Emerson –The impact of valve outlet

velocity on control valve noise

and piping systems

ValveValveWorldWorld

Volume 6, Issue 2, April 2001

Cover Story

Emerson –The impact of valve outlet

velocity on control valve noise

and piping systems

Reprint

CONTROL VALVES

Valve �� World APRIL 2001 www.valve-world.net

The majority of these noisy control valvesare found in pressure reduction applica-

tions involving gases or steam. (As a pressuredrop is taken in a control valve, most of theresultant energy is lost in heat while the restis converted to noise.) Prior to the introduc-tion of specially designed valve trims thatlower the effective noise level at the device,the first efforts at quieting these valves in-volved installing fixed restrictions in thedownstream piping to create a back pressure.This back pressure would in turn reduce thepressure drop taken across the control valve,lowering the noise generated at the valve. While this technique may work well for oneset of process conditions, it usually fails toreduce noise should the conditions vary inthe slightest. Since the inlet and outlet pres-sures of the valve vary throughout thevalve’s entire range of travel, the likelihoodthat a fixed restriction could lower the noisein all cases is extremely unlikely. Today’s technology has evolved from drilled-hole valve trim that shifts the peak frequencyof the noise above the audible range of thehuman ear to trim designs that exhibit a va-riety of noise control techniques. Thesetechniques include frequency shifting, stagingthe pressure drop, shaping the flow passagesto control turbulence, keeping the flow jetsindependent as they exit the trim and lower-ing the velocity of the process fluid. While these trims reduce the audible noisegenerated by the throttling process at thevalve, the overall noise is still dependent

upon the selection of the proper valve typeand size. If the valve body size is too small,the outlet velocity will overpower any noisereduction that occurred at the trim.

Practical exampleSeveral years ago, a gas plant in Saudi Arabiaexperienced cracking of an Acid Flare Head-er. The site engineers carried out vibrationtests to determine the cause and proposedseveral changes to the piping system to elim-inate the cause of the damage. The changeswere implemented, but the problem contin-ued and eventually led to cracks in the pip-ing as well as the header supports.A historical review of the problem was con-ducted once again, but this time with processdata also being included in the evaluation. Afluid dynamics calculation was performed uti-lizing computational fluid dynamics (CFD)software to simulate overall process condi-tions. The detailed simulation included sonicvelocity calculations, gas velocity and vibrationcorrelation and pressure drop calculations. The analysis revealed that 10” valves in theheader system were the main contributors tothe excessive vibration. While the valves in-corporated noise abatement trims, they wereundersized and generated outlet velocitiesover 0.7 Mach. This not only created unac-ceptable noise levels at the valve, but alsoled to velocity induced vibration of the pip-ing system. It was proposed to install a newtype of noise abatement trim, but adjustingonly the trim would have had no effect, as

the valve outlet velocity was the driving fac-tor behind the excessive noise. The proper valve for the application proved tobe a 16” valve that had the appropriate levelof noise abatement trim, and the correct bodysize to eliminate any outlet velocity effects.The velocity at the 16” valve was reduced to0.3 Mach, which significantly reduced thenoise at the valve and eliminated the vibrationthat could affect the downstream piping.This example is typical with some manufac-turers in the control valve industry. A con-trol valve is selected based on the requiredCv to pass the flow and address the noisegenerated by the trim. The outlet size of thevalve is normally ignored. If an inadequatelysized valve is chosen, the valve, piping sys-tem and downstream equipment are in jeop-ardy of premature failure.

How outlet velocity increases noiselevelsControl valves are typically selected in sizessmaller than the adjacent piping for econom-ic reasons, but the piping is still subject tothe typical selection process using gas densi-ty and mass flow rate. This selection processusually leads to a larger pipe downstream ofthe pressure-reducing valve. With a smaller size control valve, a reduceror expander is required to mate the controlvalve with the piping. Problems can occurwith this configuration because the velocityrelated turbulence generated by the ex-pander at the valve outlet creates its ownnoise that often exceeds that of the noiseabatement trim. This turbulence is caused bythe difference in gas velocity between thevalve outlet and downstream pipe that cre-ates an additional noise source. The effect of this velocity related turbulencecan be seen in Figure 1 as the outlet pres-

The impact of valve outletvelocity on control valvenoise and piping systems

By John Wilson, Fisher Controls International, Inc.

Control valve noise has long been an issue at many process facilities because of its effect on process equipment. During the past threedecades, the task of controlling valve noise has been a major undertak-ing by the control valve industry. It’s only been in the past few yearsthat the environmental impact of control valve noise has been recognized and regulations put into place to limit its severity.

CONTROL VALVES

APRIL 2001 Valve �� World

sure decreases in a given application wherean 8” valve with low noise trim is installedin a 12” line. The noise due to the reducerbegins to become the overriding factor atdownstream pressures of less than 22 psia.As the outlet pressure continues to drop, thenoise calculated one meter from the pipe in-creases over the predicted valve trim noise.In the case noted below, the actual noise isfour to five dBA greater than that predictedby standard noise prediction methods. Noisepredicted by combining both the valve trimnoise and reducer noise nearly matches thatof the measured values.

Figure 2 shows the effect of valve outlet ve-locity on the overall noise created by a 4”valve installed in an 8” line. The standardvalve sizing programs take into account onlyvalve trim noise, which leads to inaccuratenoise prediction. However, actual test datashow that the valve noise is much greaterand dominated by outlet velocity effects.This also shows that different piping expan-sions will have different effects on the over-all noise at the control valve.

Both valve trim and valve outlet velocitynoise must be taken into account when pre-dicting the noise at the control valve. Onlyby combining these two noise sources canone truly estimate the effect of the controlvalve on the downstream environment andequipment.

Choosing the correct solutionNow that it has been determined that con-trol valve noise is due to more than justvalve trim, how is it possible to be sure thata valve selection is the correct one? The IECestablished the guideline to reduce the veloc-ity at the valve to 0.3 Mach or less. It ispossible, however, to install a control valvewhere the velocity is greater than 0.3 Mach.This is assuming that the effect of outlet ve-locity has been taken into account during thesizing of the control valve, and that the sub-sequent additional noise source will notcause the valve to be louder than required.IEC 534-8-3, “Control valve aerodynamicnoise prediction method,” accounts for ef-fects of valve outlet noise up to 0.8 Mach.This is due to advancements in the noiseprediction technique that take into accountthe presence of reducers/expanders and pip-ing effects. Prior to these recent advance-

ments, the standard was limited to Machlevels of 0.3 or less and only included thevalve trim noise in the calculations.Only by incorporating the IEC 534-8-3 stan-dard can users be sure of the noise level toexpect from a control valve once it is in-stalled. The standard was developed to allowa noise level to be calculated for any type ofvalve with any type of standard trim ornoise abatement trim. By better understand-ing the actual system noise that will be cre-ated once a control valve is installed, valveusers can be sure that they are protectingtheir plant’s equipment and ensuring envi-ronmental compliance. ■

References

Baumann, H.D., Predicting Control Valve Noise

at High Exit Velocities, InTech, February 1997,

Page 56.

Jury, Floyd D., Understanding IEC Noise

Prediction for Industrial Process Control Valves,

September 1997.

www.valve-world.net

About the author

John Wilson is a

Severe Service Sales

Engineer with Fisher

Controls

International, Inc. in

Marshalltown, IA,

and carries a B.S. in

Chemical

Engineering from the University of Nebraska.

For the past three years, he has worked

extensively on severe service applications

focused on the Power Industry. Prior to joining

Fisher Controls, Mr. Wilson worked for the

Omaha Public Power District and the Nebraska

Boiler Company.

Figure 1. Combining Valve and Reducer Noise

Figure 2. Predicted vs. Actual Control Valve Noise

(dB

A)

(dB

A)

D350828X012