Waters 2695: Odd issue with pressure fluctuations

[juddc]

OK guys, I'm not quite stumped yet but this is one of the more challenging issues I've run across and am interested in hearing what chrom-iverse has to say.

The patient: Waters 2695, about 5 years old, under a service contract the whole time, with a single user (me) the whole time. This instrument has about 20K injections, has been regularly maintained, well treated, and stunningly reliable until last week.

The issue: Excessive but quite regular pressure fluctuations (~85 psi with a system pressure ~3K psi) when mixing gradients of methanol and water. There are no excessive fluctuations when running from a single reservoir or between two reservoirs containing like solvents. The issue ONLY happens when mixing dissimilar solvents, is worst with MeOH / water...and the piston stroke is at 100uL. It doesn't happen when the piston stroke is shorter (at a 50ul stroke, the ripple is 15-20 psi with EVERYTHING else the same).

What I've done / tested:

The first thing I did was change the filter stones on the ends of my solvent lines, thinking one or more may be slightly clogged. No improvement.

Then I tested the flow rate of the pump with methanol at ~1000psi: 1.00 mL/min set yielded 0.999 mL/min.

A static pressure test passes (accumulator decay rate = 344.3 psi/min, primary decay rate = 440.2 psi/min). A second check yielded a second passing result. No fluke there.

So, if I had to guess, I'd say that the pump, seals, check valves, and fittings are likely OK.

Next suspects: The degasser and GPV. The degasser DID have one leaky chamber which was repaired...but this had no effect on the problem (though it pumps down much more rapidly now). All others appear to be fine, and the degasser pulls down to 0.3 psia pretty quickly.

The GPV has been tested 3X and passes well within specification (error = 1.1%; specification = 3%). The line from the GPV to the inlet check valve on the primary has been replaced just as a precaution.

Just for laughs, I swapped in a priming valve from a (much) newer instrument that's running fine. Of course there was no change there...would've seen in on the static pressure test.

Our service engineer is losing his mind. He's swapped in check valves, pistons, seals, and a new primary torpedo (!!) despite passing results on tests, so those are certainly fine.

What have we missed?

(I'm kinda wondering whether the degasser is still funky, though I see no leaks and it's acting OK).

[lynzjm]

We have the same problem with our Agilent 1100 system with pressure fluctuations, our degasser was deemed "dead" but we have no budget for a new one. Having spoken to an Agilent engineer we have managed to bypass the degasser and our lines are plumbed in direct - this has made a huge difference to our pressure issues. If your degasser is playing up try priming at a high flow rate and see if you are getting a lot of air to waste (higher flow often dragged in more air for our system), also switching between different lines seemed to pull more air in too.

I assume you are filtering/sonicating your mobile phases to degas also or are you just using the filter stones as that might help if you dont.

If you suspect your solvent mixing issues, try running 2 lines, one with MeOH and one with water, and then run a single line with 50/50 MeOH/water. If you're getting more pressure fluctuations with the separate solvents on two lines than with the combined its not your solvents but your pump.

Not sure if any of this is much use but I feel your pain we have had a horrendous time trying to figure out what was wrong with ours!

lynsay

[DR]

You might also make sure you have *something* in all 4 lines. This may be easiest to do by dropping 2 in MeOH and 2 in water and doing a longer wet prime at 25% ea. Then check output from each in turn. If this doed not help, I'd shut the degasser off and He sparge my eluent pending a full degasser replacement (assuming this last step improves things).

[juddc]

If you suspect your solvent mixing issues, try running 2 lines, one with MeOH and one with water, and then run a single line with 50/50 MeOH/water. If you're getting more pressure fluctuations with the separate solvents on two lines than with the combined its not your solvents but your pump.

lynsay

Did exactly this and found that premixing improved things, which is consistent with a degasser issue, not the pump. The pump itself is fine, I believe. My theory is that when the MeOH & water mix & warm slightly prior to entering the pump, the dissolved gas comes out of solution, creating a bubble that causes the issue.

@ DR - All four lines are (and have been) fully "wet". I did as you suggested twice to no avail. I wish I still had a tank of helium!

Quick point: Unless I am dissolving a solid material (buffer salts or IP agents) in my MP, I generally don't filter prior to use. I use LC/MS grade solvents that are 0.2uM filtered and our DI water is 0.2uM filtered at the point of delivery. I use small, clear MP reservoirs (1L generally, sometimes smaller) with vented screw top caps and rinse / change MP frequently, so despite the fact that I rarely filter, I very rarely have any pump related issues. Before this, the pump has had a grand total of one seal changed outside of a PM visit in the last 5 years. As I've said, it's been extremely reliable...til now.

Perhaps I'll pop the remaining chambers from my newer instrument onto this one to see if that does the trick.

Thanks for the comments!

[juddc]

Applying some torque to the collar nut on the accumulator seemed to fix the issue, so the face seals weren't entirely sealed. Anyhow, I now have a delta of 15-17 psi at ~2600 psi with 50/50 MeOH/H2O. Go figure. Many thanks to Waters service!