Steam control to reboiler

[C4Reactor]

Dear all,
There are two types of steam flow control into a thermosyphon reboiler. One is by having the control valve at the steam inlet line and the other is controlling the exchanger condensate outlet flow.
Based on your experience, which one is more realiable and robust?

Appreciate any comments

 

[rbcoulter]

I have never seen controlling the outlet condensate flow but it may be done somewhere. Usually the steam is controlled at the inlet so that the pressure in the
HX is regulated (usually shell-side).

 

[TD2K]

Controlling the steam inlet flow will give you better response IMO. Controlling the heat transfer by controlling the condensate level (flooding some of the tubes) is slower and gives very non-linear control response if the exchanger is horizontal. Horizontal thermosyphons are not common but are out there.

 

[22082002]

If you control the condensate flow, you will have condensate flashing across the control, which can lead to serious problem of hammering in the condensate heater....

 

[TBP]

I've seen applications where vertical shell & coil heat exchangers, and horizontal shell & tubes were controlled by condensate valves instead of traps. Both worked fine. The shell & tube application was a retro-fit to deal with HXs that were GROSSLY oversized. The shell & coil was to deliberately sub-cool condensate (for efficiency) before it was sewered. No hammering in either one.

It's important to select a control valve that's tough enough to take some flashing of condensate, although most of the time on the shell & coil, the condensate was cooled to the point where it wouldn't flash anyway (about 160*F), and this unit had a 125 PSIG inlet steam pressure.

 

[25362]

Either only condensate control or only steam control plus steam trap for condensate removal, can bring about noise and hammering when steam bubbles come into contact with cold condensate. Steam control at low loads, or under sudden drops of load or when the unit is largely oversurfaced may bring about condensate flooding and its problems as with condensate control. The main advantages of condensate control are: CV is smaller and robust, low-maintenance item, cheaper , easier to design, and always has positive pressure to discharge, a fact that is not always available with steam control. The disadvantages were mentioned by TD2K, plus unpredictable, sluggish response, thermal shock.

Although not all designs are equal, it appears the most accepted approach is to have a CV on the steam side (being steam flow the primary variable thus offering rapid response) and a condensate removal pot with a balance line floating on steam pressure downstream the CV, removing the condensate with a LLCV. The level on the pot being the same as the condensate level on the reboiler. In this way control over a wide range of loads can be achieved, especially when the reboiler has been oversurfaced to satisfy contingencies such as unexpected fouling, design safety factors, varying reflux rates, etc. offering rapid response and fine control when operating conditions fluctuate. This type of setup may be also of help in horizontal units.

Only condensate control or only steam control can bring about noise and hammering when steam bubbles come into contact with cold condensate. Steam control at low loads, or under sudden drops of load or when the unit is largely oversurfaced may bring about condensate flooding as with condensate control. The main advantages of condensate control are: CV is smaller and robust, low-maintenance item, cheaper , easier to design, and always has positive pressure to discharge, a fact that is not always available with steam control.

 

[Leclerc]

It appears that both types of control can be made to work, but the preference seems to be for steam control.

Particular circumstances would drive the choice the choice one way or the other. Three particular circumstances (all favouring steam control - but that's just coincidence) are:

1. If tubeside material is heat sensitive or likely to skin the inside of the H/E tubes, use a steam control valve to drop the operating pressure of the steam chest.

2. In some circumstances when there is sufficiently high pressure drop across the steam control valve, you could use a side flow of upstream steam to operate a pumping steam trap.

3. A steam control valve and steam trap combination will allow inerts(air) to quit the steam chest. I am not so sure how this is done with a condensate flow control valve.

 

[25362]

Other factors to be considered when selecting the control system on termosiphon reboilers are:

Operation on vertical units is sensitive to process parameters such as temperature or composition, because they interact with the pressure balance between the 'pushing' static head available in the column and the friction on the two-phase flowing loop.

These conditions may bring about oscillations in the process flow loop, poor tower control, and even low heat-transfer may result from mist flow in long tubes.

Working beyond a critical delta T between the wall and the boiling liquid would mean a reduction in HTC due to film-type boiling. The steam pressure, affecting the tube wall temperature, should then be carefully selected/ monitored.

Sometimes it may be advantageous to put the steam CV on cascade with the feed inflow to get some kind of 'feed forward' control, especially when the tower bottoms are removed on LLC.

Still, it appears that a combined control on both steam input and condensate removal may be the optimal approach.

 

[mur]

I have worked on Thermo syphon reboilers using both control systems.The major problem with steam inlet control is where a Tower duty is increased (feed rate) and the subsequent need to increase steam flow. Eventually you run the risk of losing the condensate level and blowing thru, (want to see a tower profile dissapear in a hurry try this).Using Steam Inlet control you have minimal or no warning when you are reaching this point.
The best control system I have experience of is to install a Condensate controller that is Cascaded from a Level transmitter on the side of the Reboiler. (The Level Controller is the Primary loop). This confiquration has been in operation in 4 Plants I have worked in over the last 20 years.When calibrated correctly you always have accurate indication of the condensate level.(also removes the Operator fear factor about blowing thru to allow maximising Optermisation of the Towers.

 

[kbander]

I've worked with both configurations and would have to say that controlling using condensate flow gives a pretty poor result. These systems are prone to blowing the condensate seal at high loads, and as a result, Q rapidly decreases. People only put these schemes in because they are cheap...not because they work well.

Controlling the steam flow is the way to go...but you need a condensate pot beside the reboiler with a level controller to complete the installation. The condensate level control valve should normally run wide open...and the pot will typically be full. When the steam valve opens, the level in the pot will drop. When you approach the low liquid level, the valve starts to close to prevent loss of the condensate seal.

Lots of Operators get confused by this scheme, thinking the condensate valve is incorectly sized because it will usually be wide open, yet the vessel will be chock full. They forget that the pressure inside the vessel is what gives the valve it's capacity...and this increases or decreases depending on how open the steam valve is. Need more Q...the steam valve opens...more pressure inside the receiver...more condensate flow (and yet the valve is wide open...as it should be).

By the way, if you want to use a heat-off scheme to reduce flare load, the "control steam flow" installation is also a better fit. Shutting a condensate valve does not immediately stop reboiler Q, where shutting steam off does.


I hope this helps.


Regards,

Bob