Control Valve Cavitation
When selecting control valves, is the efficiency of the pressure parameter more important than valve trim fluid velocity? There's been some important debate on this point recently in the valve manufacturing community. Although the trim exit velocity needs to be considered in control valve selection, it does not describe the whole physical phenomena that occur inside a control valve.
For further reading on control valves, access the following articles:
When selecting control valves, is the efficiency of the pressure parameter more important than valve trim fluid velocity? There's been some important debate on this point recently in the valve manufacturing community.
Although the trim exit velocity needs to be considered in control valve selection, it does not describe the whole physical phenomena that occur inside a control valve. Hence, the use of the trim velocity approach may not be reliable in solving many of the problems that may occur in a control valve and, in many cases, may not provide the most economical solution.
In a control valve handling a pure liquid, cavitation may occur if the static pressure of the flowing liquid decreases to a value less than the fluid vapor pressure. At this point, continuity of flow is broken by the formation of vapor bubbles. Since all control valves exhibit some pressure recovery, the final downstream pressure is generally higher than the orifice throat static pressure. When downstream pressure is higher than vapor pressure of the fluid, the vapor bubbles revert back to liquid. This two-stage transformation is defined as cavitation.
Valve pressure drop, vapor pressure, and valve pressure recovery factor are the parameters used to determine the full choked flow and cavitation condition.
The implosion of the vapor bubbles can cause local pressure waves up to 100,000 psi. Also, during cavitation, fluid micro-jets are formed due to asymmetrical bubble collapse. The combination of the high intensity pressure waves and micro-jet impingement on valve surfaces can cause severe damage. Cavitation damage leads to rapid deterioration of the valve plug and seat and causes valve body damage as well (see photo). It also may lead to noise and vibration problems and poses a potential safety hazard. Therefore, it is necessary to understand and to prevent this phenomenon, particularly when high-pressure drop conditions are encountered.
Three parameters should be considered for predicting cavitation in a control valve: valve inlet pressure (P1), the valve outlet pressure (P2), and liquid vapor pressure (Pv). The intensity of cavitation damage depends greatly on the relationship among these three parameters. The higher the pressure drop across the valve (P1-P2) and the closer Pv is to P2, the more intense the cavitation.
These two equations for service conditions sigma and valve sigma illustrate that fluid trim velocity cannot be used to predict cavitation damage.
Pressure recovery in a valve is a function of its particular internal geometry. In general, the more streamlined a valve is, the more pressure recovery is experienced—which increases the possibility of cavitation. The pressure drop in a valve that begins at cavitation is termed a critical pressure drop. Full choked flow will exist if the actual pressure drop is greater than the critical pressure drop, and if the downstream pressure is greater than fluid vapor pressure.
In the ISA S75.01 and IEC 534-2 standards for control valves, the critical pressure drop across the valve for a full choked flow is defined in Figure 1. The equations clearly depict that the liquid velocity in the trim is not the parameter to use to predict and determine the full choked flow or full cavitation. The valve pressure drop, the vapor pressure, and the valve pressure recovery factor are the parameters commonly used to determine the full choked flow and cavitation condition.
A method that predicts cavitation damage in control valves is known as the 'sigma method.' It is recommended by ISA 75.23 and based on laboratory testing and empirical data. This method compares 'service conditions sigma' to the 'valve sigma,' which is determined from laboratory testing for a particular geometry threshold of damage (see Figure 2).
Using the sigma method, the fluid trim velocity cannot be used to predict cavitation damage. The sigma method depends on the pressure drop across the valve, the vapor pressure, the valve size, and the other pressure/valve size parameters determined from testing the reference valves in the laboratory.
Equation for calculating liquid velocity in the trim cannot predict cavitation when liquid vapor pressure is not included.
Liquid velocity in the trim is calculated in Figure 3. For multi-stage trims, the velocity can be determined at each individual stage by using the pressure drop across each stage.
Although the trim velocity depends on the pressure drop, it cannot predict the occurrence of cavitation or the damage done by cavitation because the parameter of liquid vapor pressure is not included. In other words, the critical pressure drop evaluation and the sigma method must be referenced to predict the cavitation and its damage to the valve.
Suppose that the trim velocity analysis fails to capture the cavitation problem in a control valve. For this example, let's say water flows through a valve in two ways:
Condition 1—flow rate of 900 gpm, inlet pressure of 2,000 psig, outlet pressure of 1,000 psig, and vapor pressure of 1 psia.
Condition 2—flow rate of 900 gpm, inlet pressure of 500 psig, outlet pressure of 3 psig, and vapor pressure of 1 psia.
The pressure drop in condition 1 (1,000 psi) is higher than that of condition 2 (497 psi). Therefore, according to the trim velocity analysis, and using the equation in Figure 3, a higher K factor trim (more turns or stages in the trim) is required in condition 1 than condition 2 to achieve the desired low liquid trim velocity and avoid trim damage.
However, condition 2 has much more cavitation damage than condition 1, since the outlet pressure in condition 2 (3 psig) is much closer to the vapor pressure (1 psia) than that of condition 1 (1,000 psig).
This shows that the trim velocity analysis cannot predict cavitation damage in a control valve because it is solely dependent on (P1-P2), and it is missing the parameter of vapor pressure (Pv) and the relationship between P1, P2, and Pv.
Here is another example: Consider that, in many applications, imposing a small value on trim exit velocity, such as 100 ft/sec., can be very conservative and still insufficient in preventing the valve from cavitation damage. Achieving such a low trim exit velocity of 100 ft/sec will require a multistage trim that can be very expensive and not required.
In this case, the fluid is water at 100 8F, with flow rate of 200 gpm, inlet valve pressure P1 = 800 psig, outlet valve pressure P2 = 200 psig, vapor pressure Pv = 1 psia. The valve solution for this application can be a 2-in. single stage, cage throttling valve with small drilled holes that has a pressure recovery factory F L = 0.94.
Based on the equation in Figure 1, the critical pressure drop is 707 psig, which is greater than the pressure drop across the valve of 600 psig. Therefore, choked flow and cavitation do not exist in this valve.
Consider the sigma method for this application. Using the equations in Figure 2, the service sigma is 1.35 and the sigma valve is 1.13. The empirical parameters required to determine sigma valve are determined from testing a 2-in. 41005 series valve conducted in the Masoneilan flow laboratory. The service sigma is larger than the valve sigma; thus the sigma method predicts that the valve will not be damaged by cavitation.
On the other hand, using the equation in Figure 3 with K = 1.6 (for drilled holes), the trim exit velocity in this valve is approximately 236 ft/sec—much higher than the 100 ft/sec imposed by some valve manufacturers.
These examples show that the critical pressure drop method, and the sigma method proves that the single stage valve will work without being damaged by cavitation, despite a trim exit velocity much higher than 100 ft/sec. Moreover, the selected single stage cavitation containment valve for the above application is more economical than a multistage trim.
For related coverage, see also, 'Product Research' on control valves in this issue. Read this article at www.controleng.com , under Archives, January 2005, for links to related articles.
Joseph Shahda is principal valve engineer, engineered product, Masoneilan;