Single Phase Pipe sizing criteria 3

[ChemE2912]

Hi Folks!

I have been using the V = C/(Density^0.5) criteria for sizing lines, where C ranges from 90-160 and the Density is in lb/ft3. V is the critical velocity in ft/s. If C=160, then we tend to reach erosional velocity limits in the pipe.

This rule is applicable for single and two-phase flow.

Now, I am trying to convince my Mechanical Engineer colleagues that we should not exceed the erosional limit (160/(Density^0.5)), if we want the pipe to last for a long period.

Do you happen to know any correlation that can predict the erosional wear if we exceed these erosional limits?

Also I would like to know if the above indicated correlation is applicable to fittings as well??

Regards,

Alberto J. Baumeister

 

[chicopee]

What is the medium within the pipe?

 

[ChemE2912]

Chicopee,

The fluid within the pipe is essentially water...

 

[athomas236]

ChemE2912,

In power industry (according to my experience) water lines are usually designed for a velocity of about 3 meters per second.

athomas236

 

[dbevil]

With considerable experience in the subject of erosional velocities, I offer the following memo drafted to address the subject as pertains to an ammonia pipeline:

Many companies, have in their process piping go-bys or rules-of-thumb, references to a maximum velocity of fluids in piping. This maximum recommended velocity is referred to as ‘erosional velocity’ and is typically given as some constant divided by the square root of the density of the fluid. The constant will vary from company to company but generally ranges between 100 and 160 for systems operating continuously. Higher constants are sometimes used for intermittent service. In actuality, this value has no engineering basis when dealing with clean, single-phase fluids that are not corrosive to steel.

The background of erosional velocity stems from piping failures (circa 1940’s) due to loss of metal in elbows that had been subjected to many years of constant high velocity fluids. Research has revealed that this metal loss was primarily due to two primary factors:

1) two-phase fluids in high velocity service (liquid droplets in gas service or solid particles in liquid service) OR

2) a high velocity fluid that is corrosive to steel.

In the case of clean, single-phase fluids, there is ample evidence in the physical world that there is no appreciable metal loss to steel exposed to high velocity fluids, even at velocities 10 or more times that of ‘erosional’ velocity. Some examples are impellers of pumps and seats on control valves. In clean, single-phase service, most of the damage to these devices is due to pitting caused from cavitation, which is caused by sonic velocities. However, any solid particles present in the fluid can cause rapid deterioration of steel due to the impingement of the particles on the metal surface at a change in direction (typically, at elbows). With corrosive fluids, the impact of the high velocity fluid and the eddy currents created by changes in direction, tend to concentrate the loss of metal on elbows.

If there is concern about the possibility of some metal loss due to operating at or above an ‘erosional velocity’, even with a clean, single-phase fluid such as ammonia, then remedial measures can be put into place to monitor and track any metal loss encountered. Such measures could include periodic ultrasonic wall thickness tests on elbows that would be prime candidates for metal loss. Due to the length of time required to cause damage (years of continuous service), proper monitoring will identify a potential problem long before it can progress to the point of failure.

 

[jmgmeana]

How is this erosional limit determined? what are it origins?

 

[BRIS]

dbevil makes some sound points. Fluid will not cause erosion - Erosion is either caused by solid particles within the fluid or by cavitation.

With clean fluid the only concern is cavitation. This occures when the velocity and/or the temperature are sufficiently high to cause the pressure to drop to vapour pressure. High velocities are caused at disturbances in the flow such as at elbows where the flow is contracted and then expands. But I have to disagree on the reference to sonic velocity - cavatation has nothing to do with "sonic" velocity. If the pressure is low or the temperature is high then cavtation can occur at very low velocities. In water flowing at atmospheric pressure, at 20 degree C, cavitation will not occur until velocities exceed 10m/sec (30 ft/sec). But, note velocities at elbows etc are likely to be twice that in the straight pipe.

brian

 

[zdas04]

I think dbevil and I have been looking at the same references. I've found the constant over the square root of density in a bunch of sources as well. I also can not find any scientific or experimental basis for the equation - it just shows up in engineering-standards without any explaination, derivation, or reference.

I've got a 1959 Gas Engineer's Handbook that is usually a very good source of the archaeology of this sort of equation and it has nothing on erosion. That probably means that this equation was developed and achieved wide-spread acceptance after 1959.

Brian,
The "sonic velocity" referenced above is the right physics. Cavitation is defined as "the formation and subsequent collapse of vapor bubbles in the flow". When the bubbles collapse (even in laminar flow streams) the liquid rushes into the void at or near sonic velocity. The momentum transfer of this high-velocity flow is the damage mechanism in cavitation.

Also, bulk velocity does not change in an elbow. Acceleration changes magnitude. The velocity profile changes and v(max) shifts from near the centerline to close to the outer wall so the dv between the no-flow boundry and the flowing stream increases (along with the fluid shear), but the velocity of the flow doesn't "double". The Continutiy Theory says that mass flow rate in a pipe has to be the same everywhere along the flow. If you define bulk velocity as mass flow rate over the quantity density times cross-sectional area, then bulk velocity does not change in a constant-diameter elbow. The problem with errosion at elbows (when it really occurs) is the increased shear force from the v(max) shifting towards the outer edge.

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

 

[davefitz]

There are several issues to be evaluated.

1. In today's climate, energy efficiency is important, so teh cirtieria of limiting the frictional pressure drop may be one limiting factor. For sizing steam piping, some typical limits in pressure drop ( DP/Pi) are 3% for cold reheater Xfer pipe, 5% for hot reheater Xfer pipe, 2.5-3% for HP main steam Xfer pipe.

2. For water in carbon steel, the "economic pipe size" , balancing piping costs vs pump costs , is about 10 fps, but other considerations also apply at times.

3. The max water velocity in carbon steel for erosion /corrosion is about 25 fps, and that can be raised to over 50 fps for low alloy (1% Cr) and even higher velocities have been used for incolnel rolled piping ( ie, drum feed pipes). The erosion corrosion limits are also related to water chemistry, and the combined oxygenated treatemtn system has less erosion / corrosion protential than the classical AVT alkaline with oxygen scavengers. In this latter case , it is critical that overfeed of hydrazine not occur, so a 2 ppb O2 setpoint is mandated, and startup upsets in O2 and pH are not permitted.

4. Feedpump inlet piping has other associated criteria, and a max velocity in the inlet piping ( prior to pump suction swage) is about 7 fps to avoid NPSH issues ( assumes saturated liquid sourced off bottom of a dearator).

 

[BRIS]

David

Yes the collapse of bubbles is at sonic velocity and it is that that causes cavitation erosion but cavitation is caused by the fluid pressure dropping locally to vapour pressure it is not caused by sonic velocity.

Obviously bulk velocity does not change at an elbow. The velocity that matters is the point velocity and at an elbow this may be twice the straight pipe velocity. It is the local velocities at fittings that need to be considered when estimating the velocity limit for cavitation damage. In clean fluid I would suggest that it is generally cavitation that causes erosion and not increased shear force. Cavitation will occur first where there are increased local velocities and where there are discontinuities such as at an elbow.
brian