"Ten diameters" of pipe upstream of a flow meter? 2

[nomorenames]

There's a rule of thumb that says a flow meter requires ten diameters of straight pipe upstream, and five diameters of straight pipe downstream in order to provide "clean" flow at the meters inlet/outlet. This rule is often printed explicitly in the installation manuals for commercially produced flow meters.

Question: will these lengths of straight pipe permit me to do whatever I want to the flow at positions further upstream/downstream of those limits? Or are these lengths only intended to cope with the typical minor flow perturbations introduced by simple elbows and/or manifolds?

 

[Artisi]

Are you asking about positioning the flow meter at a greater distance than 10/5 from a flow disturbance?
If yes, there is no problem as the recommendation of 10/5 is only to ensure steady flow before / after the measuring point.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[nomorenames]

>Are you asking about positioning the flow meter
>at a greater distance than 10/5 from a flow disturbance?

The specific case I'm thinking of is a flow meter with ten diameters of straight pipe upstream, with the upstream end of that pipe (the inlet) open to atmosphere. I'm wonder what sort of obstructions could be tolerated at that inlet while still providing clean flow to the meter.

 

[zdas04]

The genesis of these lengths in the code is (as most code language) a compromise. The [wishful] thinking was that fluid friction would tend to dampen swirl and correct asymmetries in the flow. There have been many experiments since the 1980's that show that ten times ten diameters will do little to affect a persistent swirl or shift an asymmetric flow towards the center of the pipe. This research has led to some very impressive flow conditioners that really work.

So to answer your basic question, the code allows any piping configuration upstream of the straight pipe requirements.

I don't understand your second post. If the upstream end is open to the atmosphere, the only way you would have any flow is to pull a vacuum on the downstream end, and vacuum is one of the cases where the AGA equations are invalid.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. ùGalileo Galilei, Italian Physicist

 

[Artisi]

Agreed, the second post needs explaining.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[nomorenames]

>Agreed, the second post needs explaining.

OK, here's what I've got:

18" diameter duct. 180-inch-long straight section, upstream end open to atmosphere, downstream end connected to flow meter. Downstream of the meter, there is a 90-inch-long straight section, and then there's a blower, pulling air through the duct.

I'm wondering what sorts of obstructions can be tolerated near the inlet end of this system.

 

[bimr]

The rule of thumb are minimum values for minor flow perturbations, not for major disturbances such as control valves.

For your application, there are a number of unknowns:

[ol 1]
[li]Pipe size[/li]
[li]Velocity[/li]
[li]Type of pipe entrance[/li]
[li]Type of flow meter[/li]
[li]Meter accuracy required[/li]
[li]Pipe orientation[/li]
[li]Air in fluid[/li]
[/ol]

The rule of thumb are minimum values, increasing these dimensions will often improve the flowmeter performance (accuracy). Where two or more disturbances are located upstream of the meter, the longest recommended upstream pipe run is to be observed as an absolute minimum. Wherever possible mount any control and butterfly shut off valves downstream of the flowmeter.

 

[hacksaw]

Putting an orifice on the inlet of a blower is a bit off.

Just what level of accurracy are you after?

 

[Artisi]

Are discussing air or water?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

First question - depends on your type of meter and what sort of accuracy and repeatability you want from your meter. +/ 5 to 10% you can probably do pretty much whatever you want - +/- 1%(fiscal) you need smooth flow, probably via a flow conditioner of some sort (packed tubes or flow conditioning plate).

don't really understand why the meter is on that side of the blower - you're either going to introduce more losses in a system which doesn't have much to spare or result in lower flow velocities / bigger diameter meter, neither of which is normally good news.

Swirl, pulsation and vorticies are normally your enemies in this instance so reduce / eliminate those and you're going in the right direction.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[zdas04]

The piping configuration you described will not give you ±10% uncertainty or repeatability. At best it will tell you if there is flow or not, but I don't think I would rely on even that very far (you could do that better by monitoring the position of a weighted flapper). This is absolutely not the place where square edged orifice measurement should be used.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. ùGalileo Galilei, Italian Physicist

 

[dbill74]

You can put whatever you want on the end. Louvers with plenums and screens are used all the time. Being open to atmosphere you need to protect your duct, meter, fan from unwanted materials such as insects, birds and rain.

 

[bimr]

Inlet filters are required to keep foreign material from entering the blower machine. Abrasive, wet or bulky material will damage the machine’s internal parts and could cause catastrophic failure. Due to the many types of filters, it is not practical to give specific instructions covering all models. In general, it is recommended to clean or replace filters when the pressure drop across the filter reaches 10” wc.

http://www.nyb.com/Catalog/Supplements/CS-491.pdf

 

[nomorenames]

I had thought I posted again with more details, but I can't find it, so I'll try again. Sorry for not putting all of this in the OP.

The flow measurement device is a smooth-approach orifice. The throat is perhaps 7" in diameter (don't recall exactly), and the main duct is 18" in diameter.

The fluid is air.

Upstream of the SAO is 180" of straight pipe, 18" diameter, with the upstream end of that pipe open to atmosphere. No louvers, filters, or anything like that.

Downstream of the SAO is 90" of straight pipe, 18" in diameter, with the downstream end connected to a blower that pulls air through the whole system.

Flow velocities are on the order of 20 feet per second.

The issue is that there is a stationary object near the inlet end of the pipe. Perhaps a foot upstream, and flush with the outer edge of the pipe.

Question: is this object likely to cause flow perturbations that will persist all the way to the SAO and influence its accuracy, or can I count on the 10 diameters of pipe between the inlet and the SAO to produce axisymmetric flow at the SAO, regardless of what's happening at the inlet?

 

[dbill74]

If louvers and filters are not going to be a problem, I really don't think a stationary object is an issue. Unless perhaps said object is a wall. But then that is what the 10x diameter rule is for.

 

[GHartmann]

That is an unusual flow measurement arrangement. The orifice is going to create a significant pressure drop in your suction line relative to the line friction loss. I would have selected some type of annubar arrangement; one of those that span the pipe diameter and have multiple holes.

http://www2.emersonprocess.com/en-U...-Products/Annubar-Flowmeters/Pages/index.aspx

You have abided the 10 diameters rule, but your flow accuracy won't be that good

 

[BigInch]

If custody metering, go for minimum of 20D upstream and 5 downstream. 15D upstream if you use flow straighteners. Better yet, give Emerson-Daniel a call.

http://www2.emersonprocess.com/en-US/brands/daniel/metering-systems/Pages/Metering-Systems.aspx

[img =

https://www.dropbox.com/s/r9mygqm70sm4hx5/OMG something else.png?dl=0

 

[BigInch]

finally

 

[bimr]

It appears that you are after extreme accuracy, so the obstruction will most likely affect your flow measurement. Here is a link to a discussion on orifice plates where high metering accuracy is required.

"The minimum recommended length of unobstructed straight pipe varies from 13 to as much as 145 pipe diameters (internal pipe diameters) upstream of the orifice plate and 4.5 pipe diameters downstream of the orifice plate (Figure 8)."

http://asgmt.com/wp-content/uploads/pdf-docs/2009/1/002.pdf

 

[if7005]

For orifice type, not always 10D upstream or 5D downstream for FO placement, it depend on piping run length, D/d (pipe bore/orifice hole), flow conditioner, and piping end configuration. As well as, any obstruction not allowed to be placed inside meter tube (upstream + downstream ) except only for metering part. This is referred to AGA report-3 part-2, please check it