Instrument maintenance and repair

[skunked_once]

This post is strictly in fun so I hope it brings a few smiles.

There are two types of chromatographers; those who perform routine maintenance and keep their instruments in tip-top shape and those who run their instruments until they stop working and scramble to get them going again.

During routine maintenance I replaced....

This appeared in a separate topic and is an example of the former group. This violates the maintenance philosophy of the second group "If it works, don't mess with it". John Dolan will cringe at this

My repair experiences fall into the following categories:

1-Instrument failure, take apart instrument, find problem, replace or repair part, reassemble instrument, instrument works. Problem solved.

2-Instrument problem, take apart instrument, can't find problem, reassemble instrument, instrument works. Don't know why but problem solved.

3-Instrument failure, take instrument apart, can't find problem, reassemble instrument, instrument still doesn't work. Blame previous user.

4-Instrument failure. Find excuse to not repair instrument (obsolete, no longer supported, tired of fixing too many times already). Beg for new instrument.

Feel free to add your experiences.

[juddc]

My HPLC experience pretty much goes like this:

Buy really good instrument and service contract.

Make sure that everything is kept clean, religiously follow filtration and cleaning procedures as appropriate & run instrument daily.

Schedule and perform annual PM's as appropriate. Continue to run instrument daily.

IF anything breaks (rare), test and diagnose prior to disassembly if possible, then fix what broke and have a look at everything else really carefully while in there. If it's anything really fun that I can't deal with (really rare), call the manufacturer and get an engineer on site within 24 hours and make them repair as needed.

Repeat as needed for 15-20 years.

Retire instrument in perfect running condition due to obsolescence or lack of parts availabilty.

It might not be the cheapest way to do it (contracts = $$), but instrument downtime is nearly nil and almost always scheduled and THAT keeps management happy.

[skunked_once]

juddc,

You operate under the ideal situation. Unfortunately, in many settings, especially the academic world, service contracts and such attention to detail are not possible due to costs and management (or lack thereof) practices.

[juddc]

juddc,

You operate under the ideal situation.

I know it...I've never worked in academia, but I do work in a situation where downtime is a BIG no-no and deadlines exist.

I work for a large corporation now but I have worked for small start-ups where money was tight and I worked with the same philosophy then, which is basically that a system works most efficiently when well maintained. Alot of folks fail to realize that.

My costs tend to be upfront, predictable, and to a certain extent, negotiable as well, which is important when doing HPLC for money.

With the added reliability, I can make more data per unit time and essentially charge more for it, thus eliminating the added cost of completely anal upkeep...not to mention that my columns tend to last longer, too, keeping my investment there down.

[JGK]

For any Lab working in a regulated environment Juddc's route is the only way to go as it best satisfies both regulatory and client concerns.

If you can't get service contracts it might pay to send people to train with the suppliers service group in instrument repair.

[Peter Apps]

My experience when I have been working in an unregualted research environment

Obtain instrument, either buy or move into a lab that has them already.

Find out what I already knew; off the shelf hardware can't do the esoteric stuff that I do

Start adding bits and pieces, if other people use the hardware curse because no change to the instrument can affect their work.

Watch actual results for signs of deteriorating performance, trouble shoot and fix as I go along.

Pray a lot

When I've been in a regulated lab I've gone the fix it before it breaks route as long as the funds were available, sometimes management chooses bread today over jam tomorrow.

I find a huge difference between GC and HPLC in maintenance strategies - mainly because an HPLC has many more moving and wearing parts in the pump, and is more susceptible to blockages.

The other major impact on strategy, besides are you doing it for money or for fun, is whether you run big batches of samples in walk away mode, if you do you have to be sure that everyting is working before you start the batch. If you are running single smaples hands on you can detect problems from the actual results of each run, and fix before you run the next one.

Peter

[lmh]

In some academic circles the chances of getting a service contract are lower than the chances of writing a new instrument into a future grant proposal. The result is a feeling at the university that they are being efficient, somewhat marred by a heap of new-but-dead instruments languishing under the stairs at the bottom of some stairwell.

Of course academia is also notorious for high staff turnover, so there's not much point in keeping an instrument in good condition. There's a strong likelihood that the resident expert will in any case run out of grant and leave, after which no one will know how to turn the instrument on. The best way to get training, again, may be to write a new grant for a new instrument that comes with training attached. You can always save 1000 Kloptian Squints spent on an over-priced training course by buying an unnecessarily new instrument for 50000 Kloptian Squints and negotiating some training on the deal. It's a bargain, isn't it?

I'm deeply grateful that, while I work in academia, I work for an organisation that is more enlightened about service contracts, and also about training.

[lmh]

Oh, and just for the record, my philosophy: hplcs are like toilets. Keep flushing them, and they stay unblocked.

[WK]

Always make sure that when you return from holiday that the detector outlet isn't hooked into the pump leak drain line - as mine was the other day......
Colleagues with little knowledge have very little respect for instruments - or "machines" as they call them.

[DR]

My favorite (target) is the type who breaks instruments in series, usually at least 3, before asking a question (like "Can I pump 50% caustic through this?").

[JJG]

The one HPLC which existed at my university was used once a year. The professors hardly knew what to do with it, but would fire it up when the time of the year came to do a lab experiment to determine caffeine in a sample by HPLC. When that time of the year came, the professor would fire up the instrument, only to find a pressure maximum error. Then he would call service in. Once he found out I had worked on HPLCs at my job in an analytical lab, the next couple years he would have me take a look at it when he had the same pressure error. Luckily, it didn't take long for me to find a blackened plugged filter. I noticed the system was stored in water...for a year. I then proceeded to write detailed instructions for storing the instrument. I'm not sure that advice was followed.

As for my take on maintenance, I like to make sure my instrument is happy before every run since I usually have many samples that have a lot of prep time. The last thing I want to have to do is re-prepare those samples. This means keeping a supply of parts on hand to replace anything that may break randomly so the run can still be run that same day.

Another thing I've learned: On a Friday, don't wait until 4pm to put your samples on an instrument to run. You are guaranteed something will break and you will be trying to fix it for 3 more hours. Better to start at 1pm so your samples are running by 4pm.

Also, if you see a co-worker pushing a lot of buttons at the end of the day, and hear a lot of instrument error beeps, better leave the lab quick and head for your car, especially if you are the go-to troubleshooter.

[Bruce Hamilton]

My strategy is leave until broken, where possible and sensible. The definition of "sensible" depends on the organisation and operating environment.

If manufacturing a high value product subject to regulatory control, the cost of a service contract is easily justified and approved. However, one consequence of such a policy is that your organisation is hostage to the whims of the contract provider - who may send the first available person, not the most skilled.

I've noticed a tendency for contract service support people to be less skilled/trained here in NZ, resulting in multiple visits to fix unusual problems, with the associated downtime and requalifications.

Note that the very first items that most auditors look for in any instrument records are the training record of the service person and operator - to ensure they are qualified to perform the work.

Forests can be chopped down to fulfil the Quality crowd. The sooner the fascist playground of DQ, IQ, OQ, PV are replaced by a rational assessent ( perhaps ASTM E2500 ), the better. Sadly I expect the Quality crowd will pervert most rational guidelines.

If you operate in a price-sensitive, competitive environment, your high level managers would sooner extract their own teeth than spend money on the possibility of breakdown. Such spending adversely impacts on their bonus.

In that case, you have to surreptitiously purchase as many spares as possible as consumables, and quickly fix problems as they occur.

I think it's a very good idiea to maintain service logs in all environments, just so tips are recorded for future victims. I'm fairly comfortable with repairing instruments on an as-required basis, provided it's well documented, and the person performing the work is competent.

For example, replacing pump and injector seals ( the original thread ) can result in significant downtime if there are problems, and usually can't usually be accurately predicted. The seal life depends on mobile phases, sample preparation, and general operational procedures.

I use a mix of normal and reverse phase separations, but the normal phase solvents often have 1-50% EtOH or IPA - which are very benign to seals, and try to flush non-volatile reverse phase buffers from the system as quickly as possible after use.

I try to ensure I understand the instrument, so I can prevent surprises by preventative actions, such as good sample and mobile phase preparation, and careful flushes when switching.

Pulling an instrument apart to fixing a possible impending problem is good fun if the alternative was sitting at a desk counting screen pixels, but if there is other work you can perform, that's probably of more value to te organisation. Unskilled maintenance can a significant cost to the organisation, especially if a service person has to be called to repair the damage you caused - which will happen to eveeryone at some stage.

If a colleague is having instrument problems, the best solution ( other than suggesting suicide ), is to provide the name of the service support person at the instrument suppliers, or support contractors. I've often found that suggestions that work became my collegue's, and those that didn't were attributed to me, and I was tainted with making the problem worse. Running out the door without offering something tangible can be misconstrued to management.

Please keep having fun,

Bruce Hamilton

[tom jupille]

http://www.murphys-laws.com/murphy/murp ... ology.html

[Nofertari]

Good evening, Everybody!

Being a new forum-visitor, im not sure whether it is the best topic to write in, but i hope u can give some help.
I have a HyperClone C18 ODS (150*3,2*3) column, and it has worked so far very well. Some days ago it looked to have contaminants; i made a blackflow with the following protocol: ACN-water 5/95, then THF, and after ACN/water 95/5. After this it looked to be ok and work. I made 2 run (ACN / water + AA; from ~20%-100%; i had been using this column such system before). Now i changed ACN to methanol (20% metanol), and the pressure flied into the sky. I tried to wash with 100% ACN, then 100% methanol and blackflow, too, it was all right, but next time with the system mentioned previously pressure has gone up again.
This is the problem. Any advice...?

[skunked_once]

Nofertari,

Welcome to the forum where all problems are solved and all answers are given to the best of our ability Your problem would be best posted in the "Liquid Chromatography" section but I will provide some input here. The high pressure may be due to the fact that aqueous mixtures with methanol have much higher viscosity than aqueous mixtures of acetonitrile. Or it could be some other factor such as column contamination. Please respost your question in the liquid chromatography section with more information on your chromatography conditions (flow rate, exact mobile phase composition, sample composition, actual pressures with the different mobile phases).