If so, save your client some grief and suggest a stainless bundle.... At low velocities resulting from the overdesign, scale buildup and plugging of tubes will occur, eventually the bundle will be undersized... and the designer will be at fault. Go figure.

 

[Montemayor]

nirav2909:

Basing a critical kettle reboiler design on heat flux is common in my experience. The heat flux is critical for proper reboiling effect and steady heat transfer. I've always resorted to proven heat fluxes when estimating or designing a submerged reboiler - especially one of the Kettle type. There is a tendency for a Kettle type reboiler to "gas-bind" and cause haphazard operation with diminished heat transfer. A proper heat flux is one thing that can cure this phenomena.

There is no "standard" or "typical" heat flux. The proper design figure depends, among other things, on:
[ul]
[li]The heating medium;[/li]
[li]The temperature of the heating medium;[/li]
[li]The process fluid and it's physical properties;[/li]
[/ul]
Empirical experience and background is essential to select the proper value for the best heat flux. HTRI can't give you this. I normally use a heat flux of 10,000 btu/hr-ft2 for amine reboilers (steam heated) and I have gone as low as 8,000 btu/hr-ft2 in crude oil heaters (direct-fired). I don't know what you call "film boiling"; perhaps this is what I indentify as "gas-binding" - the state where a vapor film is generated and maintained over the process heating surface, effectively insulating and "binding" it from heat transfer.

You don't state your contractual relationship with your client. If you have a lump-sum contract with no definition on the design basis, then you are free to design what you feel is right and can guarantee. However, if you are under a cost-plus arrangement or have a specific definition of design basis for the reboiler, then you must do what you've been told to do: design on the client's basis. "Economic" design is all relative to what your job scope is.

One thing is certain: once designed and built, if the kettle reboiler starts to demonstrate pronounced and haphazard heating effects, the design will be scrutinized and critiqued. If you are using HTRI you are using a well-proven process design tool. However, even HTRI is open to criticism when compared to actual operating experience. HTRI has no field operating experience or fabrication background that I know of - nor does it claim to do so. It is a pure process design based on sound engineering algorithms and some field recommendations.

If you have De-ethanizer or De-butanizer experience in your background or in your organization, I would strongly use that source as your basis. If not, then tie the design basis back to your client. Your client must have a good reason for insisting on a specific heat flux. If he can justify it, go with it at his risk. Don't forget that your client may know something about the fouling potential and may need more heat transfer area (or time element) to compensate for process conditions. Fouling potential is another empirical factor that you can't calculate on HTRI.

One thing I can say about your application is that using hot oil (& basing your controls on it) is a good design basis for your type of reboiler. A low and "gentle" boiling temperature is always more favorable and easier to control than, for example, a direct-fired source. The best attribute about a Kettle reboiler is that if you can achieve a steady, even, and non-gas binding boiling effect, the unit will operate forever (if there is no fouling). Thermosyphons are much more irritable, sensitive, and prone to process upsets.

Art Montemayor
Spring, TX

 

[srfish]

The limiting values of heat flux and heat transfer coefficient,that your client is requesting,come from a book named "Process Heat Transfer" by Donald Q. Kern. It was published in 1950 and has been a standard for many years.

The maximum heat flux depends on such variables as tube spacing, bundle size, triangular or square tube pitch and how close the operating pressure is to critical pressure. Generally the limiting heat transfer values of Kern are too low. But it is possible that a combination of the aforementioned variables would result in a maximum heat flux less than 12,000.

 

[NiravK]

Thank you all for valuable input.

The only concern as pointed out is about the lower velocities on tube side because of large flow area.

However, with diesel as heating medium on tube side, what are the chances of scale formation with lower velocities which is in the range of 1-2 ft/sec in my case.

I am also not able to utilise the pressure drop available on tube side (Calc/allow : 1-2 psi/15 psi).

thanks.

 

[25362]

Some kettle reboilers are provided with multiple liquid inlet nozzles to overcome any internal shell restrictions as by fouling or simply to reduce friction drop.

I've seen (low) finned tubes used successfully in kettle reboilers on fouling services such as on vaporizing furfural from extract-mixes in lube extraction units.

 

[Montemayor]

nirav:

I don’t believe you have any basis for concern in what you state. You say you have a problem with low tube-side velocities; if so, this is only because you are keeping your tube-side velocities low – when you don’t have to. Also, your statement about the pressure drop is not clear. What do you mean about not being able to use the available pressure drop? If you have a generous hot diesel pressure, then use it; I would allow for a pressure drop of at least 10-15 psi across the tube bundle - depending on the number of bundle tube passes.

Are you giving us all the basic data? For example, allow me to give you the following reasons why this application is a relatively simple one:

1) You have a Kettle type reboiler. I assume it is a CKU. This immediately identifies your tube bundle as a U-tube type in this hydrocarbon application. With the U-Tube bundle you can easily have multiple tube passes. This gives you excellent turbulence and film coefficients through the bundle with the hot liquid diesel. This is the reboiler heat input control I mentioned in my first post. You are free to control the hot diesel flowrate (reboiler load) through the bundle.
2) Since you have the ease of customizing your tube-side velocities with the hot diesel, you should be making sure that you have a reliable driving force on that same diesel in order to maximize the operating flexibility by varying the flowrate – should it be desired. Since this involves a liquid heating fluid, this makes it very easy (as compared with a gaseous fluid) – as I also mentioned previously. Why are you using 1-2 ft/sec when you can employ a higher velocity through the tubes using multiple tube passes in the U-tube bundle? Is this a client constraint? I don’t know the quality of the diesel heating fluid you plan to use, but the diesel I would employ wouldn’t produce nor leave any scaling inside the tubes.
3) Using hot diesel is a great advantage. I would not expect nor design for any measurable fouling in the tube side – unless you want to. I have to assume that you are using reasonably heated diesel – and not a 400 – 500 oF stream, for example! I may be wrong, but I don’t believe you require a diesel temperature above 300 oF. If I’m wrong, please state so. The lack of basic data makes it necessary to make assumptions and this is what takes a lot of time and confuses a lot of people. Without a high temperature to justify fouling (nor solids), I would expect a very clean diesel fuel – and, consequently, a very clean tube-side surface. You should be able to regulate the reboiler heat duty excellently with a variable speed pump on the hot diesel – or a flow control valve if you have the pressure driving force.

Again, I’m mystified by your statement that you are using a “large flow area” when you don’t have to. You can partition the U-tube bundle for multiple tube passes such that you literally “engineer” the design velocity you desire. The U-tube bundle is a concern to me when you are using steam (or a condensable) in the tube-side. In that application you have a 2-phase, vapor-liquid environment where it is particularly important to segregate the two phases in order to ensure designed heat exchange. The self-drainage of the tubes is very, very important in that case. However, with a circulating hot liquid in the U-tubes you should have a design picnic because you can accurately predict a lot of heat transfer and fluid flow results that you normally have difficulty with in a steam system. You certainly don't have to be concerned with non-condensables!

I think you have a very easy application – unless there are other facts that we don’t know about.

I hope this helps out.


Art Montemayor
Spring, TX

 

[NiravK]

Montemayor,

Thank you for your comments again.

Well, as far as certain basic information is concerned about the discussion, it is as follows.

Exchanger is BKU. Tube side passes = 4.

Diesel is heated and 525 oF. which is circulated through centrifugal pump.

There is a flow control valve in the discharge of reboiler on diesel (tube) side.

Allow. dP=15 psi.

Now, to make heat flux "low" (< 12000 btu/hr.ft2), i had to increase diameter of bundle (increase in heat transfer area and flow area).

So, pressure drop across exchanger decreased on tube side.

In general, design of exchanger should be such that calculated pressure drop is almost reaching to allowable pressure drop within the acceptable range of velocity. These approach will give high heat transfer co-efficient and optimum area of exchanger.

When i try to increase flow passes from 4 to 6, heat flux increases and goes beyond allowable limit. So, i could not increase it further.

The conclustion is that (which we all know).....

(1) There is direct relation between dP, velocity and heat transfer co-efficient.

(2) Increase in these parameters will decrease H.T.Area which is desirable. HOWEVER, it will also increase "Heat Flux" (Heat Transfer per unit area).

(3) So, when i try to utilise allowable range on dP & velocity to optimise the heat transfer area, it increases heat flux.

So, HEAT FLUX becomes "bottle neck" for the design, particularly for reboilers.

That is why i put question in my first post to have some more idea on "heat flux" constraints.

I hope i am making it more clear this time.

Thank you.

 

[25362]

When speaking of "film" boiling conditions one should remember that besides advising a maximum heat flux of 40 kW/m² (about 12,700 Btu/(h.ft²) for boiling organics in free convection, Kern also suggests a maximum heat flux of 65 kW/m² (~ 20,600 Btu/(h.ft²) for organics in forced convection, probably due to better agitation. Thus, it seems there is a margin to allow for some increase in the heat flux limit imposed on reboilers' design.

There are several Google sites on enhanced heat transfer in the boiling regime worth reviewing.

 

[StoneCold]

nirav2909
It apears that your problem is the large temperature difference between the diesel and the column bottoms. Can you send the diesel through a column feed preheater first to reduce the temperature. Or is there anywhere else you can use the diesel first to cool it off before it comes to your reboiler?

Just a thought

StoneCold

 

[jay165]

nirav:

525 F seems awfully high to be heating diesel (ie straight #2 FO). I've never used this fluid as a heat medium, but I would be concerned about thermal degradation and coking.

I could be wrong - maybe others can provide some insight.

 

[25362]

It appears nirav2909 hasn't told us the whole story. For example, whether the diesel stream could be used in series first in the de-butanizer reboiler and then in the de-ethanizing reboiler depending on the heat loads, and whether there is a de-propanizer in the middle or not.

The tube metal temperatures would be about in the middle since the individual heat transfer coefficients for the diesel stream and the boiling hydrocarbon side, would be similar.

As an example, a de-butanizer reboiler with a boiling temperature of 360 deg F, for a pressure of 115 psia, wouldn't be abnormal. If the diesel cools down in the reboiler by, say, 150 deg F, there would hardly be a temperature difference to drive heat at the cold end.

The diesel could then be sent to the deethanizer reboiler which is generally working at as low a temperature as possible (by using partial condensation at the top).

To j165, the diesel temperature is not considered of thermal cracking level. See, please, the draw-off temperature of light gas oil (LGO) from a topping crude oil tower (at about 23 psig) could be 500 deg F. A heavier gas oil (HGO) fraction would be removed at, say, 570 deg F, while the feed to the tower would be around 720 deg F.

 

[Montemayor]

We now know we didn't get the whole story in the original post. I believe there are still segments missing from the total basic data and scope. For example:

1) This is a project still in the design development stage; why is such a very hot Diesel reboil fluid being proposed? There could be other reasons besides the most obvious.

2) Flux is not the design problem. It still hasn't been explained as to why, in the course of the design, the tube bundle can't have the necessary passes. U-tube bundle length has not been mentioned and this is yet another tool to obtain the desired effect.

3) A kettle reboiler is under pool boiling effects since the convection currents are normally so small, they are negligible in stimulating a better film coefficient on the shell side. Consequently vapor binding or blanketing is a known concern.

4) No mention is made of the duty, the shellside or the tubeside flows or the estimated size of the kettle. If this is a huge reboiler duty, then the result should be a huge kettle size (complete with U-tube bundle). It comes with the territory. If the resultant size and cost seems too much, there is always the Thermosyphon alternative that can be applied - together with its tradeoffs.

5) There is a definite maximum temperature difference limitation on the reboiler's LMTD and I suspect that 500 oF Diesel is in violation of that limitation. The high Diesel temperature will also contribute to fouling as well. Again, to quote Kern (among other authors), "It is the phenomenon of vapor blanketing which poses the principal difficulty in the design and operation of vaporizing exchangers." It was true in 1950, as it remains true today. Heat flux is not a "bottleneck" in the design of reboilers - it is an acknowledged necessity to overcome the shortcomings of pool boiling.

6) I get the impression we're only getting part of the story since nirav himself states "When i try to increase flow passes from 4 to 6, heat flux increases and goes beyond allowable limit. So, i could not increase it further." This statement, and the lack of mentioning the obvious method of lengthening the tube bundle while lowering the Diesel temperature infers that there are pre-conceived design bases in place and there is a reluctance to follow the conventional design method in order to develop a smaller reboiler.

This is a design problem that has a definite design answer and resolution - albeit, with a large, bulky (& expensive) Kettle reboiler. If this results in a large tube bundle, then I also have problems in accepting a BKU as opposed to a CKU. I'm sure the maintenance department would rebel against this lower capital cost feature if they had a chance. There will be no practical way to test or repair the individual tube joints at the tubesheet.


Art Montemayor
Spring, TX