Stabilizer Reboiler (direct fired) 2

[chemieingineur]

I am evaluating a stabilizer design that has a direct fired heater (reboiler) with pump. This design combines takes the condensate off of the bottom and some of the condensate goes to a pipeline, but the rest is fed through the reboiler and combined with the liquid from the last tray!

There is no baffle in the tower and no place for liquid accumulation in the heater.

I have checked Kister's book 'Distillation operation' and could not find any real designs where the product from the reboiler is combined with the liquids from the bottom tray.

Could you please let me know if there is such a design? I have not been able to find anything in literature that would suggest this would give specification product.
Any thoughts would be welcomed.

 

[skoutso]

As I understand you are talking about a forced recirculating type, fired-coil reboiler. The reboiler feed is the trapout liquid from the bottoms tray and the recirculating liquid from the reboiler outlet. The reboiler outlet is returned below the bottom tray. A single nozzle at the tower bottom is used to draw off the bototms product and the reboiler feed. In such an arrrangement, the advantage of having the reboiler as an additional theoretical stage is lost. When the reduction in the theroretical stages is critical to the tower design, a baffle is used to separate the bottoms product from the reboiler feed and thus to maintain the effect of one theoretical stage for the reboiler.

Most simulators will, by default, assume the reboiler as being one additional theoretical stage (100% efficiency) and you should be very cautious.

I have seen quite a lot such arrangments (older designs) but I have never designed one without the baffle.

Hope it helps!

 

[25362]

Since I don't understand all of your expressions, please tell me what is your meaning of "condensate". Then kindly tell us whether the fire-heated reboiler produces vapor and where is this vapor flashed off. Any details on the process you may add will be of help to better understand your query.

 

[chemieingineur]

Thanks skoutso for the information. Do you know where I can get specific design information on the forced recirculation design? This would be helpful.

 

[chemieingineur]

Reply to 25362:

The condensate is the bottoms product which meets an RVP specification. The design I am referring to has only one outlet from the bottom of the tower. This bottoms product is then split into two streams: one goes to the pipeline and one if fed through the forced circulation reboiler. The two phase mixuture (of which about 20-25% is vapor) then enters the bottom of the tower and is combined with the liquid from the bottom of the tray.

The simulation was modelled in HYSYS but only as a normal kettle type reboiler. However, the reality is that the bottoms product is a mixture of the fluid from the reboiler and the liquids from the last tray. The equilibrium conditions would be different for this scenario than for a kettle type reboiler.

 

[ProcsssDr]

I don't see any particular problem with this design. It will work although you might want to add another tray or two to the column. The simulation assumes a theoretical tray for the reboiler that is not accounted for. Also, the actual reboiler duty may be slightly higher (but not much) than calculated because of the mixing of the reboiler return liquid with the bottom tray liquid. I assume there is level control on the bottoms controlling the liquid rate of the product withdrawl. How do you know if the bottoms product is "on spec"?

 

[25362]

As ProcsssDr says this is not an unusual design. Although it represents less than one theoretical tray, don't forget that the "colder" liquid dropping from the last tray undergoes some vaporization when meeting the "warmer" disengaged liquid coming from the reboiler.

Such an arrangement can be modeled in a similar manner than a thermosyphon system recirculating without baffle. For that purpose I suggest you get hold of the Hydrocarbon Processing article of June 1995 titled: [b[Effectively design and simulate thermosyphon reboiler systems[/b] by Martin and Stoley, which presents guidelines for modeling systems like these.

 

[skoutso]

chemieigineur

The RVP specification is customary in condensate stabilizers. Follow the 25362 lead for the design of such systems. The reboiler duty is fixed for a certain separation. In your case assume that the RVP spec stands for the bottom tray liquid. The bottoms product will be of lower RVP. Because of the recirculating liquid the vaporization ratio in the reboiler is kept low (around 30% max.). On this basis you can calculate the necessary feed to the reboiler. The bottom section should be sized for a liquid holdup (LLL to HLL) of 5 minutes on the vaporized portion and 2 minutes on the bottoms product (as a minimum if it goes to tankage).

 

[chemieingineur]

Skoutso you have provided two liquid hold-up times; one for the bottoms product and one for the vaporized portion coming from the reboiler. This would indicate a baffled bottom section of the tower if I interpret this correctly. Yet, the design I was given does not have a baffle. I would prefer the baffle because it simplifies the process and would make the control of the bottoms product easier.

 

[skoutso]

chemieingineour

The total hold-up is the sum of the two hold-up times. Hold-up times provided refer to an unbuffled bottom.

 

[sshep]

We have several forced circulation direct fired reboilers exactly as you describe (i.e. unbaffled).

The only function for a baffle in a forced circulation system is to get the extra stage of seperation. To accomplish this the baffle can be an underflow type. The reboiler inlet is taken from the bottom tray downcomer side, and will also include excess liquid fraction from reboiler via the baffle underflow. The reboiler discharge is to the other side of the baffle, and the product is also taken from that side. All of the product has therefore been once through the reboiler.

One possible disadvantage of the baffled system in your case is that a second product pump is required, as the reboiler circulation pump is now handling a composition different from the product. This may not really be a problem as the reboiler circulation will normally be high flow-low head, and the bottoms product (probably) low flow-higher head. Trying to perform both functions with the same pump may result in higher horsepower design than using two pumps. If two pumps are indicated by the service then a baffled design may be worthwhile, otherwise that extra stage of seperation is probably not worth the cost of a second pump, etc.

just a thought, sshep

 

[chemieingineur]

Reply to ProcessDr and 25362:

I have one more question on the tower size for this type of design. What tower size have you come across based on your experience? Kister states that the unbaffled design is applicable in tower diameters of less than 3 feet. (He also states that this design is recommended for fired reboilers which is one reason this design was chosen). I am assuming that this design, according to Kister, would be for towers less than 3 feet in diameter even when used with a fired reboiler.

Any recommendations on this?

 

[sshep]

We use forced circulation direct fired reboilers without any baffle in columns 8+ ft in diameter.

A 3ft diameter column is pretty small. In the most common baffled systems (natural recirculating vertical thermosyphon reboilers with an overflow baffle), the baffle is mainly to insure a constant head for stable thermosyphon action. Constant head is unneccesary in a forced circulation system. The other benifit of a baffle is an extra stage of seperation, but only if properly designed to return excess product side back material back to the reboiler feed. In most cases it would be borderline economics to install a baffle in a forced circulation system of a 3ft diameter tower. You are trading off a number of possible items for the benifit of that extra stage of seperation: less bottoms hold-up per ft of sump height, entry access to bottom is virtually impossible, may need extra controls, seperate product pumps and reboiler pumps,...

In a direct fired system it is safety critical to insure that the feed to the fired heater is always maintained. Basically the feed side (pump suction) level needs to be well measured (often using 3 transmitters) so that the pumps do not lose suction. In an unbaffled system the bottoms level is measured and controlled directly. In a baffled system the reboiler discharge must go to the product side with the excess returning to the reboiler via underflow or overflow. The underflow design as described above has the benifits of allowing direct measurement of bottoms level, the product side volume is available as hold-up, and always insuring that there will be a return to the reboiler feed pumps.

Best of luck, sshep