Arrrgh: Waters 600 - pressure variation problems

[dkreller]

Hi,

I have a Waters 600 pump, and I have been having very difficult problems with pressure. The system backpressure is so variable it prevents me from getting experiments conducted. I have lost several days. My mood goes up and down based on the wierd way the system backpressure oscillates. I have a flow restrictor to raise the system pressure at the pumps because my columns, which I pack myself, have relatively large particles (180-250 microns) and don't create much back pressure.

The pressure oscillates very periodically, and corresponds to a cycle of the reciprocating pumps. I anticipate what you're thinking - 'It's obviously a check valve problem, you loser'. I hear you, but please believe that we have certainly scrutinized the check valves thoroughly. We have taken both pump heads off and cleaned (sonication in water, then MeOH) the check valves.

We have performed the modified ramp test repeatedly and most of the time the results have indicated that the check valves are fine. However the results of a recent ramp test suggested that the left pump outlet check valve was not functioning properly. So we exchanged the positions of the left valves, i.e. put the check valve that was in the outlet position in the inlet position. I then re-ran the ramp test wondering if the results would then suggest that the left inlet check valve was the problem. But then the test results suggested that both inlet and outlet check valves were fine.

Sometimes when I am just pumping DI through the system (regardless of flow rate in the range 0.5-5.0 mL/min) the system pressure is well behaved for several minutes, even a couple of hours and then the pressure variations slowly develop and get worse and worse. In the past when the system was working well, the pressure would be about 700 psi with maybe +/- 5% fluctuations at a flow rate of 0.5 mL/min. However, when this problem is at its worst, the pressure will fluctate all the way from 700 down to 0 psi. The big drops in pressure correspond to when the L pump is on the compression cycle.

So, there definitely is something wrong with the left side of the pump, but based on what I've described, you'll understand that I can't conclude that it is a check valve. If it's a check valve perhaps one of the check valves works intermittently. Uwe, have you ever seen that? A check valve that works for a while, and then starts misbehaving? The check valves are the type with the ruby ball, i.e. the one-piece cylindrical units that have the stainless jacket and the flow direction arrow on the outside.

In one of my conversations with Waters tech support they suggested that there may be a leak in one of the lines. If there is a compromised line somewhere it is letting gas in. The pump behaves as though there is and air bubble in it, and plus I have never noticed a leak anywhere. However, if we change to a different reservoir the problems persist. This other observation tells us that if a line it bad it's a line it's not associated with a specific reservoir. Could it be a short bit of line between the pump(s) and some mixing unit?

Any thoughts?

Thanks in advance for your expert consideration,

Dave

[gstaepels]

Hello Dave,

I have a Waters 600 pump for many years and it still is working fine. The problem that you describe is also known to me. Especially, when you work with buffers, it can happen that pressure fluctuations occur. The first thing I try is to repalce the pump seals, If that does not work, I try to clean the check valves. I sonicate them, and a lot of times, this solves the problem. However, sometimes it does not! I have always a spare set of check valves available, and replacing the check valves with new ones often solved the problem.
I hope this helps

[scottythree]

I will not say that I am an expert on these but I do have a few years of experience with these pumps.

Are you degassing / sparging (forgive me but I have to mention it)?

I learned to ensure all 4 lines were primed before running (even if only running isocratic).

When/if you start seeing the pressure variation look at the mixing valve inlet lines to see if there is a pocket of air. It may be wise to change the little piece of connecting tubing between the inlet lines and the mixing valve, especially if you see any buffer salt crystals around them.

A pump seal leak will cause these issues as Gilbert has suggested.

[dkreller]

Hi,

Thanks for your input.

We certainly have thought that a pump seal could be the issue. However I was told by Waters tech support that a faulty pump seal would give erractic, rather than periodic pressure variations. The pressure fluctuations definitely are periodic and tied to the pump's cycles. I have a couple of new pump seals on hand - we could start by swapping out the seal on the left pump.

It's an interesting suggestion to prime each of the lines, even if it's isocratic work, which it is. We could try that.

I will also visually inspect the mixing valve inlet lines for air bubbles/pockets.
Dave

[Chromatographer2010]

You can definitely have intermittant failures with the check valves especially if there is a defect or particulate on the ball. Sonication doesn't always get it off.

If the plunger seals were leaking then the pressure would not rise evenly when you are doing the ramp test.

Though from what your describing it sounds like a clog or air leak on the inlet side. Usually its a problem with the filters in the solvent bottles but can also be in the proportioning valve, the line from there to the priming valve below the pump heads or the lines leading from the priming valve up to the inlet check valves. The air leaks can let air in without letting liquid out. Check all of the fittings on the inlet side for tightness.

You can also check siphon flow at the inlet to the pump heads with each of the mobile phase lines. It should be very free flowing.

What mobile phases are you running? Is there the posibility of precipitation?

And as scotty metioned are you degassing?

[Bruce Hamilton]

I'd always suspect the check valves, and would use a more rigorous cleaning regime. Ruby ball/sapphire seat HPLC check valves are a device invented by Satan to preoccupy chromatographers. Always try to provide a good hydraulic head by having the solvent reservoirs high above the pump.

I prefer warm CHCl3/MeOH or DCM/MeOH for cleaning and ultrasonication of check valves, especially if acetonitrile is being used, as the deposit that causes sticking is difficult to remove.

If your flow rate is only 0.5 ml/min, and you have a high organic solvent content, also be aware of the pump head temperature. As the head warms, so does the solvent, and dissolved gases undissolve. Ensuring you have good degassing and cool pump heads makes life more pleasant.

Incidently, I've found that old plunger seals can cause problems that mimic sticking check valves, as can worn pistons, so don't exclude those possibilities, but they are very unlikely.

Bruce Hamilton

[vofelyp]

Hi Dave

1, Check the outlet valves too. The left head pressure drop can be cause the left inlet valve or the right outlet valve.
2, Remove the inlet filters may be its clogged.
3, Check the plungers, old scratched plungers quickly destroy the new seal. Use black seal if you do not replace the plunger.

[dkreller]

Hi,

Thanks for your input. A couple of you inquired about degassing - I have He sparging.

With regard to the reservoir filters, how often to you find you have to replace them? Can they be cleaned? Sonication?


Thanks,

Dave

[juddc]

Sometimes when I am just pumping DI through the system (regardless of flow rate in the range 0.5-5.0 mL/min) the system pressure is well behaved for several minutes, even a couple of hours and then the pressure variations slowly develop and get worse and worse. In the past when the system was working well, the pressure would be about 700 psi with maybe +/- 5% fluctuations at a flow rate of 0.5 mL/min. However, when this problem is at its worst, the pressure will fluctate all the way from 700 down to 0 psi. The big drops in pressure correspond to when the L pump is on the compression cycle.

Hi Dave - In my most humble opinion, this is a classic symptom of a clogged reservoir filter. If you have them in stock, replace the filters (or remove them briefly if you're sure your MP is clean) and test. I replace filters at PM time (annually) or if my system demonstrates what you've just described, which is rare. I wouldn't bother cleaning them, though I suppose you could - a quick shot of 6N HNO3 might do it (Hey, if you can passivate your tubing with it...). I've never had too much luck sonicating them, but I've heard that folks do.

Good luck!

[dkreller]

Hi,

Thanks everyone for brainstorming with me about this issue. Here's a quick overview/update:

I removed the reservoir/filter stone completely from the reservoir from which I was pumping DI. The pressure fluctuation problems continued. Hence the problems couldn't be blamed on a filter/stone. But, we had four new ones in storage so I changed them all out. Some of them do sit in NaCl-containing solutions and were looking a bit old, so they were due for being swapped out anyway.

My reservoir bottles were sitting too low on the bench, at the same height as the pump. Waters helped me understand that it is not good to make the pump work against gravity, especially with volatile solvents. Again, I work with aqueous mobile phases, so it's presumably not a problem here, but I am working to get the reservoirs elevated, just for good practice. But, back to the issues at hand, I found that when I elevated the working reservoir the pressure fluctuation still continued..

We ordered in a couple of new check valves. I ran some more modified ramp tests and determined that the problem was in the right pump head. I changed both valves on that side of the pump and presto - things were back to normal.

I admit that in my original post I suggested that the check valves were not a clear source of the problem. The pressure ramp tests were giving inconsistent results. I am still quite certain that the check valve problems were somewhat intermittent.

Although this problem set back my experimental schedule by a few weeks, this problem has been good in the sense that it forced me to learn quite a bit more about my system. I appreciate the valuable and thought provoking input on this discussion thread.

My fingers are crossed that the problems do not re-surface.

-Dave

[tom jupille]

Dave, thanks for posting the outcome (helps us *all* learn!)

[vanwa71]

Thank you everyone for all this information. I have been having very similar issues with my Waters 600 pump. I have replaced check valves, not with new ones, but with barely used ones, I've replaced piston seals, and we just bought a new gradient proportioning valve (GPV) because air was coming into the system at the GPV when run at higher flow rates for priming (10 ml/min). Unfortunately, the new GPV did not resolve the air leak issue, and I'm afraid it was just leaking connections (I know, very unfortunate mishap if it turns out to be true). When running at typical flow rates (e.g. 1 ml/min) no air leaks are visible, but the system pressure fluctuates from about 1700 psi to 600 psi. I have noticed that the pressure always drops when the right piston is moving out toward me. I replaced the piston seals on both pump heads, sonicated check valves in methanol for at least 20 minutes, and rinsed them out, to no avail. I was very concerned that the check valves were damaged or had particulates because some degraded dichloromethane was used in the system inadvertently, and it started reacting with the seals, and even the metal solvent inlet filter/frit. The metal actually turned yellow and started releasing a gas that was formed from the reaction (i.e., the mobile phase was degassed, but gas continued to be formed and rose to the top of the solvent bottle. This corrosive degraded CH2Cl2 quickly destroyed the piston seals, so I replaced them. However, I'm worried that the check valves had permanent damage. I thus tried used barely used check valves, which didnt help. Tomorrow I'm going to try brand new check valves, and pray that it fixes everything. I've since disposed of that bad dichloromethane. This should serve as a reminder to everyone. Even simple solvents can go bad, and apparently, CH2Cl2 can degrade to HCL if given enough time. The poor column that I was trying to clean with that evil solvent was clearly damaged as well. I hope someone can comment. I will post an update regarding the brand new check valves. I will also replace all the clear tubing connectors on the GPV tomorrow, as that's another source of problems. Lesson learned: poorly maintained system exposed to degraded solvents = a plethora of issues, and lots of expense. Please pray that the tables turn tomorrow.

[tom jupille]

the system pressure fluctuates from about 1700 psi to 600 psi. I have noticed that the pressure always drops when the right piston is moving out toward me.

If the pressure always drops by that much, then you have major problems on the right side: a leaking fitting, a bad check valve, a bad seal, or a worn/scratched piston. That's not so say there may not be other problems, but your major issue is one of those four.

[juddc]

While it is certainly possible that your check valves are damaged, I've found that sonicating in straight MeOH doesn't often help much. Sonicating in warm (40-50C) 3:1 water:MeOH (or just plain warm water or even 6N HNO3, depending upon what's been through the pump) often does better for me...
Good luck!

CJ

[vanwa71]

Thanks for the prompt reply! I tried brand new check valves today, and after a few minutes the pressure seemed to stabilize. Also, I replaced all the connector tubing going into the Gradient Proportioning Valve. The new tubing was a much tighter fit than the old tubing. I have the solvent (ACN) recycling overnight at 1 ml/min to see if the pressure is maintained. I will try a higher flow rate with the ACN going to the waste line tomorrow. I don't expect to see air coming into the GPV this time. There is still a little fluctuation in the pressure, but its less than 100 psi. A world of difference. This system was installed in 1996 btw, but hasn't been used much until now.