Reboiler Steam Flow Control 5

[katmar]

All the reboilers I have worked with have had the flow control valve in the steam line feeding the reboiler, and a steam trap in the condensate exit line.

However, I recently saw a P&ID which showed the flow control valve in the condensate line, but with the flow element still in the steam line. I suppose this has the advantage of making the valve smaller and saving a steam trapping station, but I would be interested to know if this works without blowing steam into the condensate line?

I guess the control valve would cause the condensate in the reboiler to back up making the reboiler a "variable surface area" device, whereas if the valve is in the steam line it is a "variable pressure (or temperature)" device.

I would be most grateful to hear from anyone who has operated a reboiler in this way.

regards
Katmar

 

[TD2K]

I haven't operated any reboilers designed this way but have heard about them. Yes, as you suspect they are operated as variable surface to control the heat input compared to adjusting the temperature difference (which is what a control valve on the inlet gives you).

Response time is typically slower as you have to drain or raise the condensate level to change the portion of heat transfer surface in the steam zone. Also, you can have corrosion problems at the steam/condensate interface depending on how well you control your steam/condensate water chemistry.

Typically, there is a low level override on the condensate valve to avoid blowing steam into your condensate return system.

 

[pmureiko]

This is a situation of expecting many things to be done without enough measured or manipulated variables. You are correct that there is no mechanism to prevent steam from blowing through to the condensate line. I suggest an external liquid leg with steam trap be placed on the condensate outlet or a level instrument to measure the liquid level as TDK2 suggested. Even if the control valve fails open, the steam trap will prevent blow through. You may also want to relocate the control valve, albeit larger, back to the reboiler inlet. This method hase some weaknesses, but nothing as bad as what you'd get with this scheme.

This scheme has a potential failure built into it that could be very aggravating. The tower will respond to a change by increasing the heat input and steam flow rate. The condensate drain valve will open and reboiler could lose the liquid leg and get steam blowing through. At this point the steam is providing sensible and no latent heat to the tower. The controls will ask for more steam flow and increase the amount of blow through.

Also maintaining a water level in the reboiler promotes corrosion and after a while you'll get a leak in the reboiler and something is going where it's not wanted.

 

[sshep]

I just started up a whole slew of columns that had reboilers that were designed as variable area. In all cases the level control was being used to prevent a stall situation because the temperature on the process side would give a chest pressure insufficient to return the condensate if the whole area were used (especially when clean). The slower response and poor operator understanding are the most significant drawbacks to variable area. Blowthrough (from poor tuning, controller operation in manual, etc) caused hammering of the condensate line, but I did not consider it a serious consideration (i.e. always corrected within a few seconds) in my cases.

I have also seen a number of hybrids (valves on both steam and condensate) over my career. In all these cases the condensate valve controls the level to prevent stall, with the steam valve controlling the heat and giving fast response to demand changes. Some of the columns I cited above were later fitted with valves on the steam to give the faster response.

 

[byrdo]

One more thing to consider with a variable surface area exchanger:
As you decrease area, you increase heat flux and skin temperatures.

Columns with fouling problems like C3 or C4 splitters may have shorter runs or increased fouling because of this control scheme. I have not experienced this personally, but this is what I would expect to happen. Any thoughts on this?

Regards,

Matt

 

[25362]

By not regulating the steam supply one provides the reboiler with the maximum pressure (and saturation temperature) available. So instead of changing the delta T, one changes the surface. Changes in HE surface would be "linear" if the the reboiler is a vertical cylinder.

A demand for the "highest" and "constant" steam temperature for the particular duty could serve to explain the reboiler's controlling procedure.

Besides, by not regulating down the steam pressure, one may obviate the need of using a special pump/trap for the removal of condensate.

 

[hacksaw]

katmar,

you previous experience is valid, but the new scheme you have described is inherently unstable and should not be used.

 

[Montemayor]

katmar:

Wow, I couldn't believe that so many talented engineers would be interested/experienced in this technique. I've had fun reading this thread because I stuck my head in a noose as a young man when I introduced the concept in a vertical thermosyphon reboiler in 1970. The application was a steam-stripping tower in the Furfural process at Cedar Rapids, Ia. The tower was approximately 9 ft diam x 40 ft high. The bottoms product was water with approx. 0.5% acetic acid + some bad actors. I put in the reboiler to avoid the straight steam sparging and cut down the bottoms flow which was going to a treatment plant. As 25362 astutely points out, the steam chest pressure was essentially full, steady line pressure - so my heat transfer rate was the highest available at the delta temperature available. The tower operated on a very reliable, steady state feed rate and worked 100% of the time for 300 days a year. I used a condensate drum besides the reboiler that had a Fisher 2500 LevelTrol - which I empirically adjusted to the condensate height I found gave the heat transfer results I was seeking. The steam inlet was metered - but this was mainly for monitoring the operation. The heat transfer rate (actually, the reboiler duty) was controlled by the net available steam vapor heat transfer area that was left after the condensate level was set. The LevelTrol had nothing more than pneumatic Proportional Set control - as many old timers will remember.

The operation ran like a swiss clock dream. My reboiler was stainless on the tubes and my steam was controlled with a good quality - something that seems to give young engineers today a lot of trouble - but I had no corrosion on the steam side in the reboiler. However, I must repeat: my operation was a very steady one. I later designed and installed about 2 more reboilers based on condensate control - and they also worked very well.

I'm pleased that sshep has had similar experience; I agree with him that response time and operator training / understanding of the system are potential problems in some applications. However, I believe they can be overcome. There are net advantages in operating a reboiler this way. The simplicity cannot be overstressed. However, steam quality will -as in all cases- always be an issue. But, that's not (& never has been) my problem. I mention this because I hear so many nightmare stories of steam corrosion and contamination nowadays. When I tell young engineers how I ran process plants using nothing but steam to power my electrical generators, run compressors, turbines, pumps, Lentz engines, and overhead flat-belt pulleys, they walk away shaking their heads and mumbling as to why I never had the corrosion problems they have in the "Space Age". I guess I was just lucky.

I thought you might be interested in this experience.


Art Montemayor
Spring, TX

 

[RogerH]

Katmar and others
UOP is often providing the control as a FC on condensate. For low pressure applications this is fine. For MP/HP-steam applications it is normally combined with a PC on the steam inlet.
I have operated two different plants with this configuration and so far I can say I havent had any problems with this, rather the opposite. They are easy to startup/shutdown since one can control the (conendsing pressure) temperature and ramp-up in a controlled fashion.
With the addition of a PC on the steam inlet variations in steam pressure is handled smoothly(for those plants that have some instabilities in steama supply pressure).
The FC measurement on outlet condensate prevents (as long as the FI works) steam blowthrough so this is the "protection" againts upsets in the condensate system.
I have experienced more problems with the traditional type (steam FC/condensate pot+LC)than with the above mentioned type, so if I have the choice I would go for the "UOP" type.

 

[25362]

My plant experience with reboilers is similar to that of Art Montemayor. These are some more comments.

Oversurfacing of reboilers is a common -not always economical or safe- design procedure. It is frequently provided for reasons such as: undeterminate fouling, varying steam pressures and degrees of superheat coupled with J-T cooling on expansion, unsure OHTC, providing ample reserves to keep product in specification, variable input and quality of feed to the main tower, to name some factors. Steam pressure regulation to reduce loads, especially in oversurfaced units, may lead to a (low) chest positive pressure insufficient to release the condensate into its return line.

On the issue of control with flooded tubes, depending on the geometry of the reboiler and the rate of condensation, it is true that hydraulic response-times may be quite different: a load increase (a drop in condensate level on a partly flooded shell) would take less time than needed to reduce the load (flooding the shell with condensate), for the same but opposite load changes. Whatever type of control, load changes should be gradual to avoid an undesirable erratic and fluctuating equipment response.

Usually emphasized disadvantages of a flooded unit have been made largely irrelevant. For example, corrosion, it wouldn't appear when steam is free of oxygen and carbon dioxide; thermal shock, loses its importance if the unit has been designed to stand thermal differential stresses; sluggish response, it is of no great significance when condensate levels (i.e., operating loads) aren't frequently changed. Noise and hammering, that can be felt inside the reboiler, especially when the steam is inside the tubes, and which may lead to equipment damage, can be mostly prevented by avoiding condensate level surges.

The use of a LLC condensate pot with a pressure balance line (plus a valve to purge inerts), not only provides changes of HE surface while using a high steam-side positive pressure for removal of condensate, it also enables the elimination of steam traps. Some operators don't like steam traps as process-control fixtures because, they say, traps may show erratic operation and may need frequent maintenance.