Valve PMs for packing leaks?

[YungPlantEng]

Hi guys, I’m an engineer working in the ammonia industry.

I was tasked to start a valve maintenance program, which pretty much boils down to putting these valves on our walk downs and tightening the flanges and packing as they loosen over time. The idea my manager has is that tightening before a leak occurs would reduce potential safety incidents as well as help the valve packing lifetime.

While I haven’t found any info that supports tightening packing before a leak path is created extends valve lifetime it makes a bit of sense to me. However if anyone has information on just how much it does do I would appreciate it.

The other side of it is that I’m just not sure what this procedure should look like. My manager mentioned something like tightening 10% more each time before it reaches its highest allowable torque but I can’t find any best practices for that. I’m also a little worried about the idea of our technicians spending a lot more time tightening packing on in-service equipment, since a high pressure leak exposure incident seems a lot more likely at that point.

Lastly I’m not the most electrically inclined and don’t know what maintenance practices for actuated valves would look like and how they might differ,

I know this is very general but does anyone have any guidelines they could point me to for this? Much appreciated!

 

[Danlap]

Good day,

Not really clear, since it is not mentioned in your message, I assume you’re discussing about gland packing. There is no flange on this part.
Flange and packing? (btw in Dutch we call gasket also as pakking) if this is about body to body/bonnet gasket, then better to leave it as is and replaced gasket whenever it is possible. Leaking on this area most likely due to gasket already compromised, overtightening will not help.

Gland leakage:
Mainly caused by 3 things:
-Vibration on spindle --> on my place this is caused by small opening globe valve>flow induced vibration>vibration transferred to spindle>leak
Mitigation? Do not operate manual/handwheel globe valve below 25% opening. If small opening is required, suggest to use engineered globe/control valve. Kinetic energy passing through the plug suggested to be below 480 KPa. this number if from Energy Institute 'Guidelines for the Avoidance of vibration ...'

-Relaxation due to high temperature --> graphite gasket consist of graphite (93-98%), some PTFE, some Zinc, etc. above 300-350oC practically some Zinc and PTFE might be gone, hence there will be volume loss
Mitigation? Within 3x24 hours exposure high temperature after start up, suggest to tighten gland as per OEM

-[assuming spindle is not yet corroded/pitting corrosion/etc.] relaxation due to frequent operation
Background: friction between spindle and gland, over the time will relax the gland lead to small leakage (on a scale of ppm level) up to visible. It is open for counter argument, however there is general consensus which indicate 400 or more cycles, valve will start to show leakage (from zero becoming detectable ppm or by means of FLIR camera). I personally use this approach 400 cycles/ conservatively 4-6 years to re-tighten gland.
(please differentiate this with for example TUV certified manually operated special valve or control valve which can sustain 10,000-20,000 cycles without leakage, as they are more likely use some sort of engineered or PTFE gasket and for relatively ambient to moderate temperature).
A bit different with your manage approach of using torque value, we apply 2xquarter turn on each gland bolts every 4 years. Reason: it is handier/no torque moment required, assuming lifetime of valve is 25 years then this method can be applied 6x without needing gland replacement, in theory of course.
One of the downside applying moment value, in your case 110% of its original torque, is that once stud/bolts are corroded it still may not turned hence even though the moment torque is already clicking but the gland is not yet tightened.. Your manager proposal might be working for smaller population valves with good strategy though, but our refinery is about more than 700,000+ gate/globe valves.

Hope that helps, I believe there are many other best practice/reference by other valve practitioners in this forum.

Kind regards,
Danlap

 

[YungPlantEng]

Appreciate the response Danlap!

I think the current plan is to take a system, replace all the flange gaskets and torque them to the minimum OEM specification. Then perform a yearly check where the torque is tightened an extra 10% until we reach the maximum specification, in which case we replace the gasket.

Valves were the initial idea of this initiative, which I think could be managed similarly by tightening on a yearly schedule. Our plant is relatively overburdened with flange and valve leaks, but I am wary since while we might only have around 1500 valves, we have a multitude of flanges and I might underestimate the safety/reliability savings in keeping these gaskets and glands tight vs the extra work maintenance would be doing in PM's.

In terms of MTBF we generally experience a 6-9 year lifetime out of our valves. It's difficult to quantify precisely since failures are difficult to categorize.

As said previously, any research documents or books on the topic as well as personal experience are appreciated. I think this approach will end up too rigorous without much gain (would essentially need a torque label on every flange for it to be done appropriately). Would still be nice to try with our covered process though.

 

[Danlap]

Maybe its just me, but could you please advise which flange are we talking here? Maybe some pictures (without brand) might help.
Reason is the word 'flange' and 'OEM' was repeated once more.

I think you should take one step back, what is the exact root cause of the leakage. Prior start applying extra torque on the bolts.

In general, leakage paths on valve consist of: (valve) flange to pipe connection; body to bonnet gasket and gland packing/gasket.
-Flange to pipe connection --> this is not calculated or defined by the OEM. Usually each Plant have their own flange protocol with different torque values, as standard basis of calculation usually B7/B7M/B16 bolts, with Carbon steel and A182 family flange, along with Spiral Wound Gasket.
And for some special cases, EndUser might used different bolts and gasket resulting different calculation. Hence, I don’t think the word OEM suit this category.
In the event when flange-to-pipe start to leak, then directly replaced gasket if possible, retighten will not fix the issue or even worsen. Only some RTJ connection might be mitigated by re-tightened.
And also why is it leaking: gasket burst? Corroded? Or bolt relaxation? --> choose other gasket or bolts

-Body to bonnet connection --> this is calculated by OEM. However this is static connection, therefore usually no retightening is required.
Why is it leaking: gasket burst/corroded? Bolt relaxation  chose other gasket or bolts, bring along the manufacturer to investigate the issue. If bolt relaxation is the issue then adding Belleville spring washer might help a bit.

-Gland packing --> my previous message
Which leakage path do you mean on this case?
please not that the material quality of gasket/bolts and flange material might eventually contribute to the leakage as well.

In depth leakage analysis is something which is custom/case sensitive. Contributing factors: medium inside the valve, dimension of gasket, material used, bolts used, temperature exposure, frequent of operation (if one able to count in the span of years valve being used), vibration, etc.
Due to broad spectrum, at least for refinery and chemical plants, I believe there is not yet any comprehensive public documents (free/paid) which suit your need..

For more general documents
-Guidelines for the integrity management of valves for the upstream and downstream industries by Energy Institute --> pay 250 poundsterling
-OREDA (Offshore Reliability Data). There is probability of failure and also mtbf for leakage in this book, however mainly for upstream product. Fair to say that offshore environment is relatively easier for observation compared to downstream refinery or chemical. This book is written mainly driven by Alpha Piper 1988 incident.

For more specific best practice especially related to repeated leak occurrences on specific Plants, we usually have our own investigation or summary. Most of them are not available for public view.


Regards,
Danlap

 

[Andrew ONeill]

OEM's might a place to ask the same question.

I assume you are talking about gland packing on valve stems, and the flange you mention is what I call a gland follower.

There can be issues with over tightening glands, basically causing too much friction on the shaft. I general this would only be a problem for actuated valves. The other thing to consider is that you might quickly run out of adjustment on the glands if they are constantly tightened.

For problem applications (typically vibration) the only other thing I have seen is the installation of spring mechanisms between the flange and the gland. The spring is normally a series of Belleville style washers. So this could be something to look at as well, if you are about to go and replace them all anyway.

Andrew O'Neill
Specialist Mechanical Engineer
Rio Tinto
Australia

 

[YungPlantEng]

The flange I mention is a typical piping flange with a gasket. Apologies about the mix-up. These would be two separate initiatives to mitigate leaks in the process