26 replies to this topic

[irprocess]

Dear All,

In attached datasheet of control valve, Noise level is limited to 85dBA. But, based on vendor calculation in downstream pipe after the drilled disc, velocity will be supersonic for maximum flow rate condition. In normal case velocity is very close to a supersonic condition. These velocities can cause vibrations and high noise.
VENDOR suggests changing the size of pipe in downstream Valve and installing a ONE SIZE greater than existing size.

I want to know, it is possible to change in valve characteristic instead to change of size the existing pipe???

For your information,in scope work of project; New valve shall be replaced with existing Valve and we don’t like to change the existing pipe in plant.

Thanks, irprocess

Attached Files

•  control valve datasheet.pdf   18.06KB   321 downloads

Edited by irprocess, 19 October 2011 - 08:09 AM.

[GS81Process]

I believe that you have a choked flow condition across the control valve based on the high pressure drop listed on your datasheet. A choked flow condition is caused by the natural gas reaching sonic velocity within the control valve itself.

The gas cannot be supersonic downstream of the control valve as the vendor stated. Sonic flow causes a choke point. You will not have supersonic flow.

Edited by GS81Process, 15 November 2011 - 04:32 PM.

[kkala]

Meaning of "drilled disc'' is not clear to me. It will be ignored herebelow, but please revert if it has a role.
1. From the data of "control valve data sheet.pdf" we can see max & normal flow through the valve, corresponding upstream & downstream pressure, gas MW, other gas properties (assumed compr factor = 1). Sonic velocity in gas is estimated at ~ 428 m/s (not pressure dependent). Pipe size downstream valve remains at 2" Sch 80, while downstream pressure is 0.5 barg. Flow through valve is choked, as pointed out by GS81Process, since ratio of upstream to downstream absolute pressure is much higher than 2 (5.3 - 6.4).
2. Assuming that downstream pressure of 0.5 barg should be more or less retained (logical, as written in data sheet), downstream velocity is estimated at 0.77 Mach for max flow and 0.70 Mach for normal flow. Resulting frictional pressure drop is excessive, 23 or 19 bar / 100 m. Increase of pipe size to 4" would result in 0.20 - 0.18 Mach, with frictional pressure drop 0.70 - 0.58 bar / 100 m, which could be acceptable (unless resulting frictional ΔP is unacceptable, suggesting higher line diameter).
3. Supposing that another valve is installed and the flow need not be choked downstream of it. Then pressure downstream of valve can be 8.2 barg at max flow, but sufficiently lower than 7 barg at normal flow, so that the valve can take ~30% of system frictional ΔP. The latter seems realizable through a 3" downstream line, not through existing 2". But a precise calculation of ΔP is needed, since no details of piping configuration is known.
4. All above is preliminary, based on rough estimates needing confirmation through detailed work. Indicative conclusions are: (α) in case of 0.5 barg downstream pressure, check downstream line for 4" diameter (β) in the (rather improbable) case No 3, check downstream line for 3" diameter.

[ankur2061]

Assumption of compressibility factor = 1 at 8.7 bar and 7 bar is extremely questionable.

[kkala]

Assumption of compressibility factor = 1 at 8.7 bar and 7 bar is extremely questionable.

Gas has temperature = 31 oC and MW=18.09, so it is mainly CH4. Compressibility factor of CH4 is 0.9836 at 8.7 barg, as predicted by the software offered by breitzh in http://www.cheresour...h__1#entry49288. Role of compressibility factor is not significant here.

Attached Files

•  Z factor & specific gravity real gas.xls   870KB   226 downloads

[DB Shah]

Instrument engineer pointing out such issues puts a question mark on the process engineering skills. Before giving the spec sheet for procurement you should have checked the line hydraulics.

Line sizing of fuel header is not correct (as pointed out by kkala). The upstream fuel header will have velocity of >50m/s and d/s line of 2" will have >250 m/s of velocity. I will be comfortable with at least 3" inlet line size and 8" outlet header size.
(Another tip - Max CV possible in a line is @ 10d^2 (d in inch). It can be higher for butterfly valves)
In your case the calculated CV for the valve is ~24. Considering equal percent characteristics and 80% open at max flow, you get ~ 50 CV valve.

Hence the final specs which I would prefer is 3" inlet line, 50 cv control valve and 8" outlet header. (Although 4" inlet line will be required if press drop will be a constraint).

Hope this helps.

[ankur2061]

The assumption that it is mainly methane is also questionable. Since the gas is a fuel gas it could be a mixture of various hydrocarbons. In such a case for determining the compressibility factor, you require to calculate the pseudo-crtitical pressure and pseudo-critical temperature of the gas based on its composition. The datasheet for control valve should generally have the information related to the critical or pseudo-critical pressure and temperature of the gas which in this case is missing. Standard control valve datasheet template that I used to utilize had this information required to be filled in. Alternatively, the gas composition of the gas could have been provided in the notes / remarks section of the datasheet for the vendor to determine the compressibility factor and subsequently to calculate the gas density for purpose of sizing the control valve.

Sonic Velocity is calculated using the formula

Sonic Velocity, V_s = (k*R*T/M)^0.5

where:

V_s = sonic velocity, m/s (which corresponds to a Mach No. of 1 for that gas)

k = ratio of specific heats @ temperature, pressure

R = Gas constant, 8314 J/gmole-K

T = absolute gas temperature, K

M = Gas molecular weight

Compressibility factor does not appear in this equation. What is important in determining Mach No. is the correct inputs related to the specific heat ratio (k) at the given pressure and temperature and the molecular weight of the gas.

Regards,
Ankur.

Edited by ankur2061, 23 October 2011 - 11:15 PM.

[breizh]

I've attached a resource to support your query,
hope this helps
Breizh

[S.AHMAD]

The sonic flow is not at the piping but at the control valve itself.
Please note the following

1. Changing the downstream pipeline to bigger one is not going to improve your sonic problem. Should do the opposite.
2. In order to avoid sonic flow, the downstream pressure must be greater than 3.3 bar for normal flow at upstream pressure of 7 barG.
This is based on ratio as given by (2/(k+1))^(k/(k-1))
for k=1.31, the ratio is 0.54
3. This means you need to increase the downstream pressure drop such that the control valve downstream pressure is about 4 barG or greater.
4. You can achieve this by installing two globe valve or two hand control valve (HCV) downstream and install pressure gauge downstream (existing valve) to measure the pressure.
5. Reduce the sound by adjusting the HCV.
6. Alternatively, install pressure control valve.
7. One control valve may not be sufficient
8. Pressure control valve or HCV may be installed upstream as well (e.g. one upstream and another one downstream)
9. Alternatively, can also install a series of perforated plate.

Edited by S.AHMAD, 24 October 2011 - 02:47 AM.

[kkala]

I think opinions in posts will be more converging, if irprocess can give more relevant information, such as:
1. Purpose of the control valve, where downstream line goes ( a scheme is very useful).
2. Necessity to keep valve outlet pressure at 0.5 kg/cm2 g (per data sheet) or not. Are there other low pressure lines connected to downstream line?
3. Rough composition of the fuel gas handled?
4. Approximate physical length (total equivalent better) of downstream line, operating pressure at destination.
5. What is the mentioned "drilled disc"?
6. Brief description of situation before and after the problem.
Need of requested clarifications may get clearer in the following.
DB Shah: Upstream 2" line; v = 51 m/s, versus max allowable 122/SQRT(ρ) = 122/SQRT(6.95) = 46 m/s per local practices for HC vapors (ρ=density in kg/m3). ΔPf = 16 psi / 100 ft, too high. You are right that upstream line diameter should also increase.
Downstream line; Quick estimate by kkala, assumed a short 4" pipe (less than 100 m equiv length) to flare or atmosphere ("unless resulting frictional ΔP is unacceptable, suggesting higher line diameter"). So only the Mach limitation was considered, plus that ΔPf should not exceed available ΔP. For a normal gas line, a 6" pipe would be assumed, long length could increase this diameter.
Destination to flare was supposed because 0.5 barg downstream of valve looked small for any use (e.g. burners).
ankur2061: Can you suggest value for compressibility factor and sound velocity to check adopted ones? (1.0 and 428 m/s). Gas critical pressure = 44.8 barg (valve data sheet). To my opinion values adopted are adequate, but no check is bad.
S.AHMAD: It has been asked whether downstream pressure can increase, or it has some purpose to remain at 0.5 kg/cm2 g. Downstream pressure cannot go much higher than 4 barg at normal flow, for the new control valve should have adequate pressure drop for control (say pressure = 5 or 4 barg).
Can you advice need of two downstream globe valves or HCVs (at some distance?) instead of one? I understand this is more efficient, but reason may be more specific.
Hoping of settlement of this hot topic

Edited by kkala, 24 October 2011 - 03:43 PM.

[S.AHMAD]

1. By studying the valve datasheet I agree with GS81Process that the choke flow occur in the control valve, not in the piping downstream
2. If choke flow occurs in the downstream piping, the control valve will not controlling since the flow is already maximum and remains constant all the time.
3. As mentioned by Kkala, we need more information to analyze the problem
4. Irprocess, please help us in order to help you.

[DB Shah]

Yes kkala, 0.5 Barg is certainly low for any burner, downstream line will be sized based on mach number criteria if it vents.

[ankur2061]

Kkala,

The formula you have used for maximum allowable velocity (122/ Sq Root (rho)) is generally used to calculate the erosional velocity limit in 2-phase flow (API RP 14E). For single phase gas lines Norsok standards have a different view on maximum velocity in pipes

max velocity = 175*(1 / rho)^0.43 or 60 m/s whichever is lowest

This I believe would only be applicable to well supported pipes running for relatively short distances.

My personal experience says that gas velocity should be restricted to not more than 120 ft/s for any in-plant piping since invariably it has been found that the piping supports have their own limitations in preventing pipe vibrations due to high velocities.

As I have earlier mentioned compressibility factor is a function of critical or pseudo-crtitical pressure and temperature of the gas or gas mixture. Find out these values for yourself and inform whether gas compressibility is still 1 or not.

You have not provided the inputs for arriving the sonic velocity of 428 m/s. Provide these and I will inform you whether your calculation is correct or not.

Regards,
Ankur.

[S.AHMAD]

1. The sonic velocity is computed as below:

V_c = (kRT/M)^0.5 = (1.31 x 8.314 x 1000 x 304/18.09)^0.5 = 427.8 m/s

2. The z compressibility factor at very low pressure is very close to 1. Refer to the generalized compressibility chart
3. The equation V_o = A/rho^C is the formula for optimum economic velocity in pipeline where A is depending on cost ratio which does not vary so much with time (for details refer to any textbook on fluid flow)

Edited by S.AHMAD, 25 October 2011 - 01:42 AM.

[kkala]

kkala,
The formula you have used for maximum allowable velocity (122/ Sq Root (rho)) is generally used to calculate the erosional velocity limit in 2-phase flow (API RP 14E). For single phase gas lines Norsok standards have a different view on maximum velocity in pipes max velocity = 175*(1 / rho)^0.43 or 60 m/s whichever is lowest
This I believe would only be applicable to well supported pipes running for relatively short distances.
My personal experience says that gas velocity should be restricted to not more than 120 ft/s for any in-plant piping since invariably it has been found that the piping supports have their own limitations in preventing pipe vibrations due to high velocities.
As I have earlier mentioned compressibility factor is a function of critical or pseudo-crtitical pressure and temperature of the gas or gas mixture. Find out these values for yourself and inform whether gas compressibility is still 1 or not.
You have not provided the inputs for arriving the sonic velocity of 428 m/s. Provide these and I will inform you whether your calculation is correct or not.

The formula for max velocity (122/SQRT(ρ), metric units) is used for HC vapor lines according to local practices and I have seen it applied for any gas too. It is also applied for two phase lines (ρ = mixture density), but minimum velocity should be at least 60% of this (can be lower when phase separation is desirable in horisontal pipes).
Norsok would recommend 175/6.95^^0.43 =76 m/s, that is 60 m/s, versus 46 m/s per local practices, differing somehow from place to place. The concept of max gas velocity (60 or 36.5 m/s) is reasonable. In the case of upstream pipe, frictional pressure drop per unit length (not allowable velocity) seems to define the diameter.
Even though figures had better be calculated by the one making the comment, gas compressibility factor at 8.7 barg & 31 oC is estimated at 0.975 for an assumed gas composition, so that MW=18.1. See attached "FGCF.xls". Pressures can be much lower, compressibility factor does not seem to play a role, other uncertainties are significant.
Control valve data sheet contains data for sonic velocity calculation. Estimated 428 m/s has been already confirmed.

Editing note: Formula for max allowable gas velocity is not applicable for steam.
Allowable velocities are higher in PSV lines.

Attached Files

•  FGCF.xls   771.5KB   198 downloads

Edited by kkala, 25 October 2011 - 01:00 PM.

[ankur2061]

kkala,

Why is it that every design method or calculation has to be according to "local practices". Is this forum a local forum in Greece.

Regards,
Ankur.

[kkala]

kkala, Why is it that every design method or calculation has to be according to "local practices". Is this forum a local forum in Greece.Regards. Ankur.

Local companies apply them (not by law), not contradicting recognized international standards. When I refer "according to local practices", relevant figure is not arbitrary but has a basis. It is not an individual opinion.
"Local" does not mean limited to Greece, I have seen them applied here. In fact they have much wider influence (and they are influenced). Of course "every design method or calculation has to be according to local practices" contains some exaggeration. Piping sizing practices are understood to differ from organization to organization. But in general European mentality and Directives have to be followed.
Reputable Engineering companies have their own practices, sometimes simplifying or further interpreting international standards.

Edited by kkala, 25 October 2011 - 08:37 AM.

[GS81Process]

I think that it can be agreed by all that the original line sizing was performed incorrectly because of excess pressure drop, regardless of minor differences between sizing guidelines for gas service that each of us may use (pressure drop per length and velocity). There is too much pressure loss across the piping relative to the control valve. This makes for poor pressure control.

I have come across a similar situation before and I admit this gives me some sympathy for the original poster. Telling management that entire sections of existing gas piping need replacement can be very unpopular!

Edited by GS81Process, 25 October 2011 - 02:46 PM.

[ankur2061]

Local companies apply them (not by law), not contradicting recognized international standards. When I refer "according to local practices", relevant figure is not arbitrary but has a basis. It is not an individual opinion.
"Local" does not mean limited to Greece, I have seen them applied here. In fact they have much wider influence (and they are influenced). Of course "every design method or calculation has to be according to local practices" contains some exaggeration.

You say that the figure is not arbitrary but has a basis. That is what you know, but not the readers of your post. Imagine a design engineer submitting a design report and writing "According to local practices", do you think that the client will accept that. Any design report has to mention the exact reference in terms of the name, edition and revision of the reference which you have used for the equations or method of calculation. By not providing the proper reference you are providing an incomplete report or calculation and no client would accept that. Providing all the references related to any design report or calculations is an absolute must for its acceptability.

When you mention any equations which are not generic in nature such as the maximum velocity equation that you posted you need to provide the exact reference rather than just saying "According to Local Practices". This will help the reader of your post to understand the origin or source of the equation.

Regards,
Ankur.

[kkala]

In the relevant case of line upstream the valve, recommended maximum velocity would be 60 m/s per Norsok, 46 m/s per local practice, 36.5 m/s per Ankur's experience (as previous posts indicate). The value of 46 m/s is in the ball park, judged sufficient for our activities here, which differ to documented engineering. "Local practice" is noted, to get reader aware and judge the criterion if necessary. I could not give other references, not to also say that matter of velocities is insignificant in this specific case, where pressure drop per unit length determines pipe diameter.
An engineering work starts after approved design criteria by Client, including piping, so no confusion is anticipated. Looking into practices of reputable engineering companies along the world, you will most probably find this criterion for allowable gas velocity, or something not far. After all there is similarity to Norsok's criterion, being less conservative for gas densities higher than ~5 kg/m3. So the concept of allowable gas velocity not higher than 60 m/s had better be added to the local practice (mentioned in previous post), not affecting the examined case.
Above is not an important matter for the topic "Supersonic Velocity in Control Valve", but hopefully has a cooling effect. Good night.

Note: Formula for max allowable gas velocity is not applicable for steam.
Allowable velocities are higher in PSV lines.

Edited by kkala, 25 October 2011 - 12:57 PM.

[S.AHMAD]

Dear friend

1. In order to solve this type of problem - I have made suggestion earlier which is a cheaper option.
2. We can also change the control valve to a multiple-steps control valve. This is more expensive option but using the same concept of multiple pressure drop.
3. We can also predict the noise from control valve
4. Attached is a write-up by Masoneilan which was published in 1977. You probably can get latest version of the write-up from valve manufacturer.
5. This write-up suggests few solutions with regard to your problem
6. It seems that vendor unable to contribute (in fact they made a wrong suggestion), I suggest that the owner contact the valve manufacturer directly.

Attached Files

•  ControValveNoise.pdf   7.38MB   212 downloads

Edited by S.AHMAD, 26 October 2011 - 07:20 PM.

[kkala]

Masoneilan Noise Control Manual looks quite useful with its practical examples, so that we could probably make an elementary noise analysis, when required. Thanks S.AHMAD, it has not been found in web. Acceptable noise levels have been a bit stricter today, but this does not lower the value of the document.

By the way, below is an example of noise requirements, written by Process in inquiries.

1. Any piece of equipment shall produce a noise level not higher than 85 dB absolute (dBa), measured at 1 m distance.

2. The above limit is applicable to all continuous flows. Discharges from vents and safety valves shall result in no more than 95 dBa on any place or platform, where people may be during the operation . If necessary, downstream silencers shall be installed.

3. Acoustic insulation, to obtain the level of 85 dBa, shall be avoided. It may be applied in specific cases after Client's approval and evidence that there is no other alternative (e. g. low noise trim for control valves).

Note: Specific requirements for each piece of equipment (e.g. pump, compressor) are specified by other Departments.

It is noted that Masoneilan's Manual refers to acoustic insulation (differs to thermal) for whole downstream pipe. It also advises that not only sonic velocity at throttling area but also high velocities downstream valve create excessive noise. It gives guidelines to calculate noise level.

Edited by kkala, 27 October 2011 - 06:38 AM.

[breizh]

kostas ,
This link will give you an update of the paper you are looking for :


http://www.soljet.co...nertech/413.pdf

Breizh

[S.AHMAD]

Breizh, thank you for sharing. Appreciate very much.

[irprocess]

Dear KKALA,

I prepared some information as you requested, hope to be useful in order to determine supersonic velocity.
Quote:
1. Purpose of the control valve, where downstream line goes ( a scheme is very useful).
This valve is located on upstream of LP fuel gas drum and also used for transferring the extra gas to LP flare header (with back pressure= 0.5 barg).Also two sketch of valve’s location are attached.
2. Necessity to keep valve outlet pressure at 0.5 kg/cm2 g (per data sheet) or not. Are there other low pressure lines connected to downstream line?
As mentioned in item 1 , this line is connected to LP flare header with back pressure of 0.5 barg.
3. Rough composition of the fuel gas handled?
Mole Fraction (C1=0.8856,C2=0.0372,C3=0.0140,N2=0.0495 ),
MW=18.1,
Mass Density= 5.8 kg/m3
4. Approximate physical length (total equivalent better) of downstream line, operating pressure at destination.
Approximate length to near LP flare header line=50 m
5. What is the mentioned "drilled disc"? This is a vendor’s pronunciation.
6. Brief description of situation before and after the problem. Please see item 1

Attached Files

•  VALVE 2.jpg   75.37KB   58 downloads
•  VALVE 1.jpg   55.58KB   48 downloads