Interperetation of API 618 4th Ed. Pressure Drop

[Gordo2020]

I am curious if anyone in the pulsation control business can comment / provide insight.

I am reading a report references API 618 4th Edition, clause 3.9.2.2.4 which states

"Unless otherwise specified, the pressure drop based on steady flow through a pulsation suppression device at the manufacturer's rated capacity shall not exceed 0.25 percent of the average absolute line pressure at the device or the percentage determined by Equation 5. whichever is higher."

The literal interpretation of this clause would suggest that placing "pulsation suppression devices" in series is ok so long as the individual pressure drops are below the allowable. This is an absurd interpretation as you could theoretically place an infinite number of suppression devices (and thus have infinite pressure drop), without violating that clause.

I'm wondering if there is a consensus among those that do these studies as to how to interpret this clause with respect to suppression devices in series?

This also prompts the question of what exactly constitutes a "pulsation suppression device".

Thanks

 

[sms]

A pulsation suppression device as discussed in 618 is usually a Hemholz filter, so a bottle with internals that create a two volumes connected with a choke tube, but could also be an empty bottle.

As for putting them in series, I used to do so fairly often, especially on large pipeline compressors where I would mainifold the cylinders with one, and then add a secondary bottle to improve the performance of the system.
It is absurd to suggest that anyone is going to get away with many vessels, but two is not uncommon at all.

Another typical trick I have used on process compressors was to use the suction knock out drum as a secondary volume in the pulsation control. In that case the KO drum was acting as both a KO and a pulsation suppression device.

I have also put large volumes in a system for the sole purpose of isolating a compressor from say fin fan coolers that were up in a pipe rack where they would be particularly prone to pulsation induced vibration, or from a centrifugal compressor, or even from another recip that it might beat against. In the case of the fin fans, this was cheaper that trying to tie down all that piping in the racks allowed the pipe racks to be prefabbed without acoustic analysis, and kept the compressor package design going rather than waiting for all that piping design to be done. In this case I used a bottle with an internal baffle and choke for the primary pulsation control, sized the KO drum accordingly, and then added an empty vessel out a ways to isolate the compressor.






"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[Gordo2020]

So when you have two vessels in series, did you apply the allowable pressure drop to each? i.e. 2x dP as per Eq 5 or did you check their total combined pressure drop against Eq. 5?

It seems to make more sense that Eq 5 would be the total allowable of the combined pressure drop of each vessel.

Clearly API 618 considers the helmholtz filters as their sizing is outlined in Appendix N or O, but what about orifice plates? They seem to be used quite often. Particularly at the compressor cylinder flanges?

The pressure drop across an orifice plate is well known, but if there is an orifice plate at a flange between the cylinder and then a bottle with some kind of internals is that one device.

For the interstages of a multistage compressor, how is Eq 5 applied? It asks for the stage ratio which may be different for the discharge side of the first stage and the suction side of the next.

I understand the intent of the allowable pressure drop, but its implementation is still a bit fuzzy

 

[sms]

In the past I have applied the allowable to each device, and typically included the orifices (either on the cylinder flange or the outlet nozzle) with the vessel.

But just for clarification I have not done a analog design in many years, so API 618 has probably changed a bunch..

When you do an analog and you throw an orifice on the cylinder flange, it often improves the cylinder performance significantly.

"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[PulseMaster]

I am in the pulsation control business, and here is the truth of the matter.

1. Pressure drop through orifice plates under steady flow is easy to quantify, while pressure drop at the nozzle orifice requires more rigorous calculation that not every pulsation study provider does.

2. Pressure drop through a choke tube is relatively easy to quantify.

3. Pressure drop through an empty bottle is somewhat insignificant.

4. Dynamic pressure drop, when operating at or near resonance can be much higher that any static pressure drop in the manifold system.

5. Nozzle responses of large magnitude hurt your cylinder performance badly. It is not uncommon to put in a nozzle orifice and see your performance improve.

We generally add the choke tube and orifice pressure drop for a single flow path, for inlet or outlet of a single stage and compare that to API 618 allowables. Equivalent pressure drop, as outlined in API 618, can be used for multi-cylinder systems, and makes more sense.

In the future version of 618, dynamic pressure drop calculations will be a requirement. This concept will be interesting to explain to clients.

Be glad to explain further if you are interested.

 

[Gordo2020]

Do you have any papers or references on calculating the dynamic pressure loss through an orifice? It is interesting to hear that they can be more then all the other static losses

 

[PulseMaster]

No papers that I know of right offhand, but there may be some. Just some proprietary field and experimental data. Predicting dynamic pdrop effects is another matter. You might see some papers on it in the next few years.