Vapourised Feed v Reflux Ratio 2

[mrtangent]

Dear Sirs,

Can anyone provide me with a decent explaination of the following. I have a decient ASPEN simulation of a distillation column matches composition / temps/ pressures RR and boil up. Its closish to idea to system (similar chemicals etc) - but we do have activity coefficients etc programmed it.

So as I increase Feed vapourised fraction the Reflux ratio goes up.

I think this does make sence in terms of operating lines etc. or am I being thick !!

ie you vapoursie the feed so more composition in the top so you must need more RR in the top.

The top and bottom composition are fixed by the simulation and the pressure drop is fixed by the simulation.

 

[Montemayor]

james1030bruce:

The way I learned bout this at Texas A&M is the best method I have seen through the many years I spent out in industry. I think it is a tribute to Dr. Holland, one of my teachers, who was a past master at multi-stage distillation.

Basically, think of any distillation column as comprised of two sections: a stripping section (below feed tray) and a rectification section (top of feed tray). Common sense tells us that under pure, sub-cooled, liquid feed the stripping section gets inherent reflux. The rectification section, however, requires reflux via the condenser load (which increases reboiler duty). Now, start to heat up the feed, causing a fraction to be vaporized. You are now putting prior reboiler duty into the feed pre-heat duty and subjecting the rectification section to handle more vapor (increasing the the top section diameter to handle the additional load) with a related larger amount of reflux to handle the additional vapor load. At first, this is going to look as rather un-attractive engineering: you feed more vapor (reducing the reboiler size) but you increase the rectification section size and you require more condenser duty. However, the trade-off (as there always is one!) is that your overheads purity starts going up dramatically with much less height (mass transfer trays or stages). So the height is lessened, as well as the reboiler load. This is the expected mannerism of a distillation tower that I've always experienced and witnessed in the design and operating levels. There may be exceptions to the above general conceptual rules, and I would love to hear about other forum members' experiences and key learnings in this matter. I consider this subject of conceptually visionalizing the operation inside a distillation column to be of major importance for an engineer in order to dominate and anticipate results and problems in such operations. In effect, what you are doing mentally is "simulating" the various effects that can occur within a distillation column. I consider this mental exercise to be far superior than any simulation printout - mainly because it inherently proves that the engineer generating both is totally in mental control of the operation and its correct interpretation.

I hope this experience helps.

Art Montemayor
Spring, TX

 

[sshep]

I think the result you see can be readily visualized on a McCabe-Theile diagram with the feed quality change accounted for by the "q-line". Your situation is made a little more difficult to solve graphically because you are holding the product compositions fixed- a graphical solution must itteratively adjust the operating lines to exactly match your existing number of stages, feed location, and bottom product composition for the new feed condition.

A simple way to demonstrate the result you see without itterations is to keep the top operating line constant, but plot a new lower operating line based on the vaporized feed condition (pt #1) and desired bottom composition (pt #2). When you step off your existing number of stages you should find you have arrived short of the desired bottom composition. Remember to jump to the lower operating line as soon as you have stepped off all the top stages, even if the lower operating line is still close to the equilibrium curve (i.e. a pinch point).

Even with the modern simulation solutions available today, graphical visualization is still often the most convenient way to identify possible pinch points, and feeds which are poorly located or of the wrong enthalpy- i.e. a picture can give you that mental control over your process as discussed by Art.

 

[DickRussell]

To add to what already has been said, keep in mind what has to happen in the stripping section in order to meet the bottom specification, presumably some low concentration of a light component in the bottoms. To do the required stripping with a certain number of stages requires a certain V/L ratio, reflected in the slope of the lower operating line on the old McCabe-Thiele diagram. Adding heat to the feed often will but won't always reduce the reboiler load by the same amount. You have to consider also the fraction of the feed that is removed overhead. Consider the extreme case of a very small distillate rate, with most of the feed going to the bottoms. The bottom specification requires a certain boilup, and that vapor load propagated upward to the condenser is almost entirely returned as reflux. This means that the rectification section runs at nearly total reflux (very high, but mostly meaningless reflux ratio), and adding any vapor load above the feed does little to improve the top product.

As the others have suggested, use graphical representations of the distillation for interpretation of what is happening. I often normalize stage compositions from simulation output so as to study the separation between key components (sometimes multiple such pairs of keys), and plot those normalized X-Y points on a McCabe-Thiele. A line connecting the X-Y points gives the operating line for that binary, while the intersections of horizontal and vertical lines drawn through the X-Y points define the operating lines. The picture helps to locate pinch points, improperly located feed tray, and areas where more traffic either is needed or won't really help the separation. It also can give an indication of whether more or less feed heating is appropriate.

 

[pmharper]

In short terms your are feeding your column at a sub-optimal point (compared to the initial simulation). In order to investigate the behaviour I suggest creating a small series of simulations where you vary the amount of feed preheat and feed-location and record the resulting RR.

You should see the optimal feed point (minimum RR) "moving" as feed preheat is varied.

Best regards

Peter M. Harper
Principal Consultant, CapSolva

 

[alsayed]

You are right more vapors needs more reflux to cool it and less reboil.

You specified the top concentration as fixed, but your RECOVERY in the top of your product will go up (quantity of your product) as you vaporize more you can rectify more.

In reality your limitation is the top cooling system and flooding in the column as your vapor and liquid traffic increase.

In the end your column optimum operating point is a balance of operating cost and product recovery while having stable operation.