16 replies to this topic

[ianmcq28]

Hello,

 

This topic has perhaps been discussed many times and I have found some useful guidance here:

 

http://www.aft.com/d...low-Reprint.pdf

 

I would greatly appreciate some discussion to help me to better understand how to deal with choked flow in pressure drop / line sizing calculations.

 

Background:

 

I am sizing a blow-down line on a natural gas pipeline for maintenance blow-down of pipework and equipment.  I have sized a restriction orifice plate in order to give me a suitable time to achieve depressurization but also to limit the extent of the zone required for hazardous area classification so that the zone does not extend beyond the boundary fence for abnormal operation of the valve.

 

The system will normally be used for blow-down of a fixed volume during maintenance routines, but there is always a possibility that the vent is unlocked and used abnormally, so I am basing my calculations on the peak flows at maximum system operating pressure.

 

There are a number of other vents, process vents and relief valve vents and I am intending to tie-in all of the laterals from each source into a common vent header with release to atmosphere.

 

Problem 1:

 

Focussing for now just on the high pressure maintenance vent lateral design:

 

I initially I performed my blowdown calculations by treating the system as a simple orifice, but now I need to size the lateral and determine maximum system pressures in the lateral to specify design pressures.  I will be installing a flame arrestor downstream of the orifice and so need to provide operating pressures and design conditions.

 

The process conditions are as follows:

 

Fluid: natural gas (Mw = 18.5)

Z = 0.92 at upstream condition

Maximum upstream system pressure: 33.8 bara

Upstream temperature: 10ºC.  Joule thomson effect ignored for initial assessment

Orifice size: 6mm NB

Vent lateral size: Initially guessed to be 1" NB, but 2" NB likely to be selected to prevent additonal choking in the lateral piping

Common vent header size 4" NB

Common vent header disposes to atmospherice pressure

 

For the orifice calculation, I determined that the flow is choked and therefore the downstream pressure from the orifice must be equal to the critical pressure, which I calculated to be 18.45 bara.  For the given upstream pressure the maximum choked flow through the orifice is ~500 kg/hr.

 

I have built a compressible flow pressure drop spreadsheet (that I have divided into many segments to create a network for all fittings, area changes and junctions etc).  Mach no. is calculated for the inlet and outlet of each segment to determine where the choke points will occur.

 

Method 1: If I work through the pressure drop calculation with upstream pressure (downstream of orifice) set at 18.45 bara the resulting outlet pressure (release to atmosphere) will not be atmospheric (for 1" or 2" pipe sizing).

 

For this calculation the flow was based on the maximum orifice choked flow. 

 

Method 2: If I work backwards ie from the vent outlet at atmospheric pressure, using the maximum orifice flow again, I can calculate the upstream pressure (downstream of the orifice) and this is significantly lower that the critical pressure, approx 3.5 bara for a 1" pipe and 1.2 bara for a 2" pipe.

 

Question 1:  which method is correct to determine the operating pressure for the flame arrestor?

 

Discussion:

 

The reference document (link at top) states:

 

'the pressure drop across the shock wave in choked flow cannot be calculated directly. The only recourse is to use the choked flowrate as a new boundary condition on the pipe downstream of the shock wave.'

 

I have used the choked flow as the boundary condition (or limiting mass flow into the vent lateral), but I am still struggling with how to deal with the pressure discontinuity.

 

Question 2: f I was to determine the maximum operating pressure in the vent pipework downstream to determine a suitable design condition, do I state that the maximum operating pressure in the vent is equal to 18.45 bara or equal to 3.5 bara (or 1.2 bara if 2" lateral used)?

 

Problem 2:

 

As previously mentioned the HP vent lateral will tie-in to a common header.  If I size the lateral in 1" NB further choking will occur at the reducing T-piece adjoining header.  By installing 2" I can avoid this.

 

For my own interest I would like to understand the implications (for if I assumed the vent lateral is 1") and again understand how I deal with the pressure discontinuity that will occur at the reducing T-piece junction with the common 4" header.

 

Question 3: Does the shock wave caused by choking dissipate instantly at the junction or could this continue downstream for a given (unknown at this stage) distance and affect the back-pressure in the common header?

 

Again, if I calculate back-wards from the vent outlet to atmosphere, the calculated back-pressure at the junction is significantly lower and does not cause too much back-pressure to the relief valves connected to the system.

 

Any advise or discussion would be greatly appreciated.

 

Kind Regards,

 

Ian

[Bobby Strain]

You should be concerned about noise, too. So you should consult with your suppliers to provide the proper device.

 

Bobby

[ianmcq28]

Thanks Bobby,

 

There will be a number of design checks for us to consider and we will certainly consider noise limits.

 

Just to expand on my approach to the sonic flow calculations:

 

I have always considered method 2 to be the correct approach. 

 

It is generally quite common to have critical flow over an orifice or relief valve and for a relief valve the back-pressure to the relief valve is calculated from the vent outlet and I have always sized the outlet piping so that there are no other choke points (or even close to mach 1). 

 

When my manual vent sizing calculations identified a number of choking points it got me thinking and I just want to understand exactly what happens to the sonic shockwave.  Does the shock-wave (and pressure?) 'recover' as soon as the flow leaves the vena contractor of the orifice?

 

Thanks,

 

Ian

[breizh]

Hi, 

you may find information to support you query in the link attached :

 

http://www.engsoft.c...team_flow_e.htm

 

Breizh

[latexman]

Excellent reference Breizh!  As a ChE student, I did not immediately realize while at University (1975-1979) that "in case of orifice, actually there is no critical pressure", and, therefore, "there is no choked flow in orifice flow".  I learned this subtle fact only after working in industry for some years.  I cannot really recall if I was just oblivious to this fact while in college, or if it was not covered in detail in my courses.  No matter, because I know it now.  I would like to ask Ian though, was this made clear in your college education?

Edited by latexman, 17 July 2013 - 10:30 AM.

[chemsac2]

Ian,

 

I am sailing in the same boat. Just yesterday I came across an article from July 2007 issue of Chemical Engineering Progress. Article titled "Designing multiple-diameter relief piping" by Gravin, J. and O'Shea, J.D. seems to give steps on estimating flowrate in systems with multiple chokes.

 

I am yet to explore it in detail. Would get back to this post when I am through with it.

 

Regards,

 

Sachin

[ianmcq28]

Thanks Breizh I will digest the information.

 

Sachin, I will welcome your response thank you.

 

Latexman, If I completely honest I do not recall even doing the simplest compressible pressure drop calculations at University!  In fact there isn't much that I can recall that I can put to practical use!  There are plenty of topics, long derivations from first principles and mathematics that I have never had any reason to even look up though!  All of what I have learned is based on my year in industry and learning on the job using industry standard methods.  We literally left university not even knowing what a valve looked like! I wonder if this is a sign of too many academics and not enough real industrial experience at universities?

 

Thanks,

 

Ian

[fallah]

"in case of orifice, actually there is no critical pressure", and, therefore, "there is no choked flow in orifice flow". 

 

Latexman,

 

You would certainly verify that the orifice with thick plate is an exception because due to it can cofine the vena contracta by its orifice area, can choke the passing flow...

[latexman]

fallah,

 

Good point.  Yes, absolutely, a confined vena contracta will choke.  An orifice whose thickness extends to or beyond the vena contracta will choke.

[breizh]

A bit more ,

 

Breizh

[sanganipratik]

Hello Ian,

I'm sailing, today, in the similar situtation,

Were you able to find any solution and which method (1 or 2) was finally adopted?

 

Case: I have a gas swivel seal leak line of 3/4" which is connected to the LP flare header and the pipe line is fitted with an orifce and downstream to orifice is a flowmeter.

Under normal operation there shouldnt be any leaks and if there is it should be very minor @acceptable range.

 

If the seal fails, then the line see a an upstream gas pressure of 170 barg, which is the gas swivel operating pressure. An orfice is selected, inorder to safe guard the downstream LP header which operates at 1.5 barg.

 

Now the question is the significance of the choke flow/ crtical pressue. As obviously i see that under leakage condition the orifice sees a choked flow @ 97 barg critical pressure. So should I design my downstream piping and the flowmeter for pressure of 97 barg or 1.5 barg.

 

Your suggestion is appreciated.

Pratik

[serra]

the usual procedure to identify critical flow conditions in piping is to start from releasing point (where you know the pressure) 

and then calculate conditions and mach numbers on each segment,

those (max pressures and temperatures) are required to verify mechanical design (design conditions must allow these),

with Excel I use the=PIPE() macro available in Prode Properties,

the procedure is iterative,

other simulators and tools for flare piping networks adopt similar procedures.

[yazdanpanahimehrdad]

Dear All;

I have two question about above discussion

1- what about control valves? Does the chocked flow happen in the control valves or they act as an orifice?

2- In my understanding when chocked flow happens we should calculate downstream pressure of the chocked device based on the back calculation. it means that pressure of downstream may fall below critical pressure but flow will not increase (maximum flow is based on the critical pressure). is my understanding right?

Edited by yazdanpanahimehrdad, 14 November 2014 - 07:34 AM.

[serra]

you may find the details in ISA procedures or ISO standards ,

IEC 60534 (noise prediction) shows how to calculate jet diameters etc.

 

To understand critical flow consider a important parameter, the speed of sound,
pressure waves cannot travel faster than speed of sound,

from that and a law to describe fluid (for example ideal gas)

you derive all required formulations.
 

[yazdanpanahimehrdad]

Thanks serra

[flarenuf]

hi

to get back to the original question

Method 2: If I work backwards ie from the vent outlet at atmospheric pressure, using the maximum orifice flow again, I can calculate the upstream pressure (downstream of the orifice) and this is significantly lower that the critical pressure, approx 3.5 bara for a 1" pipe and 1.2 bara for a 2" pipe.

 

Question 1:  which method is correct to determine the operating pressure for the flame arrestor?

 

method 2 is correct , this is how Flarenet (AFSA) works .. tip to source .

in effect the backpressure on the Flarme arrestor is 3.5 barg , your choked flow if any will be upstream of that at the crit pressure.

flarenuf

[yazdanpanahimehrdad]

thanks alot for your clarification. I think now I get my answer