Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Slamming/banging of Non-Slam Check valve 2

Status
Not open for further replies.

Danlap

Mechanical
Sep 17, 2013
307
Dear Valve experts,

Appreciated beforehand for some insight.
I have some plants which are using Nozzle type / non-slam check valve after its pumps. Where it used to be standard swing or just spool piece.
These are high flow and low pressure application. For example one case: 8”-150# NRV for operating condition of 513 to 1135 ton/day at 0.04 bar.

Observation: during start up (of a plant) or shut down (of a pump), it was reported that the NRV(s) were banging frequently with loud noise.
Banging was not reported when it is in normal operation. However some inexperienced Plant (which on start up process) are new with this and decide to replace back the spool piece after hearing banging for 30 minutes – 1 hour. Where it’s not even had reach normal operation yet

I do believe that the decision above is just human instinct (frightened of loud noises, imagining metal Disc banging against metal Seat, etc.) not yet considering that all sequence requires (flow/pressure) built up process.
Nonetheless, may someone give me a reasonable / technical answer of:
a. Does repeated banging for Non Slam NRV for some period of time (during start up/shut down) is common?
b. Is there any calculation of when this occurrences might be expected (@what flow/pressure)?
c. What is loud and what is frequent until it reach its fatigue level?
PS: After nearly 1 hour banging several times, I found no damage whatsoever on the Sealing area (Disc and Seat)

Thank you beforehand,
Regards,
MR


Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments
 
Replies continue below

Recommended for you


The general answer is that non-slam nozzle check valves, dimensioned and placed correctly should not slam or bang at all.

There is a number of things, however, that ccould influence and give slamming noises, generally for all check valves, not so often or likely for correct soft-closing valves, but possible.

1. Lay-out of pipelines and placing of valve: Air pockets in pipeline 'pocketed and compressed' at start up caused by failing or missing air outlet valves , alone or in addition to too high/long rise after the valve, unbalanced (very small) pumping amount compared to valvesize, uneven/skew amount of intake by parallell pumps/valves or same by parallell outlet to common vassel/furthering pipeline(valves 'gulping').

2. Wrongly sized valves. Checkvalves must be dimensioned after flow, not pipeline size, if wrong, most often selected too large.

3. Soft closing valves are generally equipped with springs to close soonest possible when pressure in flow direction is approaching zero, before the flow pressure comes in a backlash in the opposite direction. (This is the whole idea!). If this is not obtained the reason is likely either as given in 1 and 2, or springs could be failing, are under dimensioned, or damaged /blocked.

The exact type of valve and exact flow data (more detailed with at least max, min and normal - measured over the pump , not theoretically) and layoutof pipeline must be given before further comments.

 
The only thing I can imagine is that the flow is starting and stopping with very low flow so that the soft closure bit isn't active. Otherwise some sort of pulsing flow is being created.

Sounds like they've solved the symptom not the cause.

Some idea of vendor, size and flow would be useful.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Is this happening on a parallel set of pumps, and is the pump Q-H curve some what flat?

Is there individual flow control capability on each pump? We need to see the overall process controls at this pump set. And if there is a FIC, is it working well (tuned okay)?

There isnt enough information to work on at the moment.
 
Hi gerhardl, Littleinch, georgeverghese,

Apology for the late reply, and thank you for the constructive response.

Sorry, for this particular case I have to revise my statement. My fault for unclear information.

This is an axial (non-slam) Check valve 18"-150#, installed vertically after an air instrument compressor.
The NRV pressure drop curve attached.Model TKZ-Y Mokveld
NRV_pressure_drop_curve_aumenb.jpg



The compressor output PI data
Density :1.71 kg/m3
Minimum flow rate :513 ton/day
Normal operating flow rate :798 ton/day
Maximum flow rate :1135 ton/day
Operating pressure :0.85 bar
Temperature :100 deg C

The incident happens for during the first 3 days of start up. Where compressor was running on reduced air flow at 180 ton/day. Try to increase it up to 250 ton/day for some period of time also still flapping
Btw, we also experience same thing years ago with Sing Check valve
Therefor (maybe it sound stupid knowing the fact is there[bigsmile]), if I may paraphrase my question:
a. Being in start up mode (180 ton/day) vs minimum flow design of NRV for stable operation (513 ton/day). Is this the main reason why this valve unstable (not really open and close) thus flapping?
b. Not really able to meet the minimum flow describe during start up, should I choose a wider flow range (say 180 ton/day to 1135 ton/day)? Not sure whether this is possible
c. gerhardl, the layout more or less as below total distance from compressor to NRV is 20 meter (horizontal, 1 elbow, vertical).Not really visible to add something in between, what should be done?
Layout_mh0hvh.jpg



Thanks again for the support and information,
Regards,
MR

Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments
 
Muktiadi,

thanks for extra info.

I'm having difficulty fully understanding your system, but this looks more like a blower than a "compressor".

It's pretty clear to me that the valve is inducing flow instability in your system.

The flow rate compared to pressure is very high and given the graph above, what would seem to be happening is that the u/s air supply needs to build up some differential pressure to open the NRV, it opens, the air rushes d/s and then there is insufficient flow to generate the DP the valve is requiring and so it closes, possibly fairly rapidly if it hasn't opened very far. These valves will have been designed to close in a certain time - what was that time?

In short for startup you either need to manually hold open the valve - if possible - or have a smaller one on a bypass line or simply a bypass line fully open. What is the downside of getting reverse flow?

Operating any item of equipment below its minimum range will lead to difficulties which even the manufacturer can't solve and normally these valves are high quality and well designed.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 

I have done a quick calculation, taking none responsibillity for calculation and advice.

At 180 ton/day (24h ?!) you have a gas speed of roughly 7,5 m/s, at 513 ton/day 21 m/s and at 1.135 ton/day 48 m/s. Roughly (depending on a lot of things) gas speeds at 20- 50 m/s are seen as a 'normal' handling range.

The 7,5 m/s is far lower than the suppliers required lowest level for normal operation. The column of fluid before the valve will act as a pressure vessel. The valve requiers an overpressure at this low flow to open, and will let out a small amount each time the pressure is high enough and then close again abruptly. This could probably be verified with running a continous pressure measurement before the valve.

It could, as a far possibillity, influence the behavior to move the valve to closer to the pressure giving source, to minimize the 'pressure vessel', but I would not recommend this as both probably costly and with unsure result.

As I see it, you have two possible solutions.

One solution would be either to find a valvetype with a wider operating range, or check if a smaller 16 inch valve is acceptable (abrupt reducing flanges fore and after the valve as a suggestion could be OK). A 16 inch valve would give a speed of about 10 to 60 m/s (check!). Combined with (if possible) a slightly higher start-up capacity this could improve the situation. Check with brand supplier and one competitor for( wider) operating range.

Extra checkpoint: Check if there is any sticking (sealings?) and that valve is suited for 100 deg C and operates as designed. Check actual temperature at the valve (less or above 100 deg C).

The other solution is to arrange a bypass with a smaller valve. For this you have to be able to close the exsisting main line before the 18 inch checkvalve and the new bypassline selctively. An 8 inch bypass valve would give you 39 m/s at 180 ton a day.


 
~If you have 20m of what looks like 30"(750) pipe u/s your NRV, then I calculate that if the flow d/s is essentially open at low flow, it would pulse at 3-4 times per second to achieve the required opening DP, vent hat d/s, valve closes again, pressure builds, opens, ....

does this equate with what you find - i.e. you have a "chattering" of one to three times per second as opposed to a "bang" every few seconds??

One option would be to move your NRV much closer to the outlet of the blower, but this would still chatter, just at a faster rate but may not have time to actually close before it needs to open again.

too fraught and I think one of the alternatives mentioned needs to be looked at first.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Gerhard, LittleInch,

Thanks again, really appreciate for the effort, calculation and help

What is the downside of getting reverse flow? Compressor (or blower as LittleInch said) is not adequately protected against backflow of acid gas from the furnace. The "Initiating event" of is the trip of this blower.
Some historical record 1990-1998 used Double Disc check valve 30" and occasionally flapping, shimmering the pipe and cause nuisance trip of the compressor.
1998 to 2015 operated without any valve internal (act as spool). And based on the recommendation (IPF and PFA study) should go to smaller NRV (18"), which now facing the problem.
Have to say not really addressed the start up flow though.

At 180 ton/day (24h ?!) Yes it is for at least 1 week. It feed up other process downstream (sequential start up), and will achieve 500 ton/day more or less within a week.

you have a "chattering" of one to three times per second as opposed to a "bang" every few seconds??
Not really sure the frequency, but yes it sounds like either 1 or several time per second. The 30" pipe was shimmering also.
LittleInch, sorry I am not a Process person and definitely not a student anymore. But how do you come with calculation of pulse 3-4 times per second? Just out of curiosity.

On top of that, I will propose the action to reduce the valve size (to increase flow) and or adding bypass.

Appreciated a lot,
Have a good weekend,
Regards,
MR

Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments
 
Muktiadi,

Fairly basic. The volume of the pipe u/s the check valve is approx. 8m3. The amount of differential air pressure needed to open the check valve is 0.02 bar. for that volume you need about 0.2m3 of air to flow in to that volume to open the valve. At 180 tons/day, the supply is 2 kg/sec or about 1.1 m3/sec so takes about 0.2 seconds to pressurise the pipe and the valve opens. Allowing a bit of time for that gas to flow downstream before the valve closes is maybe the same time so one cycle takes 0.3 - 0.4 seconds. All a bit basic, but the NRV is basically operating like a pressure relief very close to its set point and "popping" on a regular basis which I estimate is 2 to 3 times per second at your lower flow rate.

You might need a modified check valve which holds open on an electrical signal and then if the blower trips allows it to close??

Ask mokveld - they will probably have seen this before and are a very competent supplier in my experience, if a little expensive.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
When this air compressor is operating at low flow, is some excess air being vented off via the DN400 control valve which is upstream of this check valve? Is this a pressure control valve (or is this a min flow FCV)?

Is each of these compressors operating on individual pressure control via dedicated PCV on each compressor?

This may be the reason for this check valve opening / closing on each compressor run - minor changes in actual pressure upstream of the check valve ( due to action of the PIC- PCV). At higher flow, this DN400 control valve closes, and there is no more pressure instability upstream of this check valve?

If so, try to stabilise the operation of this PIC or FIC loop.
 
Or transfer the FIC or PIC loop to be downstream of the check valve (which is the same as moving the check valve upstream of the take off to this blow off control valve).
 
With this check valve in the vertical ( am assuming flow is vertically up right now at this check valve), turndown will be further reduced. Shouldnt this check valve be on a horizontal run?
 
There is another check valve at the inlet to the furnace - reactor? What type is it and how is it oriented?
 
I agree with georgeverghese, I would investigate re-orienting the valve into a horizontal run. The valve would still be only partially open (which is not ideal), but this would eliminate the banging noise as the valve disc would no longer be falling back into its seat under its own weight.

One alternate option (but less preferable solution) would be to discuss with Mokveld the potential for removing the spring from the valve. My feeling is that your minimum flow is too low for this to make a difference but they will have all the calculation sheets and will be able to make the correct decision. (Removing the spring is never a great idea - because it dampens disc bounce-back as well as aiding the dynamic characteristics of the valve, but it may solve your problem here).
 
Its just a suspicion I have that the check valve stops banging when this blowoff DN400 PCV (or FCV??) closes as the compressor loads up during startup, and not when the net forward flow through the check valve is in excess of 21 375 kg/hr (513t/d) - Muktiadi hasnt disclosed when this happens - that would indicate this is a process controls issue and not a turndown problem with the check valve. Obviously the check valve vertical orientation has made this worse.
 
georgeverghese, Raspberry,

Sorry for the late reply, due to my field load. And thanks for the additional insights.
Agree with you at some extend that vertical mounting is not really ideal for this case. However that was the only option, otherwise we have to re-route our spaghetti pipeline.
And the manufacturer has done the calculation that 18" size is not really ideal sizing however do-able shall our minimum flow indeed 513 ton/day. 30" as the size of the pipeline definitely would not work.

DN400 FCV was supplied from the blower manufacturer. It is almost always in close position (for the last 20 years) and also during startup. Not sure whether the commissioning taken place for the last 20 years, however yes shall it is leaking (crevice between the Disc and Seat), the NRV will be 'pushed' by less than 180 ton/day during start up. This FCV is directly connected to a high stack.

No there is no other Check valve at the inlet of our system (furnace/reactor).

Hope I inform you sufficiently.
regards,
MR

Greenfield and Brownfield have one thing in common; Valve(s) is deemed to "run to fail" earlier shall compared to other equipments
 
your best bet I think is some sort of smaller valve on a bypass line used for start-up or somehow holding it open during start up.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
There is a lot more mileage you'd get out of this check valve if you could move it onto to a horizontal pipe. On the other hand, I find this velocity of approx 20m/sec for min stable operation somewhat difficult to accept.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor