Columns in series - what pressure limit?

[Kristof]

Hi,

I have a method that uses two column in series - the first one has higher pressure rating (let's say - up to 250 bar), the second - lower (let's say - up to 100 bar). What pressure limit should I set? a) of the first column, because it is the one actually causing overpressure detected by sensor in the pump? - or b) of the second column - becuase it would be the one getting destroyed by overpressure?

My first reaction would be the answer b) just to be safe. But then I started to think that is the first column that causes the resistance detected by the sensor, and pressure drops after mobile phase leaves the first column thus causing no risk for the second one...

Sanity check, please?...

[tom jupille]

To be absolutely safe, use the lower of the two pressure ratings.

If you *know* the actual pressure drop across each column (i.e., if you know the flow resistance of each column), then you could go higher. You can get that information by running each column individually at the same flow and measuring the pressures.

For the sake of argument, let's assume that both columns have the same flow resistance. Flow resistance is additive, so half of your pressure drop will occur across the first column, and half across the second. If your pressure limit is 200 bar, the inlet of the first column will see a "gauge pressure" (difference from atmosphere) of 200 bar, and the second will see 100 bar. In that situation, you would be safe at 200 bar.

In the hypothetical worst-case scenario, the first column would have no flow resistance, so all of the pressure drop would occur across the second column. In that case, for a 200 bar pressure limit, the inlet of the first column could see a gauge pressure of 200 bar and the inlet of the second column would also see a gauge pressure of 200 bar. In the hypothetical best case scenario, the second column has no flow resistance, so all the pressure drop occurs across the first column. In that case, a 200 bar pressure drop would give a 200 bar gauge pressure at the inlet of the first column and 0 bar gauge pressure at the inlet of the second column.

[biochemist]

Can I revive this topic and ask:

If I have 2 similar columns in series (200 bar max backpressure each), is the pressure limit I set

a) 200 bar? or
b) 400 bar?


In other words, is the "maximum" pressure a column can withstand measured as P inlet - P outlet?

Or P inlet against P atm?

Sorry as I couldn't really grasp the previous post.

[tom jupille]

In other words, is the "maximum" pressure a column can withstand measured as P inlet - P outlet?
Or P inlet against P atm?

That depends on what part of the column will fail first. The integrity of the packing bed is limited by the pressure drop (P inlet - P outlet). The integrity of the fittings is limited by the "gauge pressure" (P inlet against P atm).

From my experience, I would say that most fittings are good up to 300 - 400 bar (we're talking conventional HPLC here, not "UHPLC"), so a 200 bar limit would most likely reflect the stability of the packing bed. Running at 400 bar would probably not collapse the packing bed, but I wouldn't be surprised if you were to blow the tube ends out of their ferrules.

[biochemist]

Thanks a bunch, Tom!

The columns I need to join are 2 identical gel permeation columns, with a pressure limit of 1000 psi each (68 bar).

So I guess if I were to connect them and set the maximum pressure as 2000 psi (138 bar), it should be ok, right? Since it's likely to be below the pressure limit of the fittings.


Or should I set it somewhat lower (e.g. 1500 psi) just to be safe?

[tom jupille]

I always try to allow safety margin; if it were my problem, I'd run at 1500 psi.