Performance Assessment of Vertical Thermosiphon Reboiler

[PetroDDP]

Once through /Natural Thermosiphon Reboiler where we know that it has got the impact of liquid level on its heat transfer rate and its performance.
For Atmospheric splitter column , where this reboiler is operated, has bottom baffle.Now when we want to assess reboiler sump level vis a vis heat transfer rate, how do we approach ?, we have steam flow FIC,Column bottom LIC. Normally thumb rule says 25-30 % level in thermoosiphon reboiler (atmospheric pressure)between tubesheet gives highest heat transfer rate,to assess it practically , can it be done via GA drawing of column and Reboiler and actual LT measurement.

There are various yardstick like osciallation phenomenan where we need delta P across Reboiler , in case where online PT is not available and process side TT is not available, how to assess that is there any scope for it performance improvment w.r.t. various fundamental parameters like weight fraction vaporised vis a vis liquid driving head % of tube length.As such reboiler is working fine which uses MP Steam , however, to identify the scope for MP Steam saving,assessing its performance from above angles as well as to assess whether it is having excess circulation/insufficient circulation etc. will give us the overall performance improvement scope if any.
How to do above practically ?

Thanks in advanced to all.
With Warm Regards,

 

[georgeverghese]

Vertical and horizontal thermosyphon reboilers have their turndown limits, and there is little you can do to change that - flow hydraulics in the transfer line back to the column goes into slug flow regime at below this limit. You've got to operate the overheads condensor and other intercondensors at a fixed minimum duty to keep reboil rate at above turndown limit. That was a choice made by your design team to reduce CAPEX at the expense of OPEX.

 

[shvet2008]

Termosiphon reboiler turndown limit is a retribution for low CAPEX.

To avoid slug flow regime in reboiler downstream pipe just try to add a new control valve in liquid line in reboiler upstream pipe. In this case you can adjust reboiler duty by decreasing flowrate and by this way to decreas heat transfer coefficient inside the safe homogenius flow regimes. I would recommend full bore control valve with handwheel for example vee-ball valve.

It will work in this way:
- operator decreases column performance
- reboiler duty and condenser duty decrease
- fluid flow velocity in reboiler downstream decreases
- flow regime in reboiler downstream pipe changes to unstable - for example to slug flow
- liquid level in column bottom starts to oscillate
- all column regime starts to oscillate
- operator goes to vee-ball and manually partially closes it
- total dP across reboiler hydraulic circuit increases
- fluid flowrate to reboiler decreases, but reboiler duty isn't changed!
- vapor fraction in reboiler downstream pipe increases
- wet surface in reboiler tubes decreases and avorall reboiler heat transfer coefficient decreases
- since total reboiler area isn't changed and heat transfer coefficient decreases - reboiler duty is constant
- since vapor fraction and temperature in reboiler downstream pipe increase - flow regime changes back to safe stable region (homogenius flow).
- liquid level in column bottom is stabilized

This little trick helps to avoid undesirable fixed limited duty and turndown limits of columns caused termosiphon reboiler flow regime. I saw the same problem in a Lieberman's book - try to find chapter 7 in Lieberman's "Process design for reliable operations". This control valve we
seek to apply in our projects. See an example.

 

[georgeverghese]

Shvet2008, Great idea - hope some folks can provide some feedback on actual operating experience with this low dp throttle valve on thermosyphon feed line.

 

[Roach44]

Shvet2008 is right on the money. Pick up Norm Lieberman's book and you'll find what you're after. Had the pleasure of working with Norm on a Vac OVHDs issue in a refinery in New Orleans. He is excellent. Make sure it is not instrument related issues. Could be, seen it before.