Noise Mitigation in piping - Downstream of Natural Gas Flow Control Valve 2

[grhead]

We have installed a flow control skid, and have measured noise at the downstream piping in excess of 100 DbA. We will likely be designing a similar installation in the future, and am looking for some input on how to mitigate noise in a gas piping system.

- See attached sketch.
- The skid is located inside a large open building, and 85 dBA max. noise limit is desirable for plant operations that may be working in the area.
- We've added acoustic insulation and cladding to the piping resulting in a reduction of about 5 Db.
- The header skid feeds fuel gas to several burners, the header will modulate various flows and pressures, but we design theoretical noise for a worst case max. flow condition.
- The Flow Control valve is sized by our valve supplier for a max. theoretical noise of 85 DbA at 3' from the valve. However, I believe what we are seeing is noise originating from the flow control valve propagating through the downstream piping, which results in a much higher value at the piping.

Some thoughts I've had:

1) Increasing the straight run of pipe after the pressure reducing skid before the 16" elbow. I'm unsure on how to come up with an estimate of length of this run.
2) Designing for a more gradual pipe bend/transition. Currently we have a 16" standard elbow, perhaps a series of 30 degree or 45 degree elbows?
3) Installing a silencer or conditioning plate downstream of the skid, I'd be curious if anyone has experience with such a device.
4) A noise "trap" (see attached image clip). I came across this in some literature from an ultrasonic meter supplier to mitigate noise upstream of the meter. I was pondering putting in a straight run of pipe with a dead-end, and then install a "T" takeoff which would go to the gas supply. I would likely acoustically insulate the straight run of pipe around the "trap"

Thanks in advance.

 

[LittleInch]

Well from my calc you're doing 30m/sec in that 16" pipe so it's not surprising it's making a loud noise.... A bigger pipe maybe? - 20" or 24"?

You can also get low noise trims for most control valves which help, but I would go for a bigger pipe to get that velocity down and possibly a bigger control valve to reduce peak velocity in the valve.

Depends on your range of flows, but maybe two different sized valves? Or two valves in parallel with one open for lower floors then both in parallel for larger flows. Sometimes one valve has difficulty at both ends if your flow range is too large. What is min flow?

What type of control valve?

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

 

[1503-44]

These should be the first three items on your list.
1. Reduce velocity
2. Reduce velocity
3. Reduce velocity

Add all three of those to your list.


You might also try reducing pressure drop across CV
(you may need two valve to do that if dP is extreme.)

 

[hacksaw]

Your valve supplier has several approaches for controlling the noise.

One question: your color image seems ro imply that that you have a ball valve.

Given that your pressure drop is 32 psi with inlet pressure of 74.7 psia, you will likely have excessive noise.

 

[1503-44]

Ball valves are extremely poor control valves. If that is a ball, change it to a globe or similar type valve suitable for control purposes..

 

[davefitz]

I would ask the valve supplier to document how he arrived at the 85dBa at 3' estimate. You can review the standard procedure for noise prediction in the ISA Handbook of Control Valves. If it is a conventional valve it likley does not have any noise reduction features internal to the valve, espescially when it is throttled.

There are several noise reduction techniques used to reduce the jet noise from control valves throttling clean gases, such as the CCI Self Drag internals ( excellent noise reduction but poor shutoff leakage) , and Fisher and others have specially designed globe valve internal perforated cages to reduce noise.

There also exists several means of reduction the noise from escaping the piping system, including lead shielding ( low frequency) and Rho-Cee plastic ( as used in submarines) for wide range jet noise. See an acoustics engineer for those techniques.

Finally, if there is only one primary frequency with harmonics, a helmholtz resonator would work, but I think your issue is wide range jet noise plus a less than honest valve salesman, so replacing the valve with a CCI valve would be the preferred fix.Hypothetically , installing multiple internal vanes at the downstream pipe will reduce the diameter of the jets , which also reduces the noise emission, but I do not think that has been used commercially.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[KevinNZ]

Dropping pressure through many small holes and non critical pressure drop stages makes a big difference to noise levels. If you have fixed-maximum flow most of the time, you can have a large control valve and a number of multiple hole back pressure plates in series. If you need low noise of the a range of flows the pressure drops need to be built into the valve.

 

[grhead]

Thanks everyone taking the time for your detailed responses, Sorry I've taken so long to respond:

The FCV (in the first sketch) is an 8" V260 V-Ball Flow control valve (Fisher) with factory integrated aerodome noise attenuator - [URL unfurl="true"]https://www.emerson.com/documents/automation/product-bulletin-fisher-v260-ball-control-valve-en-125070.pdf[/url]

The supplier is a fisher rep, my assumption is that they use Fisher software/methods to calculate their noise predictions, however, I do not see a stated method of noise prediction in their sizing sheets.

Also, the actual predicted noise is 81 Db at full load (not 85 as Initially posted).

@LittleInch - Well from my calc you're doing 30m/sec in that 16" pipe so it's not surprising it's making a loud noise.... A bigger pipe maybe? - 20" or 24"?

I would typically design for a max. of approx. 140 ft/s for similar natural gas installations, and haven't seen this level of noise. I believe the noise is being generated from the valve and propagating, and the actual pipe velocities downstream are not a huge contributing factor. We're always trying to keep pipe sizes down for cost and space considerations inside the plant.

Min. flow for this valve is 90,000 scfh @ 0.5 psi So we have a pretty substantial 16:1 turn-down. Two valves may be a solution... It complicates our flow control scheme considerably as then we have a cross over point when one valve hands over control to the other. We have done this before, but it's not preferred as we add complexity, size and cost.

@1503-44

As I mentioned above, I believe our noise is originating at the valve, not due to pipe velocities. This is only based on location of the noise, we have piping farther downstream at similar velocities where we are measure 85Db and below. But maybe I'm wrong in this assumption? Is there a target velocity "rule of thumb" that you work with?

We have a pressure regulator upstream to reduce inlet pressure (and therefore DP across the valve). We played around a bit with the inlet pressure during the design phase, with our control valve supplier. They came back with the 60 psi inlet pressure as the best option to get the turn-down and control across the range of flow conditions. Perhaps this is worth re-examining though.

@hacksaw
Yes we are using a v-ball flow control valve. Please disregard the color image as it is just to show an example of the noise trap concept... The sketch best reflects the actual installation.

@davefitz

Thanks for the info, as I mentioned above this valve does have some noise attenuation as part of its design. Your point on the theoretical noise calc is well taken, I'll need to dig into how they arrive at this number more closely.

I'm curious about the CCI valves, I had a quick look at their website. Looks like these may typically be used for fluid or steam service. I don't see anything related to fuel gas control, but perhaps they'll have more info if I contact them directly.

@KevinNZ
This application we are often modulating within in the Min/Max. flow control range of the valve in a given day. I'm assuming that this would mean the back pressure plates wouldn't be a viable solution?

 

[1503-44]

Ive used two valves in series with Fisher's Whisper Trim and that worked for me, but it was also a 1000psi letdown pressure with diesel. We were cutting a single valve to pieces every 6mo. Stopped both the erosion and noise.

I try to keep gas velocity below 80fps. About half of yours. I've never seen any company standard or recommended max as high as 140fps.

 

[grhead]

@1503-44

Wow, 1000 psi, that's hefty. Were they globe valves?
Re: Velocity - I just dug out another consultants spec that called out max. fluid velocity limits of 167 ft/s for 6" and smaller and 250 ft/s for 8" and larger for natural gas pipe and fittings.
Holding to 80 fps for our piping in the plant would push the piping sizes way larger than than practical for the allowable space around other equipment/piping. Our typical max. gas pressure are 20-30 psi, and at this low pressure its a trade-off between pipe velocity and pipe size to still get the flows we need.

 

[1503-44]

Yes globe valves on a high pressure pump recycling line that experienced lots of runtime on recycle, so it was exposed up to full 2160psig pump discharge pressure, They could start closing the recycle line at a around 1200psig.


Just to give you a rough idea on max gas velocity. I searched "maximum gas pipe velocity"
What popped up was this, Without opening any particular web page, I see numbers ranging from 20m/s to 145fps. I agree that at low pressures, you will like to use faster velocities, but as you have found out, that can bring on noise. I'd lean towards those faster velocities where you are sure that noise will not be a problem. Inside a compressor building, there is enough noise without adding gas velocity to it. Out in a corner of a plant away from offices, shops and adjacent property owners, you can turn up the volume a bit. Your low density associated with low pressures should help ... up to a point.

https://duckduckgo.com/?q=maximum+gas+pipe+velocity&t=brave&ia=web

I just did a gas pipe noise search for you too. Maybe there are some good hints there as to how to avoid it. I didnt look any closer.

https://duckduckgo.com/?q=estimate+gas+pipe+noise&t=brave&ia=web

Sorry I'm not so much help today, but here its late Friday afternoon and the beach is only 1.5km away.

 

[hacksaw]

gr,

you might be dealing with accoustic resonance near down stream branch connections, sort of the organ pipe scenario where the branch piping resonates.

the resonance effects arn't comming from your valve.

 

[1503-44]

The thin pipe walls resulting from low pressure design probably doesn't help.

It doesn't look like a long run of pipe, but if it is, do not support at uniform spacing. Stagger the support spacings so you will have different resonant frequencies in each span.

 

[hacksaw]

gr,
what's the pipe schedule, and if you have a measure of the sound spectra that helps, the frequency of you noise, the shell modes of the pipe wall come in handy when dealing with accoustic resonances.

 

[grhead]

@1503-44, @hacksaw

We are using schd. 40 pipe. I would consider schd. 80 next time around.
I never thought about the the support spacing, but it is pretty random... we don't have and long continuous runs, basically the pipe turns vertically up after the skid and then feeds a header distribution system with lots of elbows,reducers and takeoffs.

I've been told the gas pipe noise is in the 1000 Hz frequency and above range. That's about all the info I have on noise frequencies.

@ALL

Thanks everyone for your input, I've got some starting points to do a little more research on. If you have anything to add feel free to continue commenting and I'll let you guys know if I come up with any additional insights.

 

[1503-44]

You can look at the typical noise mitigation things to do as well. Distance from concrete walls, building a wall to direct the noise upward, etc.

 

[davefitz]

After reviewing the pipe configuration, it is possible that much of the noise is from the geometry of the valve outlet expander 8"x16". If the angle of the change in ID ( to the horiontal axis ) exceeds 20 degreees, then flow separation occurs and large turbulent jets vortices may may be formed, which can cause jet noise. It is possible that a custom expander with an angle to the axial horizon less than 20 degrees then a reduction in noise may occur.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[SPDL310]

I'm with Dave on this one. Even if you aren't getting jetting in the expander it will act like a megaphone. Adding a diffusor or baffles in the 16" line can increase the frequency to outside of the audible range. Is the expander integral to the valve outlet or part of the piping? If the noise is specifically centered around the elbow as you originally stated I think it is probably the velocity as previously mentioned by 1503-44. You have a serious momentum change making that 90 degree turn. I suspect you may be getting compression and re-expansion in this area. A longer radius/thicker schedule elbow may help. Sources of noise add to each other so it may take multiple corrective actions if you are looking to reduce the overall noise in the area.

 

[1503-44]

I'm with Dave. Everybody is with Dave. No arguments at all. This is just as much aerodynamics as it is pipe flow, if not more.

 

[grhead]

@davefitz @SPDL310

Good point on the expander.
I've got a standard 8x16 concentric reducer. Dimensions OD D = 406.4, OD D1 = 219, L = 356 (mm)..

I did a quick and dirty scale off a drawing and if I'm looking at the right angle, it is approx. 25 degrees. So the Jet noise phenomena may be occurring.
Just curious, where does the 20 degree number come from?

You guys got me thinking, In the past I have seen installations where the valve supplier supplied a custom transition piece with the valve which eliminated the need for an expansion fitting directly downstream of the valve. I believe this piece included some integrated baffles to reduce noise. Something I will definitely approach valve suppliers about.

Thanks!