Unexpected Packed Bed Scrubber Performance

[DonyWane]

We recently began startup testing of a new scrubber configuration at our plant. Details:

Column: 6" Pipe
Packing: RMSR 40 (essentially 1.5" IMTP) Pfactor = 24 ft-1
Packing Height: ~4 ft.
Test gas: Nitrogen quenched to approximately 100F
(gas temp could be higher)
Test liquid: 20% KOH (with cooler to keep temp 60-70F)
Gas Flow: ~10 SCFM
Liquid Flow: ~9 GPM

During testing, we saw some anomolous results (repeatable). After 10-15 minutes of low DP (~.6" water), we saw a pressure spike to ~7". Within 30 seconds, this spike had relaxed to 2.5" and after another minute had fallen to 1.5". After this 1.5 minute period, another spike occured to essentially 7" again. This process repeated for ~1 hour - the period of the spikes was almost constant at 1.5 minutes and the magnitude of peak and trough varied by only +/- 0.5 inches.

Increase in gas flow rate to 12 SCFM appeared to reduce the peaks to 6". Decrease in gas flow rate to 6 SCFM appeared to increase magnitude to 8". Neither change greatly affected the periodicity of the DP spikes. Oscillation stopped around 5 SCFM.

We are using a differential pressure transmitter for DP and nitrogen purges on the high and low tubes to mitigate against KOH accumulation in the tubes. Increasing the purge rate did not affect the period of the oscillations.

Any ideas as to the cause of this phenomena??

 

[sshep]

Hey DonyWane,
I think this hydraulic cycle may best be diagnosed and mitigated by focusing on the liquid path. It was also curious that the liquid rate was not adjusted as part of your tests.

Can you detail these items (bottom up):

1) How is the bottoms liquid drained from the tower?
2) Is there a bottoms level controller or trap? Is there a sight glass?
3) If just a gravity drain is there a seal loop?
4) How much height for liquid hold-up is in the bottom?
5) What is the design of the bottom packing support? How much open area?
6) Is there anything besides packing installed (wall wipers, thermowells, etc)?
7) Is there a top packing holddown?
8) How is the liquid distributed onto the top of the bed?
9) How are your dP taps and instrument arranged (i.e. you seem to have considered a problem with this)?

Thanks, sshep

 

[aspearin1]

It sounds like you may have some capillary forces creating a liquid barrier in your column. Low air flow rates allow the liquid to fill the capillary zones in your packing (the porosity), creating a vapor lock. Pressure then builds up until the vapor lock is broken and the gas is then allowed to escape. By increasing the flow of the gas you are preventing this from occuring, but you are likely creating channeling in your packing, which decreases your gas-liquid contact surface area. I would suggest investigating different packing material, perhaps more open and/or hydrophobic.

ChemE, M.E. EIT
"The only constant in life is change." -Bruce Lee

 

[DonyWane]

sshep...

As for the liquid rate...we were actually testing another part of our system and only noticed these anomolous results later (because we were monitoring the entire system). We will vary liquid flow in upcoming testing.

Now to your questions...

1) The bottoms from the scrubber collects in a 6" tank immediately below the scrubber and is gravity drained from that tank through a seal leg and into a larger collection tank (that is vented into the offgas of the scrubber).

2) There is not usually a sight glass - but in testing we can install one (permanent installation might be a problem). There is neither a level controller or a trap.

3) There is a seal loop that the bottoms flows through - that is approximately 4 feet or so.

4) We can stand a little less than 20 inches before we get into the bottom of the packing.

5) The design is a weld-rod cross-hatch that raises pseudo-hemispherically in the middle. The free area is large. In off-site testing, we saw little DP across the bottom packing support.

6) Just packing

7) There is not a top packing holddown.

8) The liquid is distributed through a nozzle that resembles a shower-head. We were using a BETE MP nozzle, but saw too much energy in the fluid. This is a more laminar flow into the column. It is angular with some spray on the side of the column, but most onto the top of the packing.

9) The dP taps tie into two 6" pipe crosses, one above the packed column and one below (which is just above the bottoms collection tank). Hydraulic analysis shows that at our flow rates, we would expect no liquid level at the height of the lower (high side) dP tap.

Thanks!