×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Losses through long annulus.

Losses through long annulus.

Losses through long annulus.

(OP)
Hello,

  If I have an annulus that has a large ratio of annulus radius to annulus gap (meaning the gap is small compared to the radius) what is the method of calculating K?

I am assuming that K = f (L/D) where D is hydraulic diameter.

The problem I see though is that the flow area is not well represented with hydraulic diameter and this will be problematic if i use equation 3-20 of cranes because it requires a value for "D^4". And D is hydraulic diameter which may be many times smaller than what is represented by the actual flow area. Do you see what I mean?

 Thank you!!

 

RE: Losses through long annulus.

That's exactly why hydraulic diameter is used.  With a very very small annular space, you can see immediately that you have doubled the frictional surface area of the annular flow.

Start with a 10" ID pipe and another pipe inside with an interior radius of 0, then begin increasing the interior radius to almost 10".  Annular flow contact surface (friction) increases as a function of the square of the interior radius, while at the same time flow area decreases by the square of the interior radius.  Thus, trying to maintain the same flow Q, velocity Q/A is being squared and friction is being squared again, so looks like D^4 pretty well covers the expected effect on head loss in an annular space.

**********************
"Pumping accounts for 20% of the world's energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies) http://virtualpipeline.spaces.live.com/

RE: Losses through long annulus.

(OP)
Thank you for the explanation.

So is it safe for me to use Hydraulic diameter in equation 3-20 as Dh^4? Even if my Dh = 1" and my inner and outer radii can be on the order of 100+ inches (very large flow area but small hydraulic diameter Dh=1")?

 

RE: Losses through long annulus.

(OP)
Thank you for the useful material.

But I am still confused about something,

I made a quick spreadsheet on annulus geometry and hydraulic diameter.

I start with a D=100" ID outer pipe and work my way up from D=0" inner pipe back up to 100".

My results are the following


Dout    Din     Aactual  Dequiv  Dh      A(Dh)
100    0    7853.98    100.00    100.00    7853.98
100    5    7834.35    99.87    95.00    7088.22
100    10    7775.44    99.50    90.00    6361.73
100    15    7677.27    98.87    85.00    5674.50
100    20    7539.82    97.98    80.00    5026.55
100    25    7363.11    96.82    75.00    4417.86
100    30    7147.12    95.39    70.00    3848.45
100    35    6891.87    93.67    65.00    3318.31
100    40    6597.34    91.65    60.00    2827.43
100    45    6263.55    89.30    55.00    2375.83
100    50    5890.49    86.60    50.00    1963.50
100    55    5478.15    83.52    45.00    1590.43
100    60    5026.55    80.00    40.00    1256.64
100    65    4535.67    75.99    35.00    962.11
100    70    4005.53    71.41    30.00    706.86
100    75    3436.12    66.14    25.00    490.87
100    80    2827.43    60.00    20.00    314.16
100    85    2179.48    52.68    15.00    176.71
100    90    1492.26    43.59    10.00    78.54
100    95    765.76    31.22    5.00    19.63
100    100    0.00    0.00    0.00    0.00


What we see here is that the actual area (Aactual) of the annulus and the area due to hydraulic diameter (ADh) diverge GREATLY.

On a purely flow area basis, the flow rate CANNOT be correct if hydraulic diameter is used for a very short annulus (low friction). Since friction for a short annulus is negligible and the flow rate is almost entirely dependant on the flow area, hydraulic diameter used in Eq 3-20 for a short annulus would give me an innaccurate solution. Please review this and get back to me as I am curious about this topic.

 Thank you

RE: Losses through long annulus.

(OP)
All,

 I have figured out a way to use Eq. 3-20. I use Equivalent Diameter instead of Dh.

Equivalent Diameter is the actual area times 4 over pi. This is going to give me a pseudo pipe diameter to use for Eq. 3-20.

I verified the legitimacy of this method with Flowmaster V6.5.2. Here are the results.

K    Y      w EquivD     w Dh    Flowmaster
2    0.69    1230.72    31.56    1097.47
10    0.81    641.42    16.45    626.495
20    0.83    465.76    11.94    463.189
40    0.84    332.77    8.53    335.657
100    0.84    212.38    5.45    215.594

w EquivD and Flowmaster are 12% different with low K values, this is expected because flowmaster probably uses some entrance and exit effects on the component whereas I am using purely K = 2 and Eq 3-20.

When K increases, both flowmaster and the EquivD method converge and become more accurate, because now friction is dominant and it dilutes the innaccuracy of the entrance effect.

I think this seems valid, no?

  Thank you,
 

RE: Losses through long annulus.

You mention equation 3-20 of cranes.  My version of Crane Technical Paper 410 "Flow of Fluids Through Valves, Fittings, and Pipe" (or the Metric version, which is Technical Paper 410M) has on the bottom right-hand side of page 3-5 a note which says "See page 1-4 for limitations."  When I go to page 1-4 and the discussion on hydraulic radius it says "This applies to any ordinary conduit (circular conduit not flowing full, oval, square, or rectangular) but not to extremely narrow shapes such as annular or elongated openings, where width is small relative to length.  In such cases, the hydraulic radius is approximately equal to one-half the width of the passage."

If you were looking at page 3-4 (where equation 3-20 resides), it's a shame you couldn't have read one page over...


Patricia Lougheed

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

RE: Losses through long annulus.

(OP)
I understand the limitations. Which is why I did the calculations, if you read those you will see that hydraulic diameter nor hydraulic radius cannot be used in Eq 3-20 and I am resorting to augmenting D with an effective area or an equivalent diameter. I just want to know the validity of this, if anyone has done this?

That is what my inquiry is about, calculating flow rate.

Yes, page 1-4 says hydraulic radius is equal to one half the width of the passage. This still doesnt answer my question as to what my value of D should be in Eq 3-20 (should not be hydraulic diameter). Actually, I sort of answered my own question.


Yes it is a shame I didn't read one page over, but I am not an imbecile.

   Thank you,

 

RE: Losses through long annulus.

dbecker, wouldn't your equivalent diameter be the square root of (area x 4/pi) ?

Ted

RE: Losses through long annulus.

I agree that for flow in an annulus the hydraulic diameter will give wrong answers if a pipe of a diameter equal to the hydraulic diameter is substituted for the annulus.  The "secret" of how to apply the hydraulic diameter is hinted at in the last paragraph on page 1-4 in Crane 410.  The diameter used for calculating the equivalent K value for the pipe (i.e. K = f(L/D) ) and the diameter used for calculating the fluid velocity are NOT the same.

An important factor in determining the pressure drop along a pipe is the ratio of the cross sectional area to the wall area because it is the wall that provides the friction.  The hydraulic diameter concept provides a value for a regular round pipe that would have the same ratio of cross sectional area to wall area as the particular annulus. The hydraulic diameter is therefore the right thing to use to calculate the K value for the annulus.

However, this does not mean that we should regard all the fluid flowing down the annulus as flowing through a pipe of the hydraulic diameter.  The average velocity in the annulus is easily calculated, and it is this velocity which should be applied to the K value calculated for the pipe of equivalent hydraulic diameter.  If you want to use a standard pipe pressure drop spreadsheet or program to calculate the pressure drop then simply divide the cross sectional area of the annulus by the cross sectional area of the pipe with the equivalent hydraulic diameter to get the fraction of the flow that should be used in the calculation.  This is difficult to put into words but will be obvious in the example below.  Another way of looking at the annulus is as a bunch of pipes in parallel, where each pipe has the hydraulic diameter of the annulus.

I believe that calculating an equivalent pipe diameter on the basis as having the same cross sectional area as the annulus (as per your example earlier) will significantly under estimate the pressure drop because the wall area of such a pipe would be much less than that in the annulus.

As an example I will continue with the example you have already started.  I took the outer pipe as 100" ID and the inner pipe as 99" OD.  To avoid the complication of the Y factor I have assumed the fluid to be water flowing at 6 ft/sec in the annulus.  The annulus has a cross sectional area of 156.3 inch2 and a wetted perimeter of 625.2 inch. At 6 ft/sec this gives a flow rate of 2920 USgpm.  In all the calculations below I have assumed the absolute roughness of the pipe is 0.002 inch.

Alternative 1 - Pipe of same cross sectional area as annulus
A standard round pipe of 14.1" ID would have the same cross sectional area as the annulus.  Water at 6 ft/sec gives Re = 655000,   fM = 0.01446,   Pressure drop = 0.2985 psi/100ft.

Alternative 2 - Pipe of Hydraulic Diameter
The hydraulic diameter of the annulus is 1", and a pipe with this diameter will have a cross sectional area of 0.7854 inch2.  It would require 156.3 / 0.7854 = 199 pipes of this diameter to provide the same cross sectional area as the annulus.  We can regards the flow in the pipe as being 2920 / 199 = 14.67 USgpm.  This gives Re = 46400,  fM = 0.02663,   Pressure drop = 7.73 psi/100ft.

Alternative 3 - Pipe of Equivalent Diameter according to "Petroleum Engineering" method
This method is described by ZDAS04 in FAQ378-1142: Flow in Annular Space.  De = (1992 x 13)0.2 = 8.3 inch.  Using the full flow of 2920 USgpm this gives Re = 1113000,  fM = 0.015,   Pressure drop = 4.4 psi/100ft,  Velocity = 17.3 ft/sec.


Comparing these three alternatives I would regard the first one as totally unrealistic.  But I find it amazing that the third method is able to predict an equivalent diameter that gives a value that is usable for calculating a friction factor and a velocity that gives an answer so close to alternative 2.

Some other references (eg Transport Phenomena by Bird, Stewart and Lightfoot) give another way of calculating the equivalent diameter.  This is De = R x (1 - k) where R = Outside diameter of inner pipe, and k = Outside diameter of inner pipe / Inside diameter of outer pipe.  If this diameter is used in the same way as Alternative 2 above, very similar results are obtained as with the hydraulic diameter.

Katmar Software
Engineering & Risk Analysis Software
http://katmarsoftware.com

RE: Losses through long annulus.

Annular gas flow is a ubiquitous phenomena in Oil & Gas.  I've measured actual dP in a lot of gas flows and the "Petroleum Method" that I describe in the FAQ always gets calculated values acceptably close to measured values.

Let's move into a world where you can purchase pipe (never seen 99" diameter pipe).  Let's say you have a 12" pipe (OD 12.75") inside a 16" Sched 30 pipe (ID 15.250").  Let's assume that the pressure at the head of 1 mile of pipe is 100 psia and we're flowing 1 MMCF/d.
 MMCF/d.

The wetted perimeter is 7.33 ft.  This is the key factor, but it isn't simple.  The flow at the surface of the pipe is zero.  More surface area means more of the flow is subject to surface friction.  It also means that for a given mass flow rate, the velocity will be less than for a round pipe with the same cross sectional area.  Lower velocity means less friction.  So the goal needs to be to find a pipe diameter that can be used in the standard flow equations that balance these competing factors.

Using simple difference in cross sectional area, you would get a 8.367 inch diameter and the wetted perimeter would be 2.19 ft.  The downstream pressure (using the AGA equation and a pipe roughness of 0.0015) would be 99.9 psia

Using the Hydraulic Diameter, the effective diameter would be 2.5 in, and the wetted perimeter would be 0.654 ft.  Downstream pressure would be 54.3 psia.

Using the Petroleum Method, the effective diameter is 6.571 in, and the wetted perimeter is 1.72 ft. Downstream pressure would be 99.754 psia.

None of the three calculations is anywhere close to the actual surface area, but the simple difference in area will understate dP, Hydraulic diameter will overstate it, and the Petroleum Method will be the closest for every dP I've ever measured.

Nobody has ever built this line or anything like it, but I would be very confident designing it using the Petroleum Method, in this case using the simple difference method would probably be ok, using the hydraulic diameter would be dumb.

I would expect liquid-full pipes to act like a gas (i.e., the Petroleum Method should match measured data closest).  Partially full pipes will always behave different from gas.

David Simpson, PE
MuleShoe Engineering
www.muleshoe-eng.com
Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips Fora.

"Everyone is entitled to their own opinion, but not their own facts"  Patrick Moynihan

RE: Losses through long annulus.

(OP)
Great info guys, give me some time to review these methods.

Thanks for the support!!!!



 

RE: Losses through long annulus.

(OP)
In response to hydtools, yes you are right it is square rooted!! Typo on my part on my previous post.


  Thanks!

RE: Losses through long annulus.

Hydraulic diameter is defined as the flow area divided by the wetted perimeter.
In the case of an annulus: hydraulic diameter = (D^2-d^2)/4*(D + d) or (D-d)/4.

Area = (pi/4)*(D^2-d^2)
Wetted perimeter = pi*D + pi*d
Also D-d = diametral clearance
Hydraulic diameter = 1/4 * diametral clearance = 1/2 * radial clearance

Ted

RE: Losses through long annulus.

Not according to my reference books (see Lindberg, page 36-8 for example).  The equation you referenced is the one that Patricia pointed out that on Page 1-4 of Crane 410 that it is not applicable to annular flow.  Lindberg referenced the Nusalt derivation to come up with:

HydDiam = 4{area)/(wetted Perimeter)

Area = {pi/4)(D^2-d^2) = (pi/4)(D-d)(D+d)

Wetted Perimeter = pi(D+d)

HydDiam = 4(pi/4)(D-d)(D+d) / [pi(D+d) = D-d

If you use the Area/Perimeter formulation then the area (which the Lindberg formulation results in a diameter that is far too small) is 1/4 as large.

Since the Hydraulic Diameter method is ubiquitous, I would guess that in partially full ducts it might have real value.  In totally full annular spaces it is worse than worthless.

David

 

RE: Losses through long annulus.


The definition by hydtools corresponds to the hydraulic radius, not the hydraulic diameter.

RE: Losses through long annulus.

zdas04 - thanks for the input and your more feasible example. I haven't seen any 99" pipe either!

The point I was trying to make is that when using the Hydraulic Diameter method it is wrong to assume that the full flow is flowing through a standard round pipe of diameter equal to hydraulic diameter.  This gives nonsense results, as you have shown. It is necessary to use the actual velocity in the annulus together with the K (=fL/D) value from the hydraulic diameter.

Applying this to your example with 1 MMSCF/d - the cross sectional area of the annulus is 54.98 inch2 and the cross sectional area of the 2.5" pipe (i.e. hydraulic diameter) is 4.909 inch2.  This ratio is 54.98/4.909 = 11.2.  The flow through the pipe should therefore be taken as 89285 SCF/d (=10^6 / 11.2) to give the same velocity as in the annulus.  On this basis the downstream pressure is 99.63 psia which is much more reasonable.

The flow of 1 MMSCF/d gives a velocity in the annulus of only 4.4 ft/sec so the pressure drops are small and differences between the methods are not so easily seen.  If we increase the flow to 5 MMSCF/d the pressure drop using the Petroleum Engineering method would be 6.1 psi and the pressure drop using the Hydraulic Diameter method (assuming flow is 446400 SCF/d (=5x10^6/11.2) ) is 7.8 psi.  This is a small difference and I believe both are realistic.

From a theoretical standpoint I prefer the Hydraulic Diameter method because is has the same cross sectional area to wall area ratio as the annulus, and the fluid velocity is the same whereas the Petroleum Engineering method is completely empirical (as far as I know).  If you have any real life examples where you could test this it would be very interesting.

Katmar Software
Engineering & Risk Analysis Software
http://katmarsoftware.com

RE: Losses through long annulus.

hydtools and 25362 - my references also use this definition for hydraulic radius, and then define hydraulic diameter as 4 times the hydraulic radius.

Katmar Software
Engineering & Risk Analysis Software
http://katmarsoftware.com

RE: Losses through long annulus.

Perry's Chemical Engineers Handbook refers to the hydraulic radius/hydraulic diameter concept but lists absolutely no caveats in relation to annular flow, unlike the Crane paper already listed.  

Perry's claims that the use of the hydraulic diameter in the Fanning or Darcy equation F= 4fL/D is valid for ALL long, noncircular channels in turbulent flow, full or not (as long as variation in height is negligible).

Perry's makes no reference to how one is supposed to calculate the velocity, and hence the Reynolds Number, in order to calculate f.  That's the crux of the issue.

Perrys lists its underlying references, but I don't have easy access to them:  Brightton and Jones,  J. Basic Eng. 86, p835-842, and Lawn and Elliot, J.Mech.Eng. Sci., 14, 195-254.  Does anybody have or have access to these, to determine which velocity they're using? Or am I, as usual, missing something obvious?!

RE: Losses through long annulus.


Perry's ChE Handbook VIII stresses the difference between hydraulic and equivalent diameters when dealing with non-circular channels, saying that

Quote:

the hydraulic diameter method does not work well for laminar flow because the shape affects the flow resistance in a way that cannot be expressed as a function only of the ratio of cross-sectional area to wetted perimeter.
It also gives formulas for equivalent diameters for ellipses, rectangles and annulus, as well as references for regular polygons.  

RE: Losses through long annulus.

...but makes no such distinction for TURBULENT flow, at least not that I've seen- and I re-read the section recently.

RE: Losses through long annulus.


Not so, for turbulent flow in rectangular ducts of large aspect ratio, the manual suggests the value of the friction factor from the Blasius formula for 4,000<Re<105 to be increased by about 10%.

It also says that the ΔPf in noncircular channels is a bit larger than the one estimated by the hydraulic diameter method due to secondary flows perpendicular to the axial flow direction.

This makes one wonder whether other methods like the Petroleum Method exemplified above wouldn't be underestimating ΔPf values.    

RE: Losses through long annulus.

If we're only talking about the 10-15% difference in the friction factors which IS noted in Perry's, then I'm perfectly happy to keep using the hydraulic diameter method.  Remembering that these are all correlative methods, I have a dim view of their precision and design accordingly.

katmar:  it sounds like that is what you've concluded, comparing the "petroleum method" to the hydraulic diameter method using the correct velocity, you get pretty good agreement.  Did I sum that up correctly?

We're no longer talking about, as was implied in the previous discussion, a total non-applicability of the hydraulic diameter method to annuli- we're just clearing up the difference between effective/equivalent diameter of circular pipe, and the hydraulic diameter used in the Fanning equation.  Correct?  The two are clearly different.

As katmar points out, the correct velocity to use with the hydraulic diameter method is the actual velocity in the annulus, ie. the actual volumetric flow divided by the true cross sectional area of the annulus.  The only non-circular channel example in Crane uses the actual velocity if I recall correctly.

If we're talking about really narrow annuli, it would appear that the system reduces to flow between wide parallel plates, which should be pretty well understood both in laminar and turbulent flow conditions.

RE: Losses through long annulus.

(OP)
Thank you for the helpful information.

I have studied your responses.

I may end up using the Petroleum method.

The reason I CANNOT use hydraulic diameter is because i am calculating flow rate. I do not know flow rate at this point. I only know upstream, downstream pressure and pipe geometry (annulus).

Therefore, hydraulic diameter is not suited for this calculation.

Hydraulic diameter is better suited for calculated dimensionless numbers like Reynolds number for an annulus, because it is geared more towards how turbulent the flow is versus what the flow rate should be.

The flow rate of an annulus is not well represented by hydraulic diameter because of the simple fact that the cross sectional area of Dh is going to be very small compared to actual Dh, which will yeild smal flow rates when the annulus is short (low friction).

I do believe that when the annulus gap becomes smaller, we could consider that system as two parallel plates with width = 2 pi (R) and gap = radial gap.

I am going to run some calculations, stay tuned.


    

RE: Losses through long annulus.

If you are trying to calculate the flowrate you will need to do it by trial and error - irrespective of whether you use the hydraulic diameter or the Petroleum Engineering method.  You have to guess a flowrate and calculate the pressure drop, compare it with the known pressure drop, and repeat the calculation with an adjusted flowrate until the calculated pressure drop matches the actual. I described this process more completely in thread378-235298: water flow rate @ gauge pressure

Katmar Software
Engineering & Risk Analysis Software
http://katmarsoftware.com

RE: Losses through long annulus.

(OP)
katmar,

 I am aware of this.

 I am designing software to do these iterations for me.

But before I can do that I have to make sure all my variables for equation 3-20 of Crane's make sense with what I am trying to calculate.


Thanks,

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources