Estimate Clean Pressure Drop in Cone Strainer

[sshep]

We are installing some specially designed cone strainers at the outlet of a reactor to catch any catalyst pellets which might break through. This is vapor service (H2 and vaporized hydrocarbons) at 900F and 40psig. I would like to estimate in advance the pressure drop of the proposed design (designed by others).

The strainer is the surface of a circular cone fustum built into a spool of 24"dia pipe (ID=23.25"). The cone shaped support plate is lined with wire mesh..

This is my geometry:

R1=19.25" (note: there is a 1/2" thick solid support ring holding the cone away from wall so the effective pipe cross-section at this point is 19.25" dia)
R2=5" (note: this is a flat steel plate at small end of cone)
H=47.75" (distance between base and top faces of cone)
Support plate open area=40% (as 1/8" perforations)
mesh wire size=0.025" (layed inside, flow is inside to out)
wires/inch=16 (note: this gives 40% open area in mesh)

I get the surface area of the cone part from the formula:
A = pi*(R1+R2)*(H^2+(R1-R2)^2))^.5 = 26.4sqft
with a clean open area of
26.4*0.4*0.4 = 4.2 sqft (vs 2.9 sqft cross-section area of pipe)

which all looks reasonable.

Now I am wondering if anyone can tell me how to relate the pressure drop of this spool in terms of equivalent length of 24" pipe? A bonus would be a method which can also estimate the pressure drop as the open area is reduced by plugage.

Any help is appreciated.

best wishes,
sshep

 

[unclesyd]

Here is an old approximation for a clean strainer where a cone is equal to 100% of the pressure drop for same. A truncated cone is 80% of the value.


Pressure Drop (psi)= .0005*P*V^2
P= density = #/ft^3
V=fluid velocity of approach in ft/sec

This is taken from an old Mack Iron Works Chart

 

[MikeHalloran]

I would model the wire mesh and the perf as multiple orifices in parallel. The perf is easy enough, round holes with square edges, known thickness, etc. The wire mesh is more difficult because the orifices are square, and rounded, so it's hard to guess a coefficient with any confidence. There just _has_ to be a research paper, _somewhere_...

However, when you line the perf cone with wire mesh, the wires have to occlude the perf holes, in a way that simple tools can't hope to model.

Since you have, or could soon have, a physical manifestation, how about setting it up in a standpipe and doing a water flow test?



Mike Halloran
Pembroke Pines, FL, USA

 

[panduru]

@sshep:

And, as MikeHalloran suggests, if you do set it up in a standpipe and do a water flow test, you may please post your observations and recordings on this forum for the benefit of all of us.

Thanks.

 

[sshep]

Thanks guys,

A test stand probably won't be practical, but I should be able to get clean dP data under recycle gas circulation as a prestart-up activity- i.e. before hydrocarbons are introduced and things get dirty. We have some instruments to monitor which will be calibrated to give good resolution since each 1psid = at least $50k/yr electrical cost.

Unclesyd, your formula doesn't reference the strainer open area but may prove a useful for predicting other conditions once a reference data point is known.

I'll let you know if anything interesting is found in September when these changes come on line as part of a larger project.

Thanks,
Sean

 

[unclesyd]

The approximation is for comparisons to a conical strainer of length L.

Truncated cone 175% vol; 150% area; 80% pressure drop at length L

Inverted cone 150% vol; 200% area; 95% pressure drop at length L

Cylinder 300% vol; 200% area; 20% pressure drop at length L

 

[sshep]

Gentlemen,

I contacted the designer to give me an estimated pressure drop for this service. His answer was 2 to 3psid clean pressure drop for the normal flow condition. FYI, the specs I wrote were 276000pph, MW=30, 920F, 45psig, 0.03cP. I called for 0.2psid clean, and 1psid dirty (30% or more plugged), but the available length of piping we allowed was apparently too short for a design meeting my specs.

Now I will take suggestions for possible mitigations. Again thanks to all.

best wishes,
sshep