5 replies to this topic

[HLO]

Dear all,

 

I'm dealing with an overhead pressure control scheme (defined by others) for a debutanizer column that comprises: upstream valve to the air condenser, hot-vapor bypass and air flow rate manipulation, with a drained condenser, not flooded, as per the basic engineering. The vapor valve upstream the condenser regulates column pressure and the vapor by-pass controls pressure in the reflux drum.

 

I’ve read some literature and articles, and I could found the following:

• Drained condenser by-pass alone, is not recommended by Kister (Distillation Operation, page 539) because “it may not work” according to cited literature.
• Proposed scheme is recommended by Shinskey (Distillation Control), according to paper “Quantitative Comparison of Alternative Control Schemes for air-cooled condensers”, by W.L. Luyben. I don’t have the Shinskey’s book to further analysis, BUT in the paper always is assumed a subcooled reflux, so the condenser is partially flooded or not drained. This two valve combination has some advantages, and disadvantages.

I understand how this system works with subcooled reflux, but I have doubts about how functions with a drained condenser.

My assumption about how this scheme works (with drained condenser) is that if the accumulator pressure increase (because air temperature surge) the hot-vapor bypass will tend to close and fan speed will rise. In the opposite, if accumulator pressure decreases, vapors from tower top will pressurize the vessel, through opening hot-vapor by-pass valve.

 

My question is: do you think my assumption is correct? What is your opinion about to install this scheme without a partially flooded condenser?

 

Thanks a lot.

Edited by HLO, 12 March 2013 - 05:05 PM.

[Bobby Strain]

Nothing wrong with it. Just make sure the bypass is big enough. It is best to seal the drum inlet from the condenser. You need a substantial bypass to make it effective control. The best designers place the pressure control on the liquid from the condenser so that the control is effectively a flooded condenser. In this arrangement, the hot vapor bypass has no control valve. The liquid head above the control valve provides the necessary pressure differential to operate the drum at the condenser inlet pressure. Simple. Even an idiot can't mess it up.

 

Bobby

[HLO]

Bobby, thanks for your comments. Drum inlet is sealed and I agree that flooded condenser is the best approach.

[paulhorth]

HLO, Bobby,

 

I have seen several slightly different schemes used for control of the condenser on depropanisers and debutanisers, and I have never been very clear on which was the best to choose. I have designed a couple of such systems, using ( from memory) the scheme propopsed by Bobby with a valve on the condenser outlet and a hot vapour bypass, but I cannot claim operating experience, so it's good to hear Bobby confirm that this scheme works.

HLO mentions a valve in the vapour overheads line upstream of the condenser, as well as the vapour bypass.. I did not use a valve in this location, it seems to me not necessary, I would appreciate the views of others on this point. Also I would include a means of controlling the aircooler fans to allow for air temperature variation. This can be simply on-off control on one of the fans.

 

On a recent project I was brought on when construction was almost complete, my job was to write an operating philosophy, The depropaniser and debutaniser on this project had valves in the hot vapour bypasses but not in the condenser outlet lines. I struggled to describe how this would work, but in the end I convinced myself that it was OK. I haven't had any feedback from the plant once in operation ( a few years now). Do you think this scheme would work?

 

I have a paper which describes about 15 different pressure control schemes for fractionator overheads, so there's plenty of scope for discussion here.

 

Paul

[SSWBoy]

Normally I wouldn't recommend throttling on the liquid for a condenser because it means you have to elevate the exchanger above the drum...but since we're already dealing with an air cooled exchanger, no problem. I wouldn't actually describe bobbys scheme as a hot vapour bypass really, as all that line does is equalise pressure between overheads and reflux drum.

There are various types of pressure control schemes lumped under hot vapour bypass, that are actually quite different. You can have quite a small flow of hot vapour bypass into the drum and sub-cooled reflux, and the theory goes that if the drum liquid is kept calm only the liquid at the surface is in equilibrium and the liquid below that is sub-cooled. But there's a whole load of painful design measures to get it to work. Basically it's a ball ache,

At the other end of the spectrum (which I prefer) you can have a large flow of hot vapour bypass and intentionally mix it with the sub cooled liquid, I.e. by sparger in reflux drum. This is a lot more stable, and generally makes more sense!

[S.AHMAD]

HLO
1. There is nothing wrong with you assumption. Let me explain
2. The pressure of the accumulator is the vapor pressure of the liquid out of the condenser. Vapor pressure is a function of temperature. Therefore, the accumulator pressure is actually related to the condenser outlet stream temperature. In other words, the accumulator pressure is controlled by controlling the temperature of the stream before entering the accumulator. The temperature is controlled by hot-vapor and fan speed.
3. When the stream temperature exit the condenser is low, so does the accumulator pressure since the vapor pressure is also lower. Therefore to correct this pressure, the hot-vapor bypass is opened so that more hot vapor to rise the temperature of the condenser outlet stream. Simply say, accumulator pressure drops because condenser outlet stream temperature lower.
4. Similarly, I can say that pressure rises because the condenser outlet stream temperature increases. Therefore, to reduce the pressure you need to reduce the temperature by reducing the hot-vapor bypass or increase the fan speed.
5. Hope the above simple explanation makes you more comfortable.

Edited by S.AHMAD, 13 March 2013 - 09:07 PM.