Two Columns - Two Processes, One Pump? 1

[jennymd]

Hi,

I would greatly appreciate some advice in HOW and WHY it would be better to have TWO PUMPS or ONE PUMP (both with spares) that feed two columns (different process) and a filtration system.

I am designing an application that has two main sections: H2S absorption (tray column) and SO2 absorption (packed column).

I have to send solution at different flowrates from the same source to both columns (+ some solution to a filtration section).

Currently, the design is based on a dedicated pump that sends solution to the H2S absorber and a different pump that sends solution to the SO2 absorber and the filtration system. The H2S absorber operating pressure is around 85 psig, the SO2 absorber is operating at vacuum conditions.

It has been proposed to just design one pump for both services. The equipment will be centrifugal pumps.

At this point we don’t have budget constrains and I think that having two pumps make the process more flexible.

Thanks in advance for your comments.

 

[ash9144]

Using just one pump would be budget benefit only.
Using two pumps would improve your controls. With that difference of end use pressure you could have a situation where the control valves would fight each other causing fluctuations in feed flows. Also having two pumps you could have better isolation for the two scrubbers independent of the other if you ever ran one without the other.

 

[sshep]

Jennymd,

There are capital vs operating trade-offs with the decision, and so sometimes no right answer. If both process operate continuously (or always at the same time) and require the same discharge pressure, then one pump (spared of course) should be adequate. If the columns operate independently or if one has a higher head and/or flow requirement than the other, then two pumps should be considered. If you give the head, flow, and stream factor for each service then it may be obvious whether one or two pumps should be recommended.

As example of the trade-off, consider the case of combining the column reflux and distillate into one pump which is common practice. This is obviously done for capital cost, however, the truth in some cases is it that the distillate needs more head than the reflux. The decision to combine into one pump has actually traded increased operating cost (energy) to reduce capital cost. Sometimes the distillate flow is small and high head, while reflux is large and low head so the trade-off doesn't make sense and seperate reflux and distillate pumps are used. As further example seperate bottoms (high head, low flow) and forced circulation reboiler pumps (low flow, high head) are almost always the case when forced circulation reboilers are used.

So basically it comes down to sizing and costing two pumps (with piping, electrical, etc), and then comparing the economics (capital and operating costs) vs one pump. If two pumps makes sense, there is still the option of using a common spare sized to meet either service. I find this common spare option is often used in the column examples I cited above. This is my experience anyway.

best wishes always,
sshep

 

[roker]

Hello,

1. if the source has some pressure may be you dont need a pump to the SO2 column.

2. with one pump, the pump is rated to the max discharge pressure and than you reduce pressure on the control valve to the SO2 column.

3. from Hazop point of view, with one pump and a common discharge line a reverse process flow might occure from the H2S column to the SO2 column.

4. so two separate pumps seems to be a better design.

regards,
roker

 

[jennymd]

Thanks for your comments. We are probably going with two pumps.

 

[pumpking]

Two pumps is better, 1 pump is possible by sizing for maximum conditions, and then controlling with either control valves, or safer, orifice plates to make sure pump remains on its curve and avoids possibility of operators 'tweeking' in the middle of the night as they usually do.

Alternatively, even better, and even more efficient and energy saving, use a VSD drive to control speeds !! cant remember exact figures, but slow a motor by 20% rated speed and save c.60% energy costs !!


Ash Fenn

http://www.cdrpumps.co.uk